You are on page 1of 436

Essentials of

Personality Disorders
This page intentionally left blank
Essentials of
Personality Disorders

Edited by

John M. Oldham, M.D., M.S.

Andrew E. Skodol, M.D.
Donna S. Bender, Ph.D.

Washington, DC
London, England
Note: The authors have worked to ensure that all information in this book is accurate at
the time of publication and consistent with general psychiatric and medical standards,
and that information concerning drug dosages, schedules, and routes of administration is
accurate at the time of publication and consistent with standards set by the U.S. Food and
Drug Administration and the general medical community. As medical research and prac-
tice continue to advance, however, therapeutic standards may change. Moreover, specific
situations may require a specific therapeutic response not included in this book. For these
reasons and because human and mechanical errors sometimes occur, we recommend that
readers follow the advice of physicians directly involved in their care or the care of a
member of their family.
Books published by American Psychiatric Publishing, Inc., represent the views and opin-
ions of the individual authors and do not necessarily represent the policies and opinions
of APPI or the American Psychiatric Association.
Copyright 2009 American Psychiatric Publishing, Inc.
Manufactured in the United States of America on acid-free paper
13 12 11 10 09 5 4 3 2 1
First Edition
Typeset in Adobes Palatino and Optima.
American Psychiatric Publishing, Inc.
1000 Wilson Boulevard
Arlington, VA 22209-3901
Library of Congress Cataloging-in-Publication Data
Essentials of personality disorders / edited by John M. Oldham, Andrew E. Skodol,
Donna S. Bender ; associate editors, Glen O. Gabbard ... [et al.]. 1st ed.
p. ; cm.
Companion to: The American Psychiatric Publishing textbook of personality disorders /
edited by John M. Oldham, Andrew E. Skodol, Donna S. Bender ; associate editors, Glen
O. Gabbard ... [et al.]. 1st ed. c2005.
Includes bibliographical references and index.
ISBN 978-1-58562-358-7 (alk. paper)
1. Personality disorders. I. Oldham, John M. II. Skodol, Andrew E. III. Bender, Donna S.,
1960 IV. American Psychiatric Publishing textbook of personality disorders.
[DNLM: 1. Personality Disorders. WM 190 E784 2009]
RC554.E883 2009
British Library Cataloguing in Publication Data
A CIP record is available from the British Library.
To our families, who have supported us:

Karen, Madeleine, and Michael Oldham;

Laura, Dan, and Ali Skodol; and
John and Joseph Rosegrant.

To our colleagues, who have helped us.

To our patients, who have taught us.

And to each other, for the friendship that has enriched our work together.
This page intentionally left blank

Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv

Basic Concepts

1 Personality Disorders: Recent History and

the DSM System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
John M. Oldham, M.D., M.S.

2 Theories of Personality and Personality Disorders . . . . . . . . . . 13

Amy Heim, Ph.D., and Drew Westen, Ph.D.

Clinical Evaluation

3 Manifestations, Clinical Diagnosis, and Comorbidity . . . . . . . . 37

Andrew E. Skodol, M.D.

4 Course and Outcome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

Carlos M. Grilo, Ph.D., and Thomas H. McGlashan, M.D.

5 Prevalence, Sociodemographics, and

Functional Impairment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Svenn Torgersen, Ph.D.

6 Neurobiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Emil F. Coccaro, M.D., and Larry J. Siever, M.D.

7 Developmental Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

Patricia Cohen, Ph.D., and Thomas Crawford, Ph.D.

8 Childhood Experiences and

Development of Maladaptive and
Adaptive Personality Traits . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
Jeffrey G. Johnson, Ph.D., Elizabeth Bromley, M.D., and
Pamela G. McGeoch, M.A.


9 Levels of Care in Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . 161

John G. Gunderson, M.D., Kim L. Gratz, Ph.D.,
Edmund C. Neuhaus, Ph.D., and George W. Smith, M.S.W.

10 Psychoanalysis and Psychodynamic Psychotherapy . . . . . . . . 185

Glen O. Gabbard, M.D.

11 Mentalization-Based Treatment of Borderline

Personality Disorder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
Peter Fonagy, Ph.D., F.B.A., and
Anthony W. Bateman, M.A., F.R.C.Psych.

12 Dialectical Behavior Therapy . . . . . . . . . . . . . . . . . . . . . . . . . 235

Barbara Stanley, Ph.D., and Beth S. Brodsky, Ph.D.
13 Group Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
William E. Piper, Ph.D., and John S. Ogrodniczuk, Ph.D.

14 Somatic Treatments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267

Paul H. Soloff, M.D.

15 Therapeutic Alliance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289

Donna S. Bender, Ph.D.

16 Boundary Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309

Thomas G. Gutheil, M.D.

17 Collaborative Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321

Abigail Schlesinger, M.D., and Kenneth R. Silk, M.D.

18 Assessing and Managing Suicide Risk . . . . . . . . . . . . . . . . . . . 343

Paul S. Links, M.D., F.R.C.P.C., and Nathan Kolla, M.D.

19 Substance Abuse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361

Roel Verheul, Ph.D., Louisa M.C. van den Bosch, Ph.D., and
Samuel A. Ball, Ph.D.

New Developments and
Future Directions

20 Future Directions: Toward DSM-V . . . . . . . . . . . . . . . . . . . . . 381

Andrew E. Skodol, M.D., Donna S. Bender, Ph.D., and
John M. Oldham, M.D., M.S.

Appendix: DSM-IV-TR Diagnostic Criteria for

Personality Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403
This page intentionally left blank

Samuel A. Ball, Ph.D. Thomas Crawford, Ph.D.

Professor of Psychiatry, Department of Psy- Assistant Clinical Professor of Medical Psy-
chiatry, Yale University School of Medicine; chology, Columbia University College of
Director of Research, the APT Foundation, Physicians and Surgeons, New York, New
Inc., New Haven, Connecticut York
Anthony W. Bateman, M.A., F.R.C.Psych. Peter Fonagy, Ph.D., F.B.A.
Visiting Professor, Sub-Department of Clini- Freud Memorial Professor of Psychoanalysis
cal Health Psychology, University College and Head of the Research Department of
London; Consultant Psychotherapist, Bar- Clinical, Educational, and Health Psycholo-
net, Enfield, and Haringey Mental Health gy, University College London; Chief Execu-
Trust, London, England tive of the Anna Freud Centre, London,
England; Consultant to the Child and Family
Donna S. Bender, Ph.D.
Program, Menninger Department of Psychi-
Research Associate Professor of Psychiatry,
atry, Baylor College of Medicine, Houston,
University of Arizona College of Medicine;
Chief Executive Officer and Director, Sun-
belt Collaborative, Tucson, Arizona Glen O. Gabbard, M.D.
Brown Foundation Chair of Psychoanalysis
Beth S. Brodsky, Ph.D.
and Professor, Department of Psychiatry,
Assistant Clinical Professor of Medical Psy-
Baylor College of Medicine; Training and Su-
chology, Department of Psychiatry, Colum-
pervising Analyst, Houston-Galveston Psy-
bia University College of Physicians and
choanalytic Institute, Houston, Texas
Surgeons; Research Scientist, Department of
Molecular Imaging and Neuropathology, New Kim L. Gratz, Ph.D.
York State Psychiatric Institute, New York, Assistant Professor and Director, Personality
New York Disorders Research, Department of Psychia-
try and Human Behavior, University of Mis-
Elizabeth Bromley, M.D.
sissippi Medical Center, Jackson, Mississippi
Assistant Professor in Residence, Depart-
ment of Psychiatry and Biobehavioral Scienc- Carlos M. Grilo, Ph.D.
es, Semel Institute Health Services Research Professor of Psychiatry, Department of Psy-
Center, University of California, Los Ange- chiatry, Yale University School of Medicine;
les, California Professor of Psychology, Yale University,
New Haven, Connecticut
Emil F. Coccaro, M.D.
Ellen C. Manning Professor and Chairman, John G. Gunderson, M.D.
Department of Psychiatry, University of Chi- Professor of Psychiatry, Harvard Medical
cago, Chicago, Illinois School, Boston, Massachusetts; Director,
Psychosocial and Personality Research, and
Patricia Cohen, Ph.D.
Director, Borderline Personality Disorder
Professor of Clinical Epidemiology in Psy-
Center, McLean Hospital, Belmont, Massa-
chiatry, Columbia University College of Phy-
sicians and Surgeons, New York, New York

xii E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

Thomas G. Gutheil, M.D. William E. Piper, Ph.D.

Professor of Psychiatry, Department of Psy- Professor and Head, Division of Behavioural
chiatry, Beth Israel Deaconess Medical Cen- Science; Director, Psychotherapy Program,
ter, Harvard Medical School, and Co- Department of Psychiatry, University of
Founder, Program in Psychiatry and the British Columbia, Vancouver, British Co-
Law, Massachusetts Mental Health Center, lumbia, Canada
Boston, Massachusetts
Abigail Schlesinger, M.D.
Amy Heim, Ph.D. Assistant Professor, Western Psychiatric In-
Private practice, Lexington, Massachusetts stitute and Clinic, University of Pittsburgh
School of Medicine, Pittsburgh, Pennsylva-
Jeffrey G. Johnson, Ph.D.
Associate Professor of Clinical Psychology,
Department of Psychiatry, College of Physi- Larry J. Siever, M.D.
cians and Surgeons, Columbia University; Executive Director, Mental Illness Research,
Research Scientist IV, Epidemiology of Men- Education and Clinical Center, Bronx Veter-
tal Disorders Department, New York State ans Administration Medical Center, Bronx,
Psychiatric Institute, New York, New York New York; Professor of Psychiatry, Depart-
ment of Psychiatry, The Mount Sinai School
Nathan Kolla, M.D.
of Medicine, New York, New York
Psychiatry Resident, Department of Psychia-
try, University of Toronto, Toronto, Ontario, Kenneth R. Silk, M.D.
Canada Professor and Director, Personality Disor-
ders Program, Department of Psychiatry,
Paul S. Links, M.D., F.R.C.P.C.
University of Michigan Health System, Ann
Arthur Sommer Rotenberg Chair in Suicide
Arbor, Michigan
Studies, Professor of Psychiatry, Department
of Psychiatry, St. Michaels Hospital, Uni- Andrew E. Skodol, M.D.
versity of Toronto, Toronto, Ontario, Canada Research Professor of Psychiatry, University
of Arizona College of Medicine; President,
Pamela G. McGeoch, M.A.
Sunbelt Collaborative, Tucson, Arizona
Graduate Faculty, Department of Psycholo-
gy, The New School University, New York, George W. Smith, M.S.W.
New York Director, Outpatient Personality Disorder
Services, McLean Hospital, Boston, Massa-
Thomas H. McGlashan, M.D.
Professor of Psychiatry, Department of Psy-
chiatry, Yale University School of Medicine, Paul H. Soloff, M.D.
New Haven, Connecticut Professor of Psychiatry, Western Psychiatric
Institute and Clinic, University of Pitts-
Edmund C. Neuhaus, Ph.D.
burgh, Pittsburgh, Pennsylvania
Director, Behavioral Health Partial Hospital
Program, Co-Director of Psychology Train- Barbara Stanley, Ph.D.
ing, McLean Hospital, Belmont, Massachu- Professor of Clinical Psychiatry, Department
setts; Assistant Clinical Professor, Harvard of Psychiatry, Columbia University College
Medical School, Boston, Massachusetts of Physicians and Surgeons; Research Scien-
tist, Department of Molecular Imaging and
John S. Ogrodniczuk, Ph.D.
Neuropathology, New York State Psychiat-
Associate Professor, Department of Psychia-
ric Institute, New York, New York
try, University of British Columbia, Vancou-
ver, British Columbia, Canada Svenn Torgersen, Ph.D.
Professor, Department of Psychology, Uni-
John M. Oldham, M.D., M.S.
versity of Oslo, Oslo, Norway
Senior Vice President and Chief of Staff, The
Menninger Clinic; Professor and Executive Louisa M.C. van den Bosch, Ph.D.
Vice Chairman, Menninger Department of Clinical Psychologist/Psychotherapist, CSP
Psychiatry and Behavioral Sciences, Baylor Leiden; Administrative Executive, Dialexis,
College of Medicine, Houston, Texas Noordwijkerhout, The Netherlands

Roel Verheul, Ph.D. The following contributors have no competing in-

Professor of Personality Disorders, Vier- terests to report:
sprong Institute for Studies on Personality
Disorders (VISPD); University of Amster- Samuel A. Ball, Ph.D.
dam, Department of Clinical Psychology; Anthony W. Bateman, M.A., F.R.C.Psych.
Amsterdam, Chief Executive Officer, Center Donna S. Bender, Ph.D.
for Psychotherapy De Viersprong, Halster- Beth S. Brodsky, Ph.D.
en, The Netherlands Elizabeth Bromley, M.D.
Patricia Cohen, Ph.D.
Drew Westen, Ph.D. Thomas Crawford, Ph.D.
Professor, Department of Psychiatry and Be- Glen O. Gabbard, M.D.
havioral Sciences and Department of Psy- Carlos M. Grilo, Ph.D.
chology, Emory University, Atlanta, Georgia John G. Gunderson, M.D.
Thomas G. Gutheil, M.D.
Amy Heim, Ph.D.
Jeffrey G. Johnson, Ph.D.
Pamela G. McGeoch, M.A.
Thomas H. McGlashan, M.D.
The following contributor to this book has indi-
Edmund C. Neuhaus, Ph.D.
cated a financial interest in or other affiliation
with a commercial supporter, a manufacturer of a John S. Ogrodniczuk, Ph.D.
commercial product, a provider of a commercial William E. Piper, Ph.D.
service, a nongovernmental organization, and/or Larry J. Siever, M.D.
a government agency, as listed below. Kenneth R. Silk, M.D.
George W. Smith, M.S.W.
Paul S. Links, M.D., F.R.C.P.C.Educational Paul H. Soloff, M.D.
grant: Eli Lilly Barbara Stanley, Ph.D.
Svenn Torgersen, Ph.D.
Roel Verheul, Ph.D.
John M. Oldham, M.D., M.S.
Andrew E. Skodol, M.D.
This page intentionally left blank

From as early as the fifth century B.C., it has true in some cases involving significant ge-
been recognized that every human being de- netic loading or risk. In the twentieth cen-
velops an individualized signature pattern tury, however, we became more interested in
of behavior that is reasonably persistent and the role of the environment during early de-
predictable throughout life. Hippocrates velopment in determining the shape of last-
proposed that the varieties of human behav-
ing adult behaviora view that for a while
ior could be organized into what we might
extended well beyond the realm of the per-
now call prototypesbroad descriptive pat-
terns of behavior characterized by typical, sonality disorders to include most major
predominant, easily recognizable features mental disorders. We know, of course, that
and that most individuals could be sorted the early life environment is indeed critically
into these broad categories. Sanguine, mel- importantfrom health-promoting, highly
ancholic, choleric, and phlegmatic types of nurturing environments to stressful and ne-
behavior were, in turn, thought to derive glectful environments from which only the
from body humors, such as blood, black most resilient emerge unscathed. But we also
bile, yellow bile, and phlegm, and the pre- know that variable degrees of genetic risk
dominance of a given body humor in an in- predispose many of us to become ill in very
dividual was thought to correlate with a par- specific ways, should we unluckily encoun-
ticular behavior pattern. Although we now
ter more stress than we can tolerate.
call body humors by different names (neu-
In recent years, we have begun to see an
rotransmitters, transcription factors, second
messengers), the ancient principle that fun- upsurge of empirical and clinical interest in
damental differences in biology correlate personality disorders. Improved standard-
with relatively predictable patterns of be- ized diagnostic systems have led to semi-
havior is strikingly familiar. structured research interviews that are being
In spite of long-standing worldwide in- used not only in studies of clinical popula-
terest in personality types, however, remark- tions but also in community-based studies,
ably little progress has been made, until re- to give us, for the first time, good data about
cently, in our understanding of those severe the epidemiology of these disorders. Person-
and persistent patterns of inner experience ality disorders occur in about 10% of the gen-
and behavior that result in enduring emo- eral population, and their public health sig-
tional distress and impairment in occupa- nificance has been documented by studies
tional functioning and interpersonal rela- showing their extreme social dysfunction
tionshipsthe conditions we now refer to as and high health care utilization. As clinical
personality disorders. For decades, it was populations are becoming better defined,
widely recognized that some severely dis- new and more rigorous treatment studies are
turbed individuals just seemed to have been being carried out, with increasingly promis-
born that way, a view we now know to be ing results. No longer are personality disor-

xvi E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

ders swept into the hopeless cases bin. An ment, and 5) New Developments and Future
explosion of knowledge and technology in Directions.
the neurosciences has made the formerly
black box, the brain, more and more trans-
parent. Mapping the human genome paved PART I: BASIC CONCEPTS
the way for new gene-finding technologies
that are being put to work to tackle complex Basic Concepts, the first part of Essentials of
psychiatric disorders, including the person- Personality Disorders, might be thought of as
ality disorders. New transgenic animal setting the stage for the parts that follow. In
models are providing important hints about Chapter 1, Oldham presents a brief overview
the genetic loci driving certain behavior of the recent history of the personality disor-
types, such as attachment and bonding be- ders, along with a summary look at the evo-
havior. Brain imaging studies are allowing lution of the personality disorders compo-
researchers to zero in on malfunctioning ar- nent in successive editions of the American
eas of the brain in specific personality disor- Psychiatric Association Diagnostic and Statisti-
ders. cal Manual of Mental Disorders. Heim and
A great deal of work must still be done. Westen, in the next chapter, review the major
Fundamental questions remain, such as theories that have influenced our thinking
what is the relationship between traits of about the nature of personality and person-
general, or normal, personality functioning ality disorders.
and personality psychopathology. Directly
related to this issue is the ongoing debate
about whether dimensional or categorical
systems best capture the full scope of person-
ality differences and personality pathology.
In the section on clinical evaluation begin-
Extensive impairment can be associated with
ning with Chapter 3, Skodol reviews the de-
personality disorders, but we are gaining
fining features of DSM-IV-TR personality
new knowledge regarding their longer-term
disorders, discusses complementary ap-
course and increased understanding of fac-
proaches to the clinical assessment of a pa-
tors contributing to variations in course.
tient with a possible personality disorder,
There is a strong momentum of interest inter-
provides guidance on general problems en-
nationally in these issues, as new research
countered in the routine clinical evaluation,
findings emerge daily to inform the process.
and describes patterns of Axis I and Axis II
In light of the continuing and increased
disorder comorbidity. In Chapter 4, Grilo
activity and progress in the field of personal-
and McGlashan provide an overview of the
ity studies and personality disorders, we
clinical course and outcome of personality
judged the time to be right to develop a com-
disorders, synthesizing the empirical litera-
panion volume to The American Psychiatric
ture on the stability of personality disorder
Publishing Textbook of Personality Disorders,
published in 2005 to cover and update the es-
sentials of this topic. Because of space con-
straints, it was necessary to be selective re-
garding what material to include in the PART III: ETIOLOGY
Essentials volume, and we have tried to focus
on material that is up-to-date and useful to The section on etiology of the personality
practicing clinicians. This new volume is or- disorders begins with Chapter 5. Torgersen
ganized into several parts: 1) Basic Concepts, presents the best data we have to date on the
2) Clinical Evaluation, 3) Etiology, 4) Treat- population-based epidemiology of the per-
Introduction xvii

sonality disorders. Although there are rela- and maladaptive personality traits. They em-
tively few well-designed population-based phasize the importance of stress, but also of
studies, Torgersen reviews a number of stud- protective factors that can offset and even
ies, including his own Norwegian study, and prevent the development of maladaptive
tabulates prevalence ranges and averages for traits in vulnerable individuals.
individual DSM-defined personality disor-
ders as well as for all personality disorders
taken together (showing an overall average PART IV: TREATMENT
prevalence rate for the personality disorders
of over 12%). Of particular interest in these The treatment section begins with Chapter 9,
data are cross-cultural comparisons, sug- a discussion of the levels of care available for
gesting significant cultural differences in the patients with personality disorders. Gunder-
prevalence of selected personality disorders. son, Gratz, Neuhaus, and Smith offer guide-
Substantial progress has been made in our lines for determining the appropriate inten-
understanding of the neurobiology of the sity of treatment services for individual
personality disorders, as reviewed in Chap- patients. Four levels of care are addressed:
ter 6 by Coccaro and Siever. Although a great hospitalization, partial hospitalization/day
deal more is known about the neurobiology treatment, intensive outpatient, and outpa-
of some personality disorders (e.g., schizo- tient.
typal personality disorder and borderline Chapters 10 through 14 offer a range of
personality disorder) than others (e.g., treatment options. Gabbard (Chapter 10)
Cluster C personality disorders), the under- summarizes the salient features of psycho-
lying neurobiological dysfunction involved analysis and psychodynamic psychotherapy
in personality disorders characterized by as applied to patients with character pathol-
cognitive symptomatology, impulsivity, and ogy. One psychodynamic model, mentaliza-
mood dysregulation is becoming increas- tion-based therapy, is of special interest in
ingly clear. the treatment of patients with borderline per-
Understanding the etiology of the per- sonality disorder, as described in Chapter 11
sonality disorders involves not just cross-sec- by Fonagy and Bateman. In this model, bor-
tional genetic and neurobiological analysis; derline personality disorder is seen as dys-
environmental influences shaping personal- function in self-regulation, critically related
ity must be understood as well. In Chapter 7, to interpersonal dynamics. In Chapter 12,
Cohen and Crawford provide a developmen- Stanley and Brodsky outline the core ele-
tal perspective. Although, by convention, ments of dialectical behavior therapy, which
DSM-IV-TR personality disorders are gener- includes individual and group interventions,
ally not diagnosed until late adolescence, and is chiefly used to treat parasuicidal be-
there is increasing recognition of early pat- haviors in patients with borderline personal-
terns of behavior that are thought to be pre- ity disorder.
cursors to certain personality disorders. The Apart from the realm of individual treat-
challenge to identify true early precursors of ments, there are other venues for therapeutic
personality disorders, versus the risk of inac- interventions. In Chapter 13, Piper and Og-
curate labeling of transient symptoms, is cen- rodniczuk demonstrate the application of
tral to the work ahead of us as we focus more group therapy to personality disorders. So-
and more on prevention strategies. Comple- loff (Chapter 14) takes up the issue of phar-
menting this developmental approach, the macotherapy and other somatic treatments,
authors of Chapter 8, Johnson, Bromley, and because many patients with personality dis-
McGeoch, review the relevance of childhood orders may benefit by complementing their
experiences to the development of adaptive psychosocial treatments with medication.
xviii E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

The final five chapters of this section ad-

dress issues of great importance pertaining to PART V: NEW DEVELOPMENTS AND
most, if not all, treatments. Bender (Chapter FUTURE DIRECTIONS
15) underscores the necessity of explicitly
considering alliance building across all treat- In the final section of Essentials of Personality
ment modalities, while Gutheil (Chapter 16) Disorders, Skodol, Bender, and Oldham sum-
cautions practitioners about dynamics that marize current controversies and speculate
can lead treaters to boundary violations about future directions, as the work gets un-
when working with certain patients with per- der way to consider the best strategies to clas-
sonality disorders. Schlesinger and Silk, in sify and understand the personality disor-
Chapter 17, provide recommendations about ders in DSM-V.
the best way of negotiating collaborative We are grateful to all of the authors of
treatments, since many patients with person- each chapter for their careful and thoughtful
ality disorders are engaged in several modal- contributions, and we hope that we have suc-
ities with several clinicians at the same time. ceeded in providing a current, definitive re-
In recognition of the fact that patients view of the field. We would particularly like
with personality disorders can be particu- to thank Liz Bednarowicz for her organized
larly challenging, we have included two and steadfast administrative support, with-
chapters devoted to the issues often faced out which this volume would not have been
when treating these individuals. Of prime possible.
importance is the risk for suicide. In Chapter
18, Links and Kolla provide evidence on the John M. Oldham, M.D., M.S.
association of suicidal behavior and person- Houston, Texas
ality disorders, examine modifiable risk fac-
tors, and discuss clinical approaches to the as- Andrew E. Skodol, M.D.
sessment and management of suicide risk. In Phoenix, Arizona
Chapter 19, Verheul, van den Bosch, and Ball Donna S. Bender, Ph.D.
focus on pathways to substance abuse in pa- Tucson, Arizona
tients with personality disorders, and discuss
issues of differential diagnosis and treatment.
Part I
Basic Concepts
This page intentionally left blank
Personality Disorders
Recent History and
the DSM System
John M. Oldham, M.D., M.S.

As to why we behave the way we do, we

PERSONALITY TYPES know now that a fair amount of the reason re-
AND PERSONALITY DISORDERS lates to our hardwiring. To varying de-
grees, heritable temperaments that vary
Charting a historical review of efforts to un- widely from one individual to another deter-
derstand personality types and the differ- mine the amazing range of behavior in the
ences among them would involve exploring newborn nursery, from cranky to placid.
centuries of scholarly archives, worldwide, Each individuals temperament remains a
on the varieties of human behavior. For it is key component of that persons developing
human behavior, in the end, that serves as the personality, to which is added the shaping
most v alid measurable and observable and molding influences of family, caretakers,
benchmark of personality. In many impor- and environmental experiences. This process
tant ways, we are what we do. The what of is, we now know, bidirectional, so that the
personality is easier to come by than the inborn behavior of the infant can elicit be-
why, and each of us has a personality style havior in parents or caretakers that can, in
that is unique, almost like a fingerprint. At a turn, reinforce infant behavior: placid, happy
school reunion, recognition of classmates not babies may elicit warm and nurturing behav-
seen for decades derives as much from famil- iors; irritable babies may elicit impatient and
iar behavior as from physical appearance. neglectful behaviors.

Sections of this chapter have been modified from Oldham JM, Skodol AE: Charting the Future of Axis II.
Journal of Personality Disorders 14:1729, 2000. Reprinted with permission of Guilford Press.

4 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

However, even-tempered, easy-to-care- cursors or less extreme forms of psychotic

for babies can have bad luck and land in a conditions, such as schizophrenia or manic-
nonsupportive or even abusive environment depressive illnesssystems that can clearly
that may set the stage for a personality disor- be seen as forerunners of current Axis I/Axis
der, and difficult-to-care-for babies can have II spectrum models. Schneider, on the other
good luck and be protected from future per- hand, described a set of psychopathic per-
sonality pathology by specially talented and sonalities that he viewed as separate disor-
attentive caretakers. Once these highly indi- ders co-occurring with other psychiatric dis-
vidualized dynamics have had their main ef- orders. Although these classical systems of
fects and an individual has reached late ado- descriptiv e psychopathology resonate
lescence or young adulthood, his or her strongly with the framework eventually
personality will usually have been pretty adopted by the American Psychiatric Associ-
well established. We know that this is not an ation (APA) and published in its Diagnostic
ironclad rule; there are late bloomers, and and Statistical Manual of Mental Disorders
high-impact life events can derail or reroute (DSM), they were widely overshadowed in
any of us. How much we can change if we American psychiatry during the mid-twenti-
need and want to is variable, but change is eth century by theory-based psychoanalytic
possible. How we define the differences be- concepts stimulated by the work of Sigmund
tween personality styles and personality dis- Freud and his followers.
orders, how the two relate to each other, what Freud emphasized the presence of a dy-
systems best capture the magnificent variety namic unconscious, a realm that, by defini-
of nonpathological human behavior, and tion, is mostly unavailable to conscious
how we think about and deal with extremes thought but is a powerful motivator of hu-
of behavior that we call personality disorders man behavior (key ingredients of his topo-
are all spelled out in great detail in the chap- graphical model). His emphasis on a dy-
ters of this textbook. In this first chapter, I namic unconscious was augmented by his
briefly describe how psychiatrists in the well-known tripartite structural theory, a
United States have approached the definition conflict model serving as the bedrock of his
and classification of the personality disor- psychosexual theory of pathology (Freud
ders, building on broader international con- 1926). Freud theorized that certain uncon-
cepts and theories of psychopathology. scious sexual wishes or impulses (id) could
threaten to emerge into consciousness (ego),
thus colliding with conscience-driven prohi-
bitions (superego) and producing signal
TWENTIETH-CENTURY CONCEPTS OF anxiety, precipitating unconscious defense
PERSONALITY PSYCHOPATHOLOGY mechanisms and, when these coping strate-
gies prove insufficient, leading to frank
Personality pathology has been recognized in symptom formation. For the most part, this
most influential systems of classifying psy- system was proposed as an explanation for
chopathology. The well-known contributions what were called at the time the symptom neu-
by European pioneers of descriptive psychia- roses, such as hysterical neurosis or obsessive-
try, such as Kraepelin (1904), Bleuler (1924), compulsive neurosis. During the 1940s,
Kretschmer (1926), and Schneider (1923/ 1950s, and 1960s, these ideas became domi-
1950) had an important impact on early twen- nant in American psychiatry, followed later
tieth-century American psychiatry. For the by interest in other psychoanalytic principles,
most part, Kraepelin, Bleuler, and Kretschmer such as object relations theory.
described personality types or temperaments, Freuds concentration on the symptom
such as aesthenic, autistic, schizoid, cyclothy- neuroses involved the central notion of anxi-
mic, or cycloid, that were thought to be pre- ety as the engine that led to defense mecha-
Personality Disorders: Recent History and the DSM System 5

nisms and to symptom formation, and as a other syndromes characterized by discrete,

critical factor in motivating patients to work persistent symptom patterns such as major
hard in psychoanalysis to face painful real- depressive episodes, persistent anxiety, or
izations and to tolerate stress within the dementia. General clinical experience and
treatment itself (such as that involved in the wisdom guided treatment recommendations
transference neurosis). Less prominently for these patients, at least for those who
articulated were Freuds notions of character sought treatment. Patients with paranoid,
pathology, but generally character disorders schizoid, or antisocial patterns of thinking
were seen to represent pre-oedipal pathol- and behaving often did not seek treatment.
ogy. As such, patients with these conditions Others, however, often resembled patients
were judged less likely to be motivated to with symptom neuroses and did seek help
change. Instead of experiencing anxiety re- for problems ranging from self-destructive
lated to the potential gratification of an un- behavior to chronic misery. The most se-
acceptable sexual impulse, patients with verely and persistently disabled of these pa-
fixations at the oral-dependent stage, for tients were often referred for intensive, psy-
example, experienced anxiety when not grat- choanalytically oriented long-term inpatient
ifying the impulsein this case, the need to treatment at treatment centers such as Aus-
be fed. Relief of anxiety thus could be accom- ten Riggs, Chestnut Lodge, Menninger
plished by some combination of real and Clinic, McLean Hospital, New York Hospital
symbolic feedingattention from a parent or Westchester Division, New York State Psy-
parent figure or consumption of alcohol or chiatric Institute, Sheppard Pratt, and other
drugs. Deprivations within the psychoana- long-term inpatient facilities available at the
lytic situation, theninevitable by its very time. Other patients, able to function outside
naturecould lead to patient flight and in- of a hospital setting and often hard to distin-
terrupted treatment. guish from patients with neuroses, were
In a way, social attitudes mirrored and referred for outpatient psychoanalysis or
extended these beliefs such that although intensive psychoanalytically oriented psy-
personality pathology was well known, it chotherapy. As Gunderson (2001) described,
was often thought to reflect weakness of the fact that many such patients in psycho-
character or willfully offensive or socially de- analysis regressed and seemed to get worse,
viant behavior produced by faulty upbring- rather than showing improvement in treat-
ing, rather than understood as legitimate ment, was one factor that contributed to the
psychopathology. A good example of this emerging concept of borderline personality
view could be seen in military psychiatry in disorder (BPD), thought initially to be in the
the mid-1900s, where those discharged from border zone between the psychoses and the
active duty for mental illness, with eligibility neuroses. Patients in this general category in-
for disability and medical benefits, did not cluded some who had previously been la-
include individuals with character disor- beled as having latent schizophrenia (Bleuler
ders (or alcoholism and substance abuse) 1924), ambulatory schizophrenia (Zillborg
because these conditions were seen as bad 1941), pseudoneurotic schizophrenia (Hoch
behavior and led to administrative, non- and Polatin 1949), psychotic character
medical separation from the military. (Frosch 1964), or as-if personality (Deutsch
In spite of these common attitudes, clini- 1942).
cians recognized that many patients with sig- These developments coincided with new
nificant impairment in social or occupational approaches based on alternative theoretical
functioning, or with significant emotional models that were emerging within the psy-
distress, needed treatment for psychopathol- choanalytic framework, such as the British
ogy that did not involve frank psychosis or object relations school. New conceptual
6 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

frameworks, such as Kernbergs (1975) which persist to the present. Generally, per-
model of borderline personality organization sonality disorders were viewed as more or
or Kohuts (1971) concept of the central im- less permanent patterns of behavior and hu-
portance of empathic failure in the histories man interaction that were established by
of narcissistic patients, served as the basis for early adulthood and were unlikely to change
an intensive psychodynamic treatment ap- throughout the life cycle. Thorny issues such
proach for selected patients with personality as how to differentiate personality disorders
disorders. These strategies and others are re- from personality styles or traits, which re-
viewed in detail in Chapter 10, Psychoanal- main actively debated today, were clearly
ysis and Psychodynamic Psychotherapy. identified at the time. Personality disorders
were contrasted with the symptom neuroses
in a number of ways, particularly that the
THE DSM SYSTEM neuroses were characterized by anxiety and
distress, whereas the personality disorders
Contrary to assumptions commonly encoun- were often ego-syntonic and thus not recog-
tered, personality disorders have been in- nized by those who had them. Some person-
cluded in every edition of DSM. Largely ality disorders are currently referred to as
driven by the need for standardized psychi- externalizing disordersthat is, disorders
atric diagnosis in the context of World War in which the patient disavows any problem
II, the U. S. War Department in 1943 devel- but blames all discomfort on the real or per-
oped a document labeled Technical Bulletin ceived unreasonableness of others.
203, representing a psychoanalytically ori- In DSM-I, personality disorders were
ented system of terminology for classifying generally viewed as deficit conditions reflect-
mental illness precipitated by stress (Barton ing partial developmental arrests or distor-
1987). The APA charged its Committee on tions in development secondary to inade-
Nomenclature and Statistics to solicit expert quate or pathological early caretaking. The
opinion and to develop a diagnostic manual personality disorders were grouped prima-
that would codify and standardize psychiat- rily into personality pattern disturbances,
ric diagnoses. This diagnostic system became personality trait disturbances, and socio-
the framework for the first edition of DSM pathic personality disturbances. Personality
(DSM-I; American Psychiatric Association pattern disturbances were viewed as the most
1952). This manual was widely utilized, and entrenched conditions and likely to be recal-
it was subsequently revised on several occa- citrant to change, even with treatment; these
sions, leading to DSM-II (American Psychiat- included inadequate personality, schizoid
ric Association 1968), DSM-III (American personality, cyclothymic personality, and
Psychiatric Association 1980), DSM-III-R paranoid personality. Personality trait distur-
(American Psychiatric Association 1987), bances were thought to be less pervasive and
DSM-IV (American Psychiatric Association disabling, so that in the absence of stress these
1994), and DSM-IV-TR (American Psychiat- patients could function relatively well. If un-
ric Association 2000). Figure 11 (Skodol der significant stress, however, patients with
1997) portrays the ontogeny of diagnostic emotionally unstable, passive-aggressive, or
terms relevant to the personality disorders compulsive personalities were thought to
from DSM-I through DSM-IV (DSM-IV-TR show emotional distress and deterioration in
involved only text revisions; it used the same functioning, and they were variably moti-
diagnostic terms as DSM-IV). vated for and amenable to treatment. The cat-
Although not explicit in the narrative egory of sociopathic personality disturbances re-
text, DSM-I reflected the general view of per- flected what were generally seen as types of
sonality disorders at the time, elements of social deviance at the time, including antiso-
Personality Disorders: Recent History and the DSM System 7

DSM -I (1952) DSM -II (1968) DSM -III (1980) DSM-IV (1994)/
DSM-IV-TR (2000)
Personality pattern Axis I cyclothymic Axis I cyclothymic
disturbance disorder disorder
Inadequate Inadequate Cluster A Cluster A
Paranoid Paranoid Paranoid Paranoid
Cyclothymic Cyclothymic
Schizoid Schizoid Schizoid Schizoid
Schizotypal Schizotypal

Personality trait
disturbance Cluster B Cluster B
Emotionally unstable Hysterical Histrionic Histrionic
Passive- aggressive Antisocial Antisocial
dependent type Borderline Borderline
aggressive type Passive - aggressive Narcissistic Narcissistic

Cluster C Cluster C
Compulsive Obsessive - compulsive Compulsive Obsessive- compulsive
Avoidant Avoidant
Dependent Dependent
Passive - aggressive
Sociopathic personality
disturbance Asthenic
Antisocial Antisocial
Dyssocial Explosive Axis I intermittent Axis I intermittent
explosive disorder explosive disorder

DSM-IV Appendix
Passive -aggressive
DSM - III - R Appendix*
Self- defeating
Indicates that category was discontinued.

Figure 11. Ontogeny of personality disorder classification.

*No changes were made to the personality disorder classification in DSM-III-R except for the inclusion of
self-defeating and sadistic personality disorders in Appendix A: Proposed Diagnostic Categories Need-
ing Further Study. These two categories were not included in DSM-IV or in DSM-IV-TR.
Source. Reprinted with permission from Skodol AE: Classification, Assessment, and Differential Diag-
nosis of Personality Disorders. Journal of Practical Psychiatry and Behavioral Health 3:261274, 1997.

cial reaction, dyssocial reaction, sexual devia- were observable, measurable, enduring, and
tion, and addiction (subcategorized into alco- consistent over time. The earlier view that
holism and drug addiction). patients with personality disorders did not
The primary stimulus leading to the de- experience emotional distress was dis-
velopment of a new, second edition of DSM carded, as were the DSM-I subcategories of
was the publication of the eighth edition of personality pattern, personality trait, and so-
the International Classification of Diseases ciopathic personality disturbances. One new
(World Health Organization 1967) and the personality disorder was added, called as-
wish of the APA to reconcile its diagnostic thenic personality disorder, only to be de-
terminology with this international system. leted in the next edition of the DSM.
In the DSM revision process, an effort was By the mid-1970s, greater emphasis was
made to move away from theory-derived di- placed on increasing the reliability of all di-
agnoses and to attempt to reach consensus agnoses; whenever possible, diagnostic crite-
on the main constellations of personality that ria that were observable and measurable
8 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

were developed to define each diagnosis. diagnoses were added in DSM-III: BPD and
DSM-III, the third edition of the diagnostic narcissistic personality disorder. In contrast
manual, introduced a multiaxial system. Dis- to initial notions that patients called border-
orders classified on Axis I included those line were on the border between the psy-
generally seen as episodic, characterized by choses and the neuroses, the criteria defining
exacerbations and remissions, such as psy- BPD in DSM-III emphasized emotional dys-
choses, mood disorders, and anxiety disor- regulation, unstable interpersonal relation-
ders. Axis II was established to include the ships, and loss of impulse control more than
personality disorders as well as mental retar- cognitive distortions and marginal reality
dation; both groups were seen as composed testing, which were more characteristic of
of early-onset, persistent conditions, but schizotypal personality disorder. Among
mental retardation was understood to be bi- many scholars whose work greatly influ-
ological in origin, in contrast to the person- enced and shaped our understanding of bor-
ality disorders, which were generally re- derline pathology were Kernberg (1975) and
garded as psychological in origin. The Gunderson (1984, 2001, 2008). Although con-
stated reason for placing the personality dis- cepts of narcissism had been described by
orders on Axis II was to ensure that consid- Freud, Reich, and others, the essence of the
eration is given to the possible presence of current views of narcissistic personality dis-
disorders that are frequently overlooked order emerged from the work of Millon
when attention is directed to the usually (1969), Kohut (1971), and Kernberg (1975).
more florid Axis I disorders (American Psy- DSM-III-R was published in 1987 after an
chiatric Association 1980, p. 23). It is gener- intensive process to revise DSM-III involving
ally agreed that the decision to place the per- widely solicited input from researchers and
sonality disorders on Axis II led to greater clinicians and following similar principles to
recognition of the personality disorders and those articulated in DSM-III, such as assuring
stimulated extensive research and progress reliable diagnostic categories that were clini-
in our understanding of these conditions. cally useful and consistent with research
As shown in Figure 11, the DSM-II diag- findings, thus minimizing reliance on theory.
noses of inadequate personality disorder and Efforts were made for diagnoses to be descrip-
asthenic personality disorder were discon- tive and to require a minimum of inference,
tinued in DSM-III. The diagnosis of explo- although the introductory text of DSM-III-R
sive personality disorder was changed to in- acknowledged that for some disorders, par-
termittent explosive disorder, cyclothymic ticularly the Personality Disorders, the cri-
personality disorder was renamed cyclothy- teria require much more inference on the part
mic disorder, and both of these diagnoses of the observer (American Psychiatric Asso-
were moved to Axis I. Schizoid personality ciation 1987, p. xxiii). No changes were made
disorder was felt to be too broad a category in in DSM-III-R diagnostic categories of person-
DSM-II, and it was recrafted into three per- ality disorders, although some adjustments
sonality disorders: schizoid personality disor- were made in certain criteria sets, for exam-
der, reflecting loners who are uninterested ple, making them uniformly polythetic in-
in close personal relationships; schizotypal stead of defining some personality disorders
personality disorder, understood to be on the with monothetic criteria sets (e.g., dependent
schizophrenia spectrum of disorders and personality disorder) and others with poly-
characterized by eccentric beliefs and non- thetic criteria sets (e.g., BPD). In addition, two
traditional behavior; and avoidant personality personality disorders were included in DSM-
disorder, typified by self-imposed interper- III-R in Appendix A (Proposed Diagnostic
sonal isolation driven by self-consciousness Categories Needing Further Study)self-
and anxiety. Two new personality disorder defeating personality disorder and sadistic
Personality Disorders: Recent History and the DSM System 9

personality disorderbased on prior clinical DSM-IV. The intent of DSM-IV-TR was to re-
recommendations to the DSM-III-R personal- vise the descriptive, narrative text accom-
ity disorder subcommittee. These diagnoses panying each diagnosis where it seemed in-
were considered provisional, pending further dicated and to update the information
review and research. provided. Only minimal revisions were
DSM-IV was derived after an extensive made in the text material accompanying the
process of literature review, data analysis, personality disorders.
field trials, and feedback from the profession.
Because of the increase in research stimulated
by the criteria-based multiaxial system of CURRENT CONTROVERSIES AND
DSM-III, a substantial body of evidence ex-
isted to guide the DSM-IV process. As a re-
sult, the threshold for approval of revisions There is a general consensus, at least in the
for DSM-IV was higher than that used in United States, that the placement of the per-
DSM-III or DSM-III-R. DSM-IV introduced, sonality disorders on Axis II has stimulated
for the first time, a set of general diagnostic research and focused clinical and educa-
criteria for any personality disorder (Table 1 tional attention on these disabling condi-
1), underscoring qualities such as early onset, tions. However, there is growing debate
long duration, inflexibility, and pervasive- about the continued appropriateness of
ness. Diagnostic categories and dimensional maintaining the personality disorders on a
organization of the personality disorders into separate axis in future editions of the diag-
clusters remained the same in DSM-IV as in nostic manual and about whether a dimen-
DSM-III-R, with the exception of the reloca- sional or a categorical system of classification
tion of passive-aggressive personality disor- is preferable. As new knowledge has rapidly
der from the official diagnostic list to Ap- accumulated about the personality disor-
pendix B (Criteria Sets and Axes Provided ders, these controversies take their places
for Further Study). Passive-aggressive per- among many ongoing constructive dia-
sonality disorder, as defined by DSM-III and logues, such as the relationship of normal
DSM-III-R, was thought to be too unidimen- personality to personality disorder, the pros
sional and generic; it was tentatively retitled and cons of polythetic criteria sets, how to
negativistic personality disorder, and the determine the appropriate number of criteria
criteria were revised. In addition, the two (i.e., threshold) required for each diagnosis,
provisional Axis II diagnoses in DSM-III-R, which personality disorder categories have
self-defeating personality disorder and sadis- construct validity, which dimensions best
tic personality disorder, were dropped be- cover the scope of normal and abnormal per-
cause of insufficient research data and clini- sonality, and others. Many of these discus-
cal consensus to support their retention. One sions overlap with and inform each other,
other personality disorder was proposed and and these considerations for the future are
added to Appendix B: depressive personality discussed in detail in Chapter 20 of this vol-
disorder. Although substantially controver- ume, Future Directions: Toward DSM-V.
sial, this provisional diagnosis was proposed
as a pessimistic cognitive style; its validity
and its distinction from passive-aggressive
personality disorder on Axis II or dysthymic
disorder on Axis I, however, remain to be es- This brief review of recent notions of person-
tablished. ality pathology serves as a window on the
DSM-IV-TR, published in 2000, did not rapid progress in our field and in our under-
change the diagnostic terms or criteria of standing of psychiatric disorders. Increas-
10 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

Table 11. General diagnostic criteria for a personality disorder

A. An enduring pattern of inner experience and behavior that deviates markedly from the
expectations of the individual's culture. This pattern is manifested in two (or more) of the
following areas:
(1) cognition (i.e., ways of perceiving and interpreting self, other people, and events)
(2) affectivity (i.e., the range, intensity, lability, and appropriateness of emotional
(3) interpersonal functioning
(4) impulse control
B. The enduring pattern is inflexible and pervasive across a broad range of personal and
social situations.
C. The enduring pattern leads to clinically significant distress or impairment in social,
occupational, or other important areas of functioning.
D. The pattern is stable and of long duration, and its onset can be traced back at least to
adolescence or early adulthood.
E. The enduring pattern is not better accounted for as a manifestation or consequence of
another mental disorder.
F. The enduring pattern is not due to the direct physiological effects of a substance (e.g., a
drug of abuse, a medication) or a general medical condition (e.g., head trauma).

Source. Reprinted with permission from American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders, 4th Edition, Text Revision. Washington, DC, American Psychiatric Associa-
tion, 2000.

ingly, a stress/diathesis framework seems somewhat unique to the personality disor-

applicable in medicine in general as a unify- ders is their correlation and continuity with
ing model of illnessa model that can easily normal functioning, which could be an im-
encompass the personality disorders (Paris portant consideration in future revisions of
1999). Variable genetic vulnerabilities pre- our diagnostic system. As we learn more
dispose us all to potential future illness that about the etiologies and pathology of the
may or may not develop depending on the personality disorders, it will no longer be
balance of specific stressors and protective necessary, or even desirable, to limit our di-
factors. agnostic schemes to atheoretical, descriptive
The personality disorders represent mal- phenomena, and we can look forward to an
adaptive exaggerations of nonpathological enriched understanding of these disorders.
personality styles resulting from predispos-
ing temperaments combined with stressful
circumstances. Neurobiology can be altered REFERENCES
in at least some Axis II disorders, as it can be
in Axis I disorders. Our challenge for the fu- American Psychiatric Association: Diagnostic and
ture is to recognize that not all personality Statistical Manual of Mental Disorders. Wash-
disorders are alike, nor are personality disor- ington, DC, American Psychiatric Association,
ders fundamentally different from many American Psychiatric Association: Diagnostic and
other psychiatric disorders. What may be Statistical Manual of Mental Disorders, 2nd
Personality Disorders: Recent History and the DSM System 11

Edition. Washington, DC, American Psychiat- Gunderson JG: Borderline Personality Disorder.
ric Association, 1968 Washington, DC, American Psychiatric Pub-
American Psychiatric Association: Diagnostic and lishing, 2008
Statistical Manual of Mental Disorders, 3rd Hoch PH, Polatin P: Pseudoneurotic forms of
Edition. Washington, DC, American Psychiat- schizophrenia. Psychiatr Q 23:248276, 1949
ric Association, 1980 Kernberg OF: Borderline Conditions and Patho-
American Psychiatric Association: Diagnostic and logical Narcissism. New York, Jason Aronson,
Statistical Manual of Mental Disorders, 3rd 1975
Edition, Revised. Washington, DC, American Kohut H: The Analysis of the Self: A Systematic
Psychiatric Association, 1987 Approach to the Treatment of Narcissistic Per-
American Psychiatric Association: Diagnostic and sonality Disorder. New York, International
Statistical Manual of Mental Disorders, 4th Universities Press, 1971
Edition. Washington, DC, American Psychiat- Kraepelin E: Lectures on Clinical Psychiatry (En-
ric Association, 1994 glish translation). New York, Wood Press, 1904
American Psychiatric Association: Diagnostic and Kretschmer E: Hysteria (English translation). New
Statistical Manual of Mental Disorders, 4th York, Nervous and Mental Disease Publishers,
Edition, Text Revision. Washington, DC, 1926
American Psychiatric Association, 2000 Millon T: Modern Psychopathology: A Biosocial
Barton WE: The History and Influence of the Approach to Maladaptive Learning and Func-
American Psychiatric Association. Washing- tioning. Philadelphia, PA, WB Saunders, 1969
ton, DC, American Psychiatric Press, 1987 Paris J: Nature and Nurture in Psychiatry: A Pre-
Bleuler E: Textbook of Psychiatry (English transla- disposition-Stress Model of Mental Disorders.
tion). New York, Macmillan, 1924 Washington, DC, American Psychiatric Press,
Deutsch H: Some forms of emotional disturbance 1999
and their relationship to schizophrenia. Psy- Schneider K: Psychopathic Personalities (1923).
choanal Q 11:301321, 1942 London, Cassell, 1950
Freud S: Inhibitions, Symptoms and Anxiety, Skodol AE: Classification, assessment, and differ-
Standard Edition. London, Hogarth Press, ential diagnosis of personality disorders. Jour-
1926 nal of Practical Psychology and Behavioral
Frosch J: The psychotic character: clinical psychi- Health 3:261274, 1997
atric considerations. Psychiatr Q 38:8196, World Health Organization: International Classi-
1964 fication of Diseases, 8th Revision. Geneva,
Gunderson JG: Borderline Personality Disorder. World Health Organization, 1968
Washington, DC, American Psychiatric Press, Zillborg G: Ambulatory schizophrenia. Psychia-
1984 try 4:149155, 1941
Gunderson JG: Borderline Personality Disorder: A
Clinical Guide. Washington, DC, American
Psychiatric Publishing, 2001
This page intentionally left blank
Theories of Personality and
Personality Disorders
Amy Heim, Ph.D.
Drew Westen, Ph.D.

P ersonality refers to enduring patterns of need not be broadly generalized to be con-

cognition, emotion, motivation, and be- sidered aspects of personality (or to lead to
havior that are activated in particular cir- dysfunction), because many aspects of
cumstances (see Mischel and Shoda 1995; personality are triggered by specific situa-
Westen 1995). This minimalist definition tions, thoughts, or feelings. For example, a
(i.e., one that most personality psycholo- tendency to bristle and respond with op-
gists would accept, despite widely differ- position, anger, or passive resistance to
ing theories) underscores two important perceived demands of male authority fig-
aspects of personality. First, personality is ures may or may not occur with female au-
dynamic, characterized by an ongoing in- thorities, peers, lovers, or subordinates.
teraction of mental, behavioral, and envi- Nevertheless, this response tendency rep-
ronmental events. Second, inherent in per- resents an enduring way of thinking, at-
sonality is the potential for variation and tending to information, feeling, and re-
flexibility of responding (activation of spe- sponding that is clearly an aspect of
cific processes under particular circum- personality (and one that can substantially
stances). Enduring ways of responding affect adaptation).

Preparation of this manuscript was supported in part by National Institute of Mental Health grants
MH62377 and MH62378 (Westen).

14 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

Among the dozens of approaches to per- had understood psychological problems in

sonality advanced over the past century, two terms of conflict and defense using Freuds
are of the most widespread use in clinical topographic model (conscious, precon-
practice: the psychodynamic and the cogni- scious, unconscious) or his structural model
tive-social or cognitive-behavioral. Two (id, ego, superego). In classical psychoana-
other approaches have gained increased in- lytic terms, most symptoms reflect maladap-
terest among personality disorder research- tive compromises, forged outside of aware-
ers: trait psychology, one of the oldest and ness, among conflicting wishes, fears, and
most enduring empirical approaches to the moral standards. For example, a patient with
study of normal personality; and biological anorexia nervosa who is uncomfortable with
approaches, which reflect a long-standing her impulses and who fears losing control
tradition in descriptive psychiatry as well as over them may begin to starve herself as a
more recent developments in behavioral ge- way of demonstrating that she can control
netics and neuroscience. Although most the- even the most persistent of desires, hunger.
ories have traditionally fallen into a single Some of the personality disorders currently
camp, several other approaches are best identified in DSM-IV (American Psychiatric
viewed as integrative. These include Ben- Association 1994) and its update, DSM-IV-
jamins (1996a, 1996b) interpersonal ap- TR (American Psychiatric Association 2000),
proach, which integrates interpersonal, psy- have their roots in early psychoanalytic the-
chodynamic, and social learning theories; orizing about conflictnotably dependent,
Millons (1990) evolutionarysocial learning obsessive-compulsive, and to some extent
approach, which has assimilated broadly histrionic personality disorders (presumed
from multiple traditions (e.g., psychoana- to reflect fixations at the oral, anal, and phal-
lytic object relations theory); and Westens lic stages, respectively).
(1995, 1998) functional-domains model, Although some psychoanalysts have ar-
which draws on psychodynamic, evolution- gued that a conflict model can account for se-
ary, behavioral, cognitive, and developmen- vere personality pathology (e.g., Abend et al.
tal research. In this chapter we briefly con- 1983), most analytic theorists have turned to
sider how each approach conceptualizes ego psychology, object relations theory, self
personality disorders. psychology, and relational theories to help
understand patients with personality disor-
ders. According to these approaches, the
PSYCHODYNAMIC THEORIES problems seen in patients with character dis-
orders run deeper than maladaptive compro-
Psychoanalytic theorists were the first to mises among conflicting motives, and reflect
generate a concept of personality disorder derailments in personality development re-
(also called character disorder, reflecting the garding temperament, early attachment ex-
idea that personality disorders involve char- periences, and their interaction (e.g., Balint
acter problems not isolated to a specific 1969; Kernberg 1975b). Many of the DSM-IV
symptom or set of independent symptoms). personality disorders have roots in these
Personality disorders began to draw consid- later approaches, notably schizoid, border-
erable theoretical attention in psychoanaly- line, and narcissistic personality disorders.
sis by the middle of the twentieth century Psychoanalytic ego psychology focuses
(e.g., Fairbairn 1952; Reich 1933/1978), in on the psychological functions (in contempo-
part because they were common and difficult rary cognitive terms, the skills, procedures,
to treat, and in part because they defied un- and processes involved in self-regulation)
derstanding using the psychoanalytic mod- that must be in place for people to behave
els prevalent at the time. For years, analysts adaptively, attain their goals, and meet exter-
Theories of Personality and Personality Disorders 15

nal demands (see Bellak et al. 1973; Blanck stantial body of research supports many of
and Blanck 1974; Redl and Wineman 1951). these propositions, particularly vis--vis bor-
From this perspective, patients with person- derline personality disorder (BPD), the most
ality disorders may have various deficits in extensively studied personality disorder
functioning, such as poor impulse control, (e.g., Baker et al. 1992; Gunderson 2001; Wes-
difficulty regulating affects, and deficits in ten 1990a, 1991a).
the capacity for self-reflection. These deficits From a psychodynamic point of view,
may render them incapable of behaving con- perhaps the most important features of per-
sistently in their own best interest or of tak- sonality disorders are the following: a) they
ing the interests of others appropriately into represent constellations of psychological
account (e.g., they lash out aggressively processes, not distinct symptoms that can be
without forethought or cut themselves when understood in isolation; b) they can be lo-
they become upset). cated on a continuum of personality pathol-
Object relations, relational, and self psy- ogy from relative health to relative sickness;
chological theories focus on the cognitive, af- c) they can be characterized in terms of char-
fective, and motivational processes pre- acter style, which is orthogonal to level of
sumed to underlie functioning in close disturbance (e.g., a patient can have an ob-
relationships (Aron 1996; Greenberg and sessional style but be relatively sick or rela-
Mitchell 1983; Mitchell 1988; Westen 1991b). tively healthy); d) they involve both implicit
From this point of view, personality disor- and explicit personality processes, only some
ders reflect a number of processes. Internal- of which are available to introspection (and
ization of attitudes of hostile, abusive, criti- thus amenable to self-report); and e) they re-
cal, inconsistent, or neglectful parents may flect processes that are deeply entrenched,
leave patients with personality disorder vul- often serve multiple functions, and/or have
nerable to fears of abandonment, self-hatred, become associated with regulation of affects
a tendency to treat themselves as their par- and are hence resistant to change.
ents treated them, and so forth (Benjamin The most comprehensive theory that em-
1996a, 1996b; Masterson 1976; McWilliams bodies these principles is the theory of per-
1998). Patients with personality disorder of- sonality structure or organization developed
ten fail to develop mature, constant, multi- by Otto Kernberg (1975a, 1984, 1996). In his
faceted representations of the self and others. theory, Kernberg proposed a continuum of
As a result, they may be vulnerable to emo- pathology, from chronically psychotic levels
tional swings when significant others are of functioning, through borderline function-
momentarily disappointing, and they may ing (severe personality disorders), through
have difficulty understanding or imagining neurotic to normal functioning. In Kern-
what might be in the minds of the people bergs view, people with severe personality
with whom they interact (Fonagy and Target pathology are distinguished from people
1997; Fonagy et al. 1991, 2003). Those with whose personality is organized at a psy-
personality disorder often appear to have chotic level by their relatively intact capacity
difficulty forming a realistic, balanced view for reality testing (the absence of hallucina-
of themselves that can weather momentary tions or psychotic delusions) and their rela-
failures or criticisms and may have a corre- tive ability to distinguish between their own
sponding inability to activate procedures thoughts and feelings and those of others
(hypothesized to be based on loving, sooth- (the absence of beliefs that their thoughts are
ing experiences with early caregivers) that being broadcast on the radio; their recogni-
would be useful for self-soothing in the face tion, although sometimes less than complete,
of loss, failure, or threats to safety or self- that the persecutory thoughts in their heads
esteem (e.g., Adler and Buie 1979). A sub- are voices from the past rather than true hal-
16 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

lucinations, etc.). What distinguishes indi- According to Kernberg, whereas borderline

viduals with severe personality pathology patients lack an integrated identity, narcissis-
from people with neurotic (that is, health- tic patients are typically developmentally
ier) character structures includes 1) their more advanced, in that they have been able
more maladaptive modes of regulating their to develop a coherent (if distorted) view of
emotions through immature, reality-distort- themselves. Narcissistic phenomena, in
ing defenses such as denial and projection Kernbergs view, lie on a continuum from
(e.g., refusing to recognize the part they play normal (characterized by adequate self-
in generating some of the hostility they en- esteem regulation) to pathological (narcissis-
gender from others); and 2) their difficulty in tic personality disorder) (Kernberg 1984,
forming mature, multifaceted representa- 1998). Individuals with narcissistic personal-
tions of themselves and significant others ity disorder need to construct a grossly in-
(e.g., believing that a person they once loved flated view of themselves to maintain self-
is really all bad, with no redeeming features, esteem and may appear grandiose, sensitive
and is motivated only by the desire to hurt to the slightest attacks on their self-esteem
them). Kernberg refers to these two aspects (and hence vulnerable to rage or depression),
of borderline personality organization as or both. Not only are the conscious self-rep-
primitive defenses and identity diffu- resentations of narcissistic patients inflated
sion. This level of severe personality distur- but so too are the representations that consti-
bance shares some features with the DSM- tute their ideal selves. Actual and ideal self-
IVs BPD diagnosis. However, borderline representations stand in dynamic relation to
personality organization is a broader con- one another. Thus, one reason narcissistic pa-
struct, encompassing patients with para- tients must maintain an idealized view of self
noid, schizoid, schizotypal, and antisocial is that they have a correspondingly grandi-
personality disorders as well as some pa- ose view of who they should be, a divergence
tients who would receive a DSM-IV diagno- that leads to tremendous feelings of shame,
sis of narcissistic, histrionic, or dependent failure, and humiliation.
personality disorder. (Some schizotypal and The concept of a grandiose self is central to
borderline patients may at times fall south the self psychology of Heinz Kohut, a major
of the border into the psychotic range.) theorist of narcissistic personality pathology
Some research supports the notion that pa- whose ideas, like those of Kernberg, contrib-
tients fall on a continuum of severity of per- uted to the DSM-III diagnosis of narcissistic
sonality pathology (see Millon and Davis personality disorder (Goldstein 1985). Ko-
1995; Tyrer and Johnson 1996), with disor- huts theory grew out of his own and others
ders such as paranoid and borderline per- clinical experiences with patients whose prob-
sonality disorder representing more severe lems (such as feelings of emptiness or unstable
forms, and disorders such as obsessive- self-esteem) did not respond well to existing
compulsive personality disorder less severe (psychoanalytic) models. Narcissistic pathol-
(Westen and Shedler 1999a). ogy, according to Kohut, results from faulty
Although many of Kernbergs major con- self-development. Kohuts concept of the self
tributions have been in the understanding of refers to the nucleus of a persons central am-
borderline phenomena, his theory of narcis- bitions and ideals and the talents and skills
sistic disturbance contributed substantially used to actualize them (Kohut 1971, 1977;
to the development of the diagnosis of nar- Wolf 1988). It develops through two pathways
cissistic personality disorder in DSM-III (poles) that provide the basis for self-es-
(American Psychiatric Association 1980), just teem. The first is the grandiose selfan ideal-
as his understanding of borderline phenom- ized representation of self that emerges in chil-
ena contributed to the borderline diagnosis. dren through empathic mirroring by their
Theories of Personality and Personality Disorders 17

parents (Mommy, watch!) and provides the stant interplay between environmental de-
nucleus for later ambitions and strivings. The mands and the way the individual processes
second is the idealized parent imagoan ideal- information about the self and the world
ized representation of the parents that pro- (Bandura 1986).
vides the foundation for ideals and standards Cognitive-social theorists have only re-
for the self. Parental mirroring allows the child cently begun to write about personality dis-
to see his or her reflection in the eyes of a lov- orders (e.g., Beck et al. 2003; Linehan 1993a;
ing and admiring parent; idealizing a parent Pretzer and Beck 1996; Young 1990). In large
or parents allows the child to identify with part this late entrance into the study of per-
and become like them. In the absence of ade- sonality disorders reflects the assumption,
quate experiences with parents who can mir- initially inherited from behaviorism, that
ror the child or serve as appropriate targets of personality is composed of relatively dis-
idealization (for example, when the parents crete, learned processes that are more mallea-
are self-involved or abusive), the childs self- ble and situation specific than implied by the
structure cannot develop, preventing the concept of personality disorder. Cognitive-
achievement of cohesion, vigor, and normal social theories focus on a number of variables
self-esteem (described by Kohut as healthy presumed to be most important in under-
narcissism). As a result, the child develops a standing personality disorders, including
disorder of the self, of which pathological nar- schemas, expectancies, goals, skills and com-
cissism is a prototypic example. petencies, and self-regulation (Bandura 1986,
1999; Cantor and Kihlstrom 1987; Mischel
1973, 1979; Mischel and Shoda 1995). Al-
COGNITIVE-SOCIAL THEORIES though particular theorists have tended to
emphasize one or two of these variables in
Cognitive-social theories (Bandura 1986; explaining personality disorders, such as the
Mischel 1973, 1979) offered the first compre- schemas involved in encoding and process-
hensive alternative to psychodynamic ap- ing information about the self and others
proaches to personality. First developed in (Beck et al. 2003) or the deficits in affect reg-
the 1960s, these approaches are sometimes ulation seen in borderline patients (Linehan
called social learning theories, cognitive- 1993a), a comprehensive cognitive-social ac-
social learning theory, social cognitive theo- count of personality disorders would likely
ries, and cognitive-behavioral theories. Cog- address all of them.
nitive-social theories developed from behav- For example, patients with personality
iorist and cognitive roots. From a behaviorist disorders have dysfunctional schemas that
perspective, personality consists of learned lead them to misinterpret information (as
behaviors and emotional reactions that tend when patients with BPD misread and misat-
to be relatively specific (rather than highly tribute peoples intentions); attend to and en-
generalized) and tied to particular environ- code information in biased ways (as when pa-
mental contingencies. Cognitive-social theo- tients with paranoid personality disorder
ries share the behaviorist belief that learning maintain vigilance for perceived slights or at-
is the basis of personality and that person- tacks); or view themselves as bad or incompe-
ality dispositions tend to be relatively spe- tent (pathological self-schemas). Related to
cific and shaped by their consequences. They these schemas are problematic expectancies,
share the cognitive view that the way people such as pessimistic expectations about the
encode, transform, and retrieve information, world, beliefs about the malevolence of oth-
particularly about themselves and others, is ers, and fears of being mocked. Patients with
central to personality. From a cognitive- personality disorders may have pathological
social perspective, personality reflects a con- self-efficacy expectancies, such as the depen-
18 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

dent patients belief that she or he cannot sur- p. 55). Becks theory highlights three aspects
vive on his or own, the avoidant patients of cognition: 1) automatic thoughts (beliefs
belief that he or she is likely to fail in social and assumptions about the world, the self,
circumstances, or the narcissistic patients and others); 2) interpersonal strategies; and
grandiose expectations about what she or he 3) cognitive distortions (systematic errors in
can accomplish. Equally important are com- rational thinking). Beck and colleagues have
petenciesthat is, skills and abilities used for described a unique cognitive profile charac-
solving problems. In social-cognitive terms, teristic of each of the DSM-IV personality
social intelligence includes a variety of com- disorders. For example, an individual diag-
petencies that help people navigate interper- nosed with schizoid personality disorder
sonal waters (Cantor and Harlow 1994; Can- would have a view of himself as a self-suffi-
tor and Kihlstrom 1987), and patients with cient loner, a view of others as unrewarding
personality disorders tend to be notoriously and intrusive, and a view of relationships as
poor interpersonal problem solvers. messy and undesirable, and his primary in-
Of particular relevance to severe person- terpersonal strategy would involve keeping
ality disorders is self-regulation, which refers his distance from other people (Pretzer and
to the process of setting goals and subgoals, Beck 1996). He would use cognitive distor-
evaluating ones performance in meeting tions that minimize his recognition of how
these goals, and adjusting ones behavior to relationships with others can be sources of
achieve these goals in the context of ongoing pleasure. A study of dysfunctional beliefs (as
feedback (Bandura 1986; Mischel 1990). assessed by the Personal Beliefs Question-
Problems in self-regulation, including a defi- naire [A.T. Beck, J.S. Beck, unpublished as-
cit in specific skills, form a central aspect of sessment instrument, The Beck Institute for
Linehans (1993a, 1993b) work on BPD. Line- Cognitive Therapy and Research, Bala Cyn-
han regards emotion dysregulation as the wyd, Pennsylvania, 1991]) provides some
essential feature of BPD. The key character- initial support for the link between particular
istics of emotion dysregulation include dif- beliefs and the DSM-IV personality disorders
ficulty 1) inhibiting inappropriate behavior (Beck et al. 2001).
related to intense affect, 2) organizing oneself Building on Becks cognitive theory,
to meet behavioral goals, 3) regulating phys- Young and colleagues (Young and Gluhoski
iological arousal associated with intense 1996; Young and Lindemann 2002; Young et
emotional arousal, and 4) refocusing atten- al. 2003) have added a fourth level of cog-
tion when emotionally stimulated (Linehan nition: early maladaptive schemas, which
1993b). Many of the behavioral manifesta- they have defined as broad and pervasive
tions of BPD (e.g., cutting) can be viewed as themes regarding oneself and ones relation-
consequences of emotional dysregulation. ships with others, developed during child-
Deficits in emotion regulation lead to other hood and elaborated throughout ones life
problems, such as difficulties with interper- (Young and Lindemann 2002, p. 95). The au-
sonal functioning and with the development thors distinguish these schemas from auto-
of a stable sense of self. matic thoughts and underlying assumptions,
According to another cognitive-behav- noting that the schemas are associated with
ioral approach, Becks cognitive theory (Beck greater levels of affect, are more pervasive,
1999; Beck et al. 2003; Pretzer and Beck 1996), and involve a strong interpersonal aspect.
dysfunctional beliefs constitute the primary Young and colleagues have identified 16
pathology involved in the personality disor- early maladaptive schemas, each of which
ders (Beck et al. 2001), which are viewed as comprises cognitive, affective, and behav-
pervasive, self-perpetuating cognitive- ioral components. They have also identified
interpersonal cycles (Pretzer and Beck 1996, three cognitive processes involving schemas
Theories of Personality and Personality Disorders 19

that define key features of personality disor- ductive empirical research programs. Traits
ders: schema maintenance, which refers to the are emotional, cognitive, and behavioral ten-
processes by which maladaptive schemas are dencies on which individuals vary (e.g., the
rigidly upheld (e.g., cognitive distortions, tendency to experience negative emotions).
self-defeating behaviors); schema avoidance, According to Gordon Allport (1937), who pi-
which refers to the cognitive, affective, and oneered the trait approach to personality, the
behavioral ways individuals avoid the nega- concept of trait has two separate but comple-
tive affect associated with the schema; and mentary meanings: it is both an observed ten-
schema compensation, which refers to ways of dency to behave in a particular way and an
overcompensating for the schema (e.g., be- inferred underlying personality disposition
coming a workaholic in response to a schema that generates this behavioral tendency. In
of self as failure). the empirical literature, traits have largely
Mischel and Shoda (1995) have offered a been defined operationally, as the average of
compelling social-cognitive account of per- a set of self-report items designed to assess a
sonality that focuses on ifthen contingen- given trait (e.g., items indicating a tendency
ciesthat is, conditions that activate parti- to feel anxious, sad, ashamed, guilty, self-
cular thoughts, feelings, and behaviors. doubting, and angry that all share a common
Although they have not linked this model to core of negative affectivity or neuroticism).
personality disorders, one could view per- Researchers have begun recasting person-
sonality disorders as involving a host of ality disorders in terms of the most promi-
rigid, maladaptive ifthen contingencies. For nent contemporary trait theory, the Five-
example, for some patients, the first hints of Factor Model (FFM) of personality (McCrae
trouble in a relationship may activate con- and Costa 1997; Widiger 2000; Widiger and
cerns about abandonment. These in turn may Costa 1994). (We address other trait models
elicit anxiety or rage, to which the patient re- that have been more closely associated with
sponds with desperate attempts to lure the biological theories later.) The FFM is a de-
person back that often backfire (such as ma- scription of the way personality descriptors
nipulative statements and suicidal gestures). tend to covary and hence can be understood
From an integrative psychodynamic-cogni- in terms of latent factors (traits) identified via
tive viewpoint, Horowitz (1988, 1998) of- factor analysis. Based on the lexical hypothe-
fered a model that similarly focused on the sis of personalitythat important personal-
conditions under which certain states of ity attributes will naturally find expression in
mind become active, which he has tied more words used in everyday languagethe FFM
directly to a model of personality disorders; emerged from factor analysis of adjectival de-
and Wachtel (1977, 1997) has similarly de- scriptions of personality originally selected
scribed cyclical psychodynamics, in which from Websters Unabridged Dictionary (Allport
people manage to elicit from others precisely and Odbert 1936). Numerous studies, includ-
the kind of reactions of which they are the ing cross-cultural investigations, have found
most vigilant and afraid. that when participants in nonclinical (nor-
mal) samples are asked to rate themselves on
dozens or hundreds of adjectives or brief sen-
TRAIT THEORIES tences, the pattern of self-descriptions can of-
ten be reduced to five overarching constructs
Trait psychology focuses less on personality (Costa and McCrae 1997; Goldberg 1993):
processes or functions than do psychody- 1) neuroticism or negative affect (how much
namic or cognitive-social approaches, and they tend to be distressed); 2) extraversion or
hence has not generated an approach to treat- positive affect (the extent to which they tend
ment, although it has generated highly pro- to be gregarious, high-energy, and happy);
20 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

3) conscientiousness; 4) agreeableness; and ever, are that it integrates the understanding

5) openness to experience (the extent to and assessment of normal and pathological
which they are open to emotional, aesthetic, personality and that it establishes dimen-
and intellectual experiences). sions of personality pathology using well-
McCrae and Costa (1990, 1997) proposed understood empirical procedures (factor
a set of lower-order traits, or facets, within analysis).
each of these broadband traits that can allow Another way to proceed using the FFM is
a more discriminating portrait of personality. to translate clinically derived categories into
Thus, an individuals personality profile is five-factor language (Coker et al. 2002;
represented by a score on each of the five fac- Lynam and Widiger 2001; Widiger and Costa
tors plus scores on six lower-order facets or 1994). For example, Widiger et al. (2002) de-
subfactors within each of these broader con- scribed antisocial personality disorder
structs (e.g., anxiety and depression as facets (ASPD) as combining low agreeableness
of neuroticism). Advocates of the FFM argue with low conscientiousness. Because analy-
that personality disorders reflect extreme sis at the level of five factors often lacks the
versions of normal personality traits, so that specificity to characterize complex disorders
the same system can be used for diagnosing such as BPD (high neuroticism plus high ex-
normal and pathological personality. From traversion), proponents of the FFM have of-
the perspective of the FFM, personality dis- ten moved to the facet level. Thus, whereas
orders are not discrete entities separate and all six neuroticism facets (anxiety, hostility,
distinct from normal personality. Rather, depression, self-consciousness, impulsivity,
they represent extreme variants of normal and vulnerability) are characteristic of pa-
personality traits or blends thereof. tients with BPD, patients with avoidant per-
In principle, one could classify personal- sonality disorder are characterized by only
ity disorders in one of two ways using the four of these facets (anxiety, depression, self-
FFM. The first, and that more consistent with consciousness, and vulnerability).
the theoretical and psychometric tradition Similarly, Widiger et al. (1994, 2002) de-
within which the FFM developed, is simply scribed obsessive-compulsive personality
to identify personality pathology by extreme disorder as primarily an extreme, maladap-
values on each of the five factors (and per- tive variant of conscientiousness. They add,
haps on their facets). For example, extremely however, that obsessive-compulsive patients
high scores on the neuroticism factor and its tend to be low on the compliance and altru-
facets (anxiety, hostility, depression, self- ism facets of agreeableness (i.e., they are op-
consciousness, impulsivity, and vulnerabil- positional and stingy) and low on some of
ity) all represent aspects of personality pa- the facets of openness to experience as re-
thology. Whether this strategy is appropriate flected in being closed to feelings and closed
for all factors and facets, and when to con- to values (i.e., morally inflexible). Numerous
sider extreme responses on one or both poles studies have shown predicted links between
of a dimension pathological, are matters of DSM-IV Axis II disorders and FFM factors
debate. Extreme extraversion, for example, and facets (Axelrod et al. 1997; Ross et al.
may or may not be pathological, depending 2002; Trull et al. 2001), although other studies
on the social milieu and the persons other have found substantial overlap among the
traits. Similarly, extreme openness to experi- FFM profiles of patients with very different
ence could imply a genuinely open attitude disorders (e.g., borderline and obsessive-
toward emotions, art, and so forth or an un- compulsive) using major FFM self-report in-
critical, flaky, or schizotypal cognitive ventories (Morey et al. 2002).
style. The advantages of this approach, how-
Theories of Personality and Personality Disorders 21

pulsivity/aggression (impulse control disor-

BIOLOGICAL PERSPECTIVES ders); affective instability (mood disorders);
and anxiety/inhibition (anxiety disorders).
The first biological perspectives on personal- Conceptualized in dimensional terms, Axis I
ity disorders, which influenced the current disorders such as schizophrenia represent
Axis II classification, stemmed from the ob- the extreme end of a continuum. Milder ab-
servations of the pioneering psychiatric tax- normalities can be seen in patients with per-
onomists in the early twentieth century, sonality disorder, either directly (as sub-
notably Bleuler (1911/1950) and Kraepelin threshold variants) or through their influence
(1896/1919). These authors and others on adaptive strategies (coping and defense).
noticed, for example, that the relatives of Siever and Davis linked each dimension
schizophrenic patients sometimes appeared to biological correlates and indicators, some
to have attenuated symptoms of the disorder presumed to be causal and others to provide
that endured as personality traits, such as in- markers of underlying biological dysfunction
terpersonal and cognitive peculiarity. More (e.g., eye movement dysfunction in schizo-
recently, researchers have used the methods phrenia, which is also seen in individuals
of trait psychology (particularly the reliance with schizotypal personality disorder and in
on self-report questionnaires and factor anal- nonpsychotic relatives of schizophrenic
ysis) to study personality disorders from a probands). They also pointed to suggestive
biological viewpoint. In some cases, they data on neurotransmitter functioning that
have developed item sets with biological might link Axis II disorders with Axis I syn-
variables in mind (e.g., neurotransmitters dromes such as depression. More recently,
and their functions) or have reconsidered Siever and colleagues (New and Siever 2002;
patterns of covariation among different traits Siever et al. 2002) proposed an approach to
in light of hypothesized neurobiological sys- BPD that tries to circumvent the problems
tems or circuits. In other cases, they have ap- created by the heterogeneity of the diagnosis
plied behavior-genetic approaches to study by examining the neurobiology of specific di-
personality traits (as well as DSM-IV disor- mensions thought to underlie the disorder
ders). We explore each of these approaches (endophenotypes), especially impulsive ag-
in turn. (Researchers are just beginning to gression and affective instability.
use neuroimaging to study personality dis- The major attempt thus far to develop a
orders, particularly BPD [e.g., Herpertz et al. trait model of personality disorders based on
2001], but the results at this point are prelim- a neurobiological model is Cloninger s
inary, and hence we do not address them fur- seven-factor model of personality (Cloninger
ther here.) 1998; Cloninger et al. 1993). In his model,
Cloninger divided personality structure into
Traits and Neural Systems two domains: temperament (automatic asso-
ciative responses to basic emotional stimuli
Siever and Davis (1991) provided one of the that determine habits and skills) and charac-
first attempts to reconsider the personality ter (self-aware concepts that influence vol-
disorders from a neurobiology perspective. untary intentions and attitudes) (Cloninger
They proposed a model based on core char- 1998, p. 64). According to Cloninger, each of
acteristics of Axis I disorders relevant to per- these domains is defined by a mode of learn-
sonality disorders and related these charac- ing and the underlying neural systems in-
teristics to emerging knowledge of their volved in that learning: temperament is asso-
underlying neurobiology. They focused on ciated with associative/procedural learning,
cognitive/perceptual organization (schizo- and character is associated with insight learn-
phrenia and other psychotic disorders); im- ing. The temperament domain includes four
22 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

dimensions, each theoretically linked to par- emergent phenotypes arising from the inter-
ticular neurotransmitter systems: 1) novelty action of basic neurobehavioral systems that
seeking (exploration, extravagance, impul- underlie major personality traits (Depue and
sivity), associated with dopamine; 2) harm Lenzenweger 2001, p. 165). Through an ex-
avoidance (characterized by pessimism, fear, tensive examination of the psychometric lit-
timidity), associated with serotonin and - erature on the structure of personality traits
aminobutyric acid (GABA); 3) reward depen- as well as a theoretical analysis of the neu-
dence (sentimentality, social attachment, robehavioral systems likely to be relevant to
openness), associated with norepinephrine personality and personality dysfunction,
and serotonin; and 4) persistence (industri- they identified five trait dimensions that may
ousness, determination, ambitiousness, per- account for the range of personality disorder
fectionism), associated with glutamate and phenotypes. They labeled these five traits
serotonin (Cloninger 1998, p. 70). The charac- 1) agentic extraversion (reflecting both the
ter domain includes three dimensions: 1) self- activity and gregariousness components of
directedness (responsibility, purposefulness, extraversion); 2) neuroticism; 3) affiliation;
self-acceptance), considered the major de- 4) nonaffective constraint (the opposite pole
terminant of the presence or absence of per- of which is impulsivity); and 5) fear. For ex-
sonality disorder (Cloninger et al. 1993, ample, the neurobehavioral system underly-
p. 979); 2) cooperativeness (empathy, com- ing the trait of agentic extraversion is posi-
passion, helpfulness); and 3) self-transcen- tive incentive motivation, which is common
dence (spirituality, idealism, enlightenment). to all mammalian species and involves posi-
Cloninger (1998) proposed that all per- tive affect and approach motivation. The
sonality disorders are low on the character dopaminergic system has been strongly im-
dimensions of self-directedness and cooper- plicated in incentive-motivated behavior,
ativeness. What distinguishes patients with such that individual differences in the former
different disorders are their more specific predict differences in the latter. Research on
profiles. In broad strokes, the Cluster A per- this model is just beginning, but the model is
sonality disorders (schizotypal, schizoid, promising in its integration of research on
paranoid) are associated with low reward neural systems involved in fundamental
dependence; the Cluster B personality disor- functions common to many animal species
ders (borderline, antisocial, narcissistic, his- (such as approach, avoidance, affiliation
trionic) are associated with high novelty with conspecifics, and inhibition of punished
seeking; and the Cluster C personality disor- behavior) with individual differences re-
ders (dependent, avoidant, obsessive-com- search in personality psychology.
pulsive) are associated with high harm
avoidance. Individual personality disorders Behavior-Genetic Approaches
may be described more fully by profiles ob-
tained from Cloningers self-report Temper- The vast majority of behavior-genetic studies
ament and Character Inventory (Cloninger of personality have focused on normal per-
and Svrakic 1994). For example, BPD would sonality traits, such as those that compose
consist of high harm avoidance, high novelty the FFM and Eysencks (1967, 1981) three-
seeking, and low reward dependence as well factor model (extraversion, neuroticism, and
as low scores on the character dimensions. psychoticism). These studies have generally
More recently, a dimensional neurobe- shown moderate to high heritability (30%
havioral model was offered by Depue, Len- 60%) for a range of personality traits (Lives-
zenweger, and colleagues (e.g., Depue and ley et al. 1993; Plomin and Caspi 1999) rele-
Collins 1999; Depue and Lenzenweger 2001). vant to personality disorders. The most fre-
Their model regards personality disorders as quently studied traits, extraversion and
Theories of Personality and Personality Disorders 23

neuroticism, have produced heritability esti- the higher- and lower-order factors they
mates of 54%74% and 42%64%, respec- identified.
tively (Eysenck 1990). Compared with research on normal per-
Behavior-genetic data are proving in- sonality traits (as well as many Axis I disor-
creasingly useful in both etiological and tax- ders), behavior-genetic studies of personality
onomic work (e.g., Krueger 1999; Livesley et disorders are relatively rare. The most com-
al. 1998). Livesley et al. (2003) noted that be- mon designs have been family studies in
havior-genetic data can help address the per- which researchers begin with the personality
sistent lack of consensus among trait psy- disorder proband and then assess other fam-
chologists regarding which traits to study by ily members. The major limitation of this
helping them study the causes of trait cova- method is that familial aggregation of disor-
riation (as opposed to simply describing it). ders can support either genetic or environ-
Establishing congruence between a pro- mental causes. As in all behavior-genetic re-
posed phenotypic model of personality traits search, twin and adoption studies provide
and the genetic structure underlying it more definitive data. Most of these studies
would support the validity of a proposed have examined only a subset of the DSM per-
factor model. The same holds true for models sonality disorders, particularly schizotypal,
of personality disorders. antisocial, and borderline personality disor-
To test this approach, Livesley et al. ders. These disorders appear to reflect a con-
(1998) administered the Dimensional Assess- tinuum of heritability, with schizotypal most
ment of Personality PathologyBasic Ques- strongly linked to genetic influences, antiso-
tionnaire to a large sample of individuals cial linked both to environmental and genetic
with and without personality disorders, in- variables, and borderline showing the small-
cluding twin pairs. This self-report measure est estimates of heritability in the majority of
consists of 18 traits considered to underlie studies (see Nigg and Goldsmith 1994).
personality disorder diagnoses (e.g., identity Research on the heritability of schizo-
problems, oppositionality, social avoidance). typal personality disorder provides the clear-
Factor analysis indicated a four-factor solu- est evidence of a genetic component to a per-
tion: emotional dysregulation, dissocial be- sonality disorder. (Schizotypal personality
havior, inhibition, and compulsivity. Results disorder is defined by criteria such as odd be-
also showed high congruence for all four fac- liefs or magical thinking, unusual perceptual
tors between the phenotypic and behavior- experiences, odd thinking and speech, suspi-
genetic analyses, indicating strong support ciousness, inappropriate or constricted af-
for the proposed factor solution. In addition, fect, and behavior or appearance that is odd
the data showed substantial residual herita- or eccentric.) As mentioned earlier, Bleuler
bility for many lower-order traits, suggesting and Kraepelin noted peculiarities in lan-
that these traits likely are not simply compo- guage and behavior among some relatives of
nents of the higher-order factors but include their schizophrenic patients. Bleuler called
unique components (specific factors) as well. this presentation latent schizophrenia and
Krueger and colleagues (e.g., Krueger 1999) considered it to be a less severe and more
have similarly found, using structural equa- widespread form of schizophrenia. Further
tion modeling with a large twin sample, that research into the constellation of symptoms
broadband internalizing and externalizing characteristic of relatives of schizophrenic
personality factors account for much of the patients ultimately resulted in the creation of
variance in many common Axis I disorders the DSM diagnosis of schizotypal personal-
(e.g., mood, anxiety, and substance use) and ity disorder (Spitzer et al. 1979). A genetic re-
that genetic and environmental sources of lationship between schizoph renia an d
variance are associated with many of both schizotypal personality disorder is now well
24 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

established (Kendler and Walsh 1995; Len- nent to several personality disorders, with
zenweger 1998). In one study, Torgersen most heritability estimates between 0.50 and
(1984) found that 33% (7 of 21) of identical co- 0.60, including BPD. Increasingly, research-
twins had schizotypal personality disorder, ers are suggesting that specific components
whereas only 4% (1 of 23) of fraternal co- of BPD may have higher heritability than the
twins shared the diagnosis. Data from a later BPD diagnosis taken as a whole. For exam-
twin study (Torgersen et al. 2000), which ple, several authors (Nigg and Goldsmith
used structural equation modeling, esti- 1994; Widiger and Frances 1994) suggest that
mated heritability at 0.61. neuroticism, which is highly heritable, is at
ASPD, in contrast, appears to have both the core of many borderline features (e.g.,
genetic and environmental roots, as docu- negative affect and stress sensitivity). Other
mented in adoption studies (Cadoret et al. components of BPD have shown substantial
1995). An adult adoptee whose biological heritability as well (e.g., problems with iden-
parent has an arrest record for antisocial be- tity, impulsivity, affective lability) (Livesley
havior is four times more likely to have prob- et al. 1993; Skodol et al. 2002).
lems with aggressive behavior than a person A caveat worth mentioning, however, is
without a biological vulnerability. At the that behavior-genetic studies that systemati-
same time, a person whose adoptive parent cally measure environmental influences di-
has ASPD is more than three times more rectly (e.g., measuring developmental toxins
likely to develop the disorder, regardless of such as sexual abuse), rather than deriving
biological history. As is the case with other estimates of shared and nonshared environ-
behavior-genetic findings, twin studies sug- ment statistically from residual terms, often
gest that environmental and genetic factors obtain very different estimates of environ-
grow more predictive as individuals get mental effects, and this may well be the case
older (Lyons et al. 1995). In considering the with many personality disorders. For exam-
data on ASPD and other personality disor- ple, if one child in a family responds to sex-
ders, however, it is important to remember ual abuse by becoming avoidant and con-
that all estimates of heritability are sample stricted and another responds to the same
dependent. Turkheimer et al. (2003) found, experience by becoming borderline and
for example, that genes account for most of impulsive, researchers will mistakenly con-
the variability in IQ among middle-class chil- cludeunless they actually measured de-
dren but that over 60% of the variance in IQ velopmental variablesthat shared envir-
in samples from low socioeconomic back- onment has no effect, because a shared
grounds reflects shared environment. Socio- environmental event led to nonshared re-
economic status may similarly moderate the sponses (see Turkheimer and Waldron 2000;
relation between genes and environment and Westen 1998). Recent work by Caspi et al.
antisocial behavior. (2002) showing genes and environmental
Data on the behavioral genetics of BPD events (e.g., sexual abuse) interacting in pre-
are mixed. Several studies have found only dicting subsequent personality and psycho-
modest evidence of heritability (e.g., Dahl pathology emphasize the same point.
1993; Nigg and Goldsmith 1994; Reich 1989).
A rare twin study conducted by Torgersen
(1984) failed to find evidence for the genetic INTEGRATIVE THEORIES
transmission of the disorder, although the
sample was relatively small. A more recent Of all the disorders identified in DSM-IV-TR,
twin study by Torgersen et al. (2000) focused the personality disorders are likely to be
on the heritability of several personality dis- among those that most require biopsychoso-
orders, finding a substantial genetic compo- cial perspectives. They are also disorders for
Theories of Personality and Personality Disorders 25

which we may gain substantially by integrat- which refer to the extent to which the person
ing data from both clinical observation and focuses on individuation or nurturance of
research, from classical theories of personal- others and which are reflected in the self/
ity that delineate personality functions, and other polarity; and 4) processes of abstrac-
from more contemporary research that em- tion, which refer to the ability for symbolic
phasizes traits. The emergence of several in- thought and which are represented by the
tegrative models is thus perhaps not surpris- thinking/feeling polarity.
ing. We briefly describe three such models in Millon identified 14 personality proto-
the following discussion: Millons evolution- types that can be understood in terms of the
arysocial learning model, Benjamins inter- basic polarities. For example, patients with
personal model, and Westens functional- schizoid personality disorder tend to have
domains model. little pleasure, to have little involvement
with others, to be relatively passive in their
Millons EvolutionarySocial stance to the world, and to rely on abstract
thinking over intuition. In contrast, patients
Learning Model
with histrionic personality disorder are plea-
Millon developed a comprehensive model of sure seeking, interpersonally focused (al-
personality and personality disorders that he though in a self-centered way), highly active,
initially framed in social learning terms (Mil- and short on abstract thinking. Millons the-
lon 1969), describing personality in terms of ory led to the distinction between avoidant
three polarities: pleasure/pain, self/other, and schizoid personality disorder in DSM-
and passive/active. These polarities reflect III. Whereas schizoid personality disorder
the nature of reinforcement that controls the represents a passive-detached personality
persons behavior (rewarding or aversive), style, avoidant personality disorder repre-
the source or sources that provide reinforce- sents an active-detached style characterized
ment (oneself or others), and the instrumen- by active avoidance motivated by avoidance
tal behaviors and coping strategies used to of anxiety. Millon also developed both a com-
pursue reinforcement (active or passive). prehensive measure to assess the DSM per-
Millon (Davis and Millon 1999; Millon 1990; sonality disorders and his own theory-
Millon and Davis 1996; Millon and Gross- driven personality disorder classification,
man 2005) eventually reconceptualized his the Millon Clinical Multiaxial Inventory
original theory in evolutionary terms. In do- (Millon and Davis 1997). The instrument,
ing so, he added a fourth polarity, thinking/ now in its third edition, has been used in
feeling, which reflects the extent to which hundreds of studies and is widely used as an
people rely on abstract thinking or intuition. assessment tool in clinical practice (e.g., Es-
Millons reconceptualized theory out- pelage et al. 2002; Kristensen and Torgersen
lined four basic evolutionary principles con- 2001).
sistent with the polarities described by his
earlier theory: 1) aims of existence, which re- Benjamins Interpersonal Model
fer to life enhancement and life preservation,
and which are reflected in the pleasure/pain Benjamins (1993, 1996a, 1996b) interpersonal
polarity; 2) modes of adaptation, which he theory, called Structural Analysis of Social
described in terms of accommodation to, ver- Behavior (SASB), focuses on interpersonal
su s m od if ica t ion o f , th e e n v iron me n t processes in personality and psychopathol-
(whether one adjusts or tries to adjust the ogy and their intrapsychic causes, correlates,
world, particularly other people) and which and sequelae. Influenced by Sullivans (1953)
are reflected in the passive/active polarity; interpersonal theory of psychiatry, by object
3) strategies of replication or reproduction, relations approaches, and by research using
26 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

the interpersonal circumplex (e.g., Kiesler neglect or abandonment. Anger in narcissis-

1983; Leary 1957; Schaefer 1965), the SASB is a tic personality disorder tends to follow from
three-dimensional circumplex model with perceived slights or failures of other people to
three surfaces, each of which represents a give the patient everything he or she wants
specific focus. The first surface focuses on ac- (entitlement). Anger in patients with ASPD is
tions directed at a person (e.g., abuse by a often cold, detached, and aimed at control-
parent toward the patient). The second sur- ling the other person. Second, the SASB
face focuses on the persons response to real model is able to represent multiple, often con-
or perceived actions by the other (e.g., recoil- flicting aspects of the way patients with a
ing from the abusive parent). The third sur- given disorder behave (or complex, multifac-
face focuses on the persons actions toward eted aspects of a single interpersonal interac-
him- or herself, or what Benjamin calls the tion) simultaneously. Thus, a single angry
introject (e.g., self-abuse). The notion be- outburst by a borderline patient could reflect
hind the surfaces is that the first two are inter- an effort to get distance from the other, to hurt
personal and describe the kinds of interaction the other, and to get the other to respond and
patterns (self with other) in which the patient hence be drawn back into the relationship.
engages with significant others (e.g., parents, Benjamin has devised several ways of opera-
attachment figures, therapists). The third sur- tionalizing a persons dynamics or an inter-
face represents internalized attitudes and ac- personal interaction (e.g., in a therapy hour),
tions toward the self (e.g., self-criticism that ranging from direct observation and coding
began as criticism from parents). According of behavior to self-report questionnaires, all
to Benjamin, children learn to respond to of which yield descriptions using the same
themselves and others by identifying with circumplex model.
significant others (acting like them), recapitu-
lating what they experienced with significant Westens Functional-Domains
others (e.g., eliciting from others what they
experienced before), and introjecting others
(treating themselves as others have treated Westen (1995, 1996, 1998) described a model
them). of domains of personality functioning that
As with all circumplex models, each sur- draws substantially on psychoanalytic clini-
face has two axes that define its quadrants. In cal theory and observation as well as on em-
the SASB (as in other interpersonal circum- pirical research in personality, cognitive,
plex models), love and hate represent the two developmental, and clinical psychology. Al-
poles of the horizontal axis. Enmeshment and though some aspects of the model are linked
differentiation are the endpoints of the verti- to research on etiology, the model is less a
cal axis. The SASB offers a translation of each theory of personality disorders than an at-
of the DSM Axis II criteria (and disorders) tempt to delineate and systematize the major
into interpersonal terms (Benjamin 1993, elements of personality that define a patients
1996b). In this respect, it has two advantages. personality, whether or not the patient has a
First, it can reduce the number of co-occur- personality disorder. The model differs from
ring conditions that need to be identified in a trait approaches in its focus on personality
given patient by specifying the interpersonal processes and functions (e.g., the kinds of af-
antecedents that elicit the patients responses. fect regulation strategies the person uses, the
For example, maladaptive anger is character- ways she represents the self and others men-
istic of many of the DSM-IV personality dis- tally, as well as more behavioral dispositions,
orders but has different interpersonal triggers such as whether she engages in impulsive or
and meanings (Benjamin 1993). Anger in pa- self-destructive behavior). However, it shares
tients with BPD often reflects perceived with trait approaches the view that a single
Theories of Personality and Personality Disorders 27

model should be able to accommodate rela- dure Q-Sort, a personality pathology mea-
tively healthy as well as relatively disturbed sure for use by expert informants, although
personality styles and dynamics. the model and the measure are not closely
The model suggests that a systematic linked (i.e., one does not require the other).
personality case formulation must answer From this point of view, individuals with
three questions, each composed of a series of particular personality disorders are likely to
subquestions or variables that require assess- be characterized by 1) distinct constellations
ment: 1) What does the person wish for, fear, of motives and conflicts, such as chronic wor-
and value, and to what extent are these mo- ries about abandonment in BPD or a conflict
tives conscious or unconscious, collaborating between the wish for and fear of connected-
or conflicting? 2) What psychological re- ness to others in avoidant personality disor-
sourcesincluding cognitive processes (e.g., der; 2) deficits in adaptive functioning, such
intelligence, memory, intactness of thinking as poor impulse control, lack of self-reflective
processes), affects, affect regulation strate- capacities (see Fonagy and Target 1997), and
gies (conscious coping strategies and uncon- difficulty regulating affect (Linehan 1993a;
scious defenses), and behavioral skillsdoes Westen 1991a) in BPD or subclinical cogni-
the person have at his or her disposal to meet tive disturbances in schizotypal personality
internal and external demands? 3) What is disorder; and 3) problematic ways of think-
the persons experience of the self and others, ing, feeling, and behaving toward them-
and how able is the individualcognitively, selves and significant others, such as a ten-
emotionally, motivationally, and behavior- dency to form simplistic, one-dimensional
allyto sustain meaningful and pleasurable representations of the self and others, to mis-
relationships? understand why people (including the self)
From a psychodynamic perspective, behave as they do, and to expect malevo-
these questions correspond roughly to the is- lence from other people (characteristics seen
sues raised by classical psychoanalytic theo- in patients with many personality disorders,
ries of motivation and conflict (Brenner such as paranoid, schizoid, and borderline)
1982); ego-psychological approaches to (Kernberg 1975a, 1984; Westen 1991a). In this
adaptive functioning; and object-relational, model, a persons level of personality health
self-psychological, attachment, and contem- sickness (from severe personality disorder to
porary relational (Aron 1996; Mitchell 1988) relatively healthy functioning), which can be
approaches to understanding peoples expe- assessed reliably using a personality health
rience of self with others. Each of these ques- prototype or a simple rating of level of per-
tions and subdimensions, however, is also sonality organization derived from Kern-
associated with a number of research tradi- bergs work (Westen and Muderrisoglu 2003;
tions in personality, clinical, cognitive, and Westen and Shedler 1999b), reflects his or her
developmental psychology (e.g., on the de- functioning in each of these three domains.
velopment of childrens representations of People who do not have severe enough
self, representations of others, moral judg- pathology to receive a personality disorder
ment, attachment styles, ability to tell coher- diagnosis can similarly be described using
ent narratives) (see Damon and Hart 1988; this approach. For example, a successful
Fonagy et al. 2002; Harter 1999; Livesley and male executive presented for treatment with
Bromley 1973; Main 1995; Westen 1990a, troubles in his marriage and his relationships
1990b, 1991b, 1994). Westen and Shedler at work, as well as low-level feelings of anx-
(1999a) used this model as a rough theoreti- iety and depression. None of these character-
cal guide to ensure comprehensive coverage istics approached criteria for a personality
of personality domains in developing items disorder (or any Axis I disorders, except the
for the Shedler-Westen Assessment Proce- relatively nondescript diagnosis of adjust-
28 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

ment disorder with mixed anxious and de- very distant, whom he saw twice a
pressed mood). Using this model, one would month and with whom he rarely had
sex. Prior to her, his sexual experiences
note that he was competitive with other peo-
had all been anxiety provoking and
ple, a fact of which he was unaware (Ques- short lived, in every sense.
tion 1); had impressive capacities for self- Mr. A tended to be inhibited in
regulation but was intellectualized and many areas of his life. He was emotion-
afraid of feelings and often used his enjoy- ally constricted and seemed particularly
ment of his work as a way of retreating from uncomfortable with pleasurable feel-
his family (Question 2); and had surprisingly ings. He tended to speak in intellectual-
ized terms about his life and history and
noncomplex representations of others
seemed afraid of affect. He felt stifled in
minds (for a person who could solve nonin- his chosen profession, which did not al-
terpersonal problems in complex ways) and low him to express many of his intellec-
consequently would often became angry and tual abilities or creative impulses. He al-
attack at work without stopping to empa- ternated between overcontrol of his
thize with the other persons perspective impulses, which was his modal stance in
(Question 3). This description is, of course, life, and occasional breakthroughs of
poorly thought-out, impulsive actions
highly oversimplified, but it gives a sense of
(as when he bought an expensive piece
how the model can be used to describe per- of equipment with little forethought
sonality dynamics in patients without a diag- about how he would pay for it).
nosable personality disorder (Westen 1998; Mr. A came from a working-class
Westen and Shedler 1999b). family in Boston and had lost his father,
a policeman, as a young boy. He was
reared by his mother and later by a step-
father with whom he had a positive re-
CASE EXAMPLE lationship. He also described a good re-
lationship with his mother, although
To see how some of the models discussed she, like several members of her ex-
here operate in practice, consider the follow- tended family, struggled with depres-
ing brief case description: sion, and she apparently suffered a
lengthy major depressive episode after
her husbands death.
Mr. A was a man in his early 20s who
came to treatment for lifelong problems
with depression, anxiety, and feelings For purposes of brevity, we briefly expli-
of inadequacy. He was a kind, intro- cate this case from two theoretical stand-
spective, sensitive man who neverthe- points that provide very different approaches
less had tremendous difficulty making to case formulation: the FFM and the func-
friends and interacting comfortably
tional-domains viewpoint. (In clinical prac-
with people. He was constantly worried
that he would misspeak, he would ru- tice, a functional-domains account and a psy-
minate after conversations about what chodynamic account are similar, because the
he had said and the way he was per- former reflects an attempt to systematize and
ceived, and he had only one or two integrate with empirical research [and mini-
friends with whom he felt comfortable.
mal jargon] the major domains emphasized
He wanted to be closer to people, but he
was frightened that he would be re- by classical psychoanalytic, ego-psychologi-
jected and was afraid of his own anger cal, and object-relational/self-psychological/
in relationships. While interacting with relational approaches.)
people (including his therapist), he From a five-factor perspective, the most
would often have a running commen-
salient features of Mr. As personality profile
tary with them in his mind, typically
filled with aggressive content. He was were his strong elevations in neuroticism and
in a 2-year relationship with a woman introversion (low extraversion). He was high
who was emotionally and physically on most of the facets of neuroticism, notably
Theories of Personality and Personality Disorders 29

anxiety, depression, anger, self-conscious- pacity to lead a fulfilling life. He wanted to

ness, and vulnerability. He was low on most connect with people, but he was inhibited by
facets of extraversion as well, particularly social anxiety, feelings of inadequacy, and an
gregariousness, assertiveness, activity, and undercurrent of anger toward people that he
happiness. This combination of high negative could not directly express (which emerged in
affectivity and low positive affectivity, which his running commentaries in his mind).
left him vulnerable to feelings of depression, Although he worried that he would fail oth-
captures his anxious, self-conscious social ers, he always felt somehow unfulfilled in his
avoidance. relationships with them and could be subtly
No other broadband factors describe Mr. critical. He likely had high standards with
A adequately, although specific FFM facets which he compared himself and others and
provide insight into his personality. He was against which both frequently fell short. He
moderately high in agreeableness, being also had trouble handling his anger, aggres-
compliant, modest, and tender-minded; sive impulses, and desires for self-assertion.
however, he was not particularly high on He would frequently behave in passive or
trust, altruism, or straightforwardness (re- self-punitive ways rather than appropriately
flecting his tendency to behave passive- asserting his desires or expressing his anger.
aggressively). He was moderately conscien- This pattern contributed in turn to a linger-
tious, showing moderate scores on the facets ing hostile fantasy life and a tendency at
of orderliness and discipline. He similarly times to behave passive-aggressively.
showed moderate openness to experience, Sex was particularly conflictual for Mr. A,
being artistically oriented but low on com- not only because it forced him into an inti-
fort with feelings. His scores on facets such as mate relationship with another person but
intellectual curiosity would likely be moder- because of his feelings of inadequacy, his dis-
ate, reflecting both an interest and an inhibi- comfort in looking directly at a womans
tion. Indeed, a tendency to receive moderate body (because of his associations to sex and
scores because of opposing dynamics would womens bodies), and his worries that he
be true of his facet scores on several traits, was homosexual. When with a woman, he
such as achievement orientation. frequently worried that he would acciden-
A functional-domains perspective would tally touch her anus and be repulsed, al-
offer a similar summary diagnosis to that of a though interestingly, his sexual fantasies
psychodynamic approach, along with a de- (and humor) had a decidedly anal tone. Ho-
scription of his functioning on the three major mosexual images would also jump into his
domains outlined in the model. In broadest mind in the middle of sexual activity, which
outline, from this point of view Mr. A had a led to considerable anxiety.
depressive, avoidant, and obsessional per- With respect to adaptive resources, Mr. A
sonality style organized at a low-functioning had a number of strengths, notably his im-
neurotic level. In other words, he did not have pressive intellect, a dry sense of humor, a ca-
a personality disorder, as evidenced by his pacity to introspect, and an ability to perse-
ability to maintain friendships and stable em- vere. Nevertheless, his overregulation of his
ployment, but he had considerable psycho- feelings and impulses left him vulnerable to
logical impediments to love, work, and life breakthroughs of anger, anxiety, and impul-
satisfaction, with a predominance of depres- sive action. He distanced himself from emo-
sive, avoidant, and obsessional dynamics. tion, in an effort to regulate anxiety, depres-
With respect to motives and conflicts sion, excitement, and pleasure, which seemed
(and interpersonal issues, around which to him both undeserved and threatening.
many of his conflicts centered), Mr. A had a With respect to his experience of self and
number of conflicts that impinged on his ca- relationships, Mr. As dominant interpersonal
30 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

concerns centered around rejection, shame,

and aloneness. He was able to think about REFERENCES
himself and others in complex ways and to
show genuine care and concern toward other Abend SM, Porder MS, Willick MS: Borderline Pa-
tients: Psychoanalytic Perspectives. New
people, although these strengths were often
York, International Universities Press, 1983
not manifest because of his interpersonal Adler G, Buie D: Aloneness and borderline psy-
avoidance. He had low self-esteem, although chopathology: the possible relevance of child
he had some intellectual awareness that his development issues. Int J Psychoanal 60:83
feelings toward himself were unrealistically 96, 1979
negative. He often voiced identity concerns, Allport G: Personality: A Psychological Interpre-
tation. New York, Henry Holt, 1937
wondering what he was going to do with his Allport G, Odbert H: Trait-names: a psycho-lexical
life and where he would fit in and feeling study, in Psychological Monographs, Vol 47.
adrift without either meaningful work or Princeton, NJ, Psychological Review, 1936
love relationships that were sustaining. (This American Psychiatric Association: Diagnostic and
is, of course, a very skeletal description of Statistical Manual of Mental Disorders, 3rd
Edition. Washington, DC, American Psychiat-
functional domains in Mr. A; for a more thor-
ric Association, 1980
ough description, and an empirical descrip- American Psychiatric Association: Diagnostic and
tion using the Shedler-Westen Assessment Statistical Manual of Mental Disorders, 4th
Procedure Q-Sort, see Westen 1998.) Edition. Washington, DC, American Psychiat-
ric Association, 1994
American Psychiatric Association: Diagnostic and
Statistical Manual of Mental Disorders, 4th
CONCLUSION Edition, Text Revision. Washington, DC,
American Psychiatric Association, 2000
Clearly, there are a number of ways to con- Aron L: A Meeting of Minds: Mutuality in Psycho-
ceptualize personality disorders. The case ex- analysis, Vol 4. New York, Analytic Press, 1996
Axelrod S, Widiger T, Trull T, et al: Relations of
ample applied, in a highly schematic way,
five-factor model antagonism facets with per-
two specific approaches: the FFM and the sonality disorder symptomatology. J Pers As-
functional-domains (or psychodynamic) sess 69:297313, 1997
model. Using these two very different meth- Baker L, Silk KR, Westen D, et al: Malevolence,
ods provides some sense of how one might splitting, and parental ratings by borderlines.
J Nerv Ment Dis 180:258264, 1992
formulate a case from two very different the-
Balint M: The Basic Fault: Therapeutic Aspects of
oretical perspectivesnotably a case on Regression. Evanston, IL, Northwestern Uni-
which Axis II personality disorder diagnostic versity Press, 1969
thresholds would not be met because the pa- Bandura A: Social Foundations of Thought and Ac-
tients pathology is not severe enough for a di- tion. Englewood Cliffs, NJ, Prentice Hall, 1986
agnosis of personality disorder. Nevertheless, Bandura A: Social cognitive theory of personality,
in Handbook of Personality: Theory and Re-
some of the patients problematic personality search, 2nd Edition. Edited by Pervin L, John
attributes obviously affected his well-being O. New York, Guilford, 1999, pp 154196
and Axis I symptoms and contributed to his Beck A: Cognitive aspects of personality disorders
seeking treatment. Theory, research, and this and their relation to syndromal disorders: a
brief case example all suggest that including a psychoevolutionary approach, in Personality
and Psychopathology. Edited by Cloninger
broader range of personality functioning is
CR. Washington, DC, American Psychiatric
crucial as the field evolves and should be one Association, 1999, pp 411429
of the primary goals guiding the revision of Beck A, Butler A, Brown G, et al: Dysfunctional
Axis II in DSM-V to increase its clinical utility beliefs discriminate personality disorders. Be-
and improve the delivery of care. hav Res Ther 39:12131225, 2001
Beck AT, Freeman A, Davis DD, et al: Cognitive
Therapy of Personality Disorders, 2nd Edi-
tion. New York, Guilford, 2003
Theories of Personality and Personality Disorders 31

Bellak L, Chassan JB, Gediman HK, et al: Ego Dahl A: The personality disorders: a critical re-
function assessment of analytic psychother- view of family, twin, and adoption studies.
apy combined with drug therapy. J Nerv Ment J Personal Disord 7 (suppl):8699, 1993
Dis 157:465469, 1973 Damon W, Hart D: Self-Understanding in Child-
Benjamin L[S]: Interpersonal Diagnosis and Treat- hood and Adolescence. New York, Cambridge
ment of Personality Disorders. New York, University Press, 1988
Guilford, 1993 Davis R: Millon: Essentials of his science, theory,
Benjamin LS: Interpersonal Diagnosis and Treat- classification, assessment, and therapy. J Pers
ment of Personality Disorders, 2nd Edition. Assess 72:330352, 1999
New York, Guilford, 1996a Depue R, Collins P: Neurobiology of the structure
Benjamin LS: An interpersonal theory of personal- of personality: dopamine, facilitation of incen-
ity disorders, in Major Theories of Personality tive motivation, and extraversion. Behav
Disorder. Edited by Clarkin JF, Lenzenweger Brain Sci 22:491569, 1999
MF. New York, Guilford, 1996b, pp 141220 Depue R, Lenzenweger M: A neurobehavioral di-
Blanck G, Blanck R: Ego Psychology: Theory and mensional model, in Handbook of Personality
Practice. New York, Columbia University Disorders: Theory, Research, and Treatment.
Press, 1974 Edited by Livesley J. New York, Guilford,
Bleuler E: Dementia Praecox or the Group of 2001, pp 136176
Schizophrenias (1911). New York, Interna- Espelage DL, Mazzeo SE, Sherman R, et al:
tional Universities Press, 1950 MCMI-II profiles of women with eating disor-
Brenner C: The Mind in Conflict. New York, Inter- ders: a cluster analytic investigation. J Per-
national Universities Press, 1982 sonal Disord 16:453463, 2002
Cadoret RJ, Yates WR, Troughton E, et al: Genetic Eysenck H: The Biological Basis of Personality.
environmental interaction in the genesis of ag- Springfield, IL, Charles C Thomas, 1967
gressivity and conduct disorders. Arch Gen Eysenck H: A Model for Personality. New York,
Psychiatry 52:916924, 1995 Springer-Verlag, 1981
Cantor N, Harlow RE: Personality, strategic be- Eysenck HJ: Biological dimensions of personality,
havior, and daily life problem solving. Cur Dir in Handbook of Personality: Theory and Re-
Psychol Sci 3:169172, 1994 search. Edited by Pervin LA. New York, Guil-
Cantor N, Kihlstrom JF: Personality and Social In- ford, 1990, pp 244276
telligence. Englewood Cliffs, NJ, Prentice- Fairbairn WR: Psychoanalytic Studies of the Per-
Hall, 1987 sonality. London, Tavistock, 1952
Caspi A, McClay J, Moffitt T, et al: Role of geno- Fonagy P, Target M: Attachment and reflective
type in the cycle of violence in maltreated chil- function: their role in self-organization. Dev
dren. Science 297:851854, 2002 Psychopathol 9:679700, 1997
Cloninger CR: The genetics and psychobiology of Fonagy P, Steele H, Steele M: Maternal representa-
the seven-factor model of personality, in Biol- tions of attachment during pregnancy predict
ogy of Personality Disorders. Edited by Silk KR the organization of infant-mother attachment
(Review of Psychiatry Series, Vol 17; Oldham at one year of age. Child Dev 62:891905, 1991
JM, Riba MB, series eds). Washington, DC, Fonagy P, Gergely G, Jurist EL, et al: Affect Regu-
American Psychiatric Press, 1998, pp 6392 lation, Mentalization, and the Development of
Cloninger CR, Svrakic D: Differentiating normal the Self. New York, Other Press, 2002
and deviant personality by the seven-factor Fonagy P, Target M, Gergely G, et al: The develop-
personality model, in Differentiating Normal mental roots of borderline personality disor-
and Abnormal Personality. Edited by Strack S, der in early attachment relationships: a theory
Lorr M. New York, Springer, 1994, pp 4064 and some evidence. Psychoanalytic Inquiry
Cloninger CR, Svrakic DM, Przybeck TR: A psy- 23:412459, 2003
chobiological model of temperament and char- Goldberg L: The structure of phenotypic personal-
acter. Arch Gen Psychiatry 50:975990, 1993 ity traits. Am Psychol 48:2634, 1993
Coker L, Samuel D, Widiger T: Maladaptive per- Goldstein W: DSM-III and the narcissistic person-
sonality functioning within the Big Five and ality. Am J Psychother 39:416, 1985
the five-factor model. J Personal Disord Greenberg JR, Mitchell S: Object Relations in Psy-
16:385401, 2002 choanalytic Theory. Cambridge, MA, Harvard
Costa P, McCrae R: Longitudinal stability of adult University Press, 1983
personality, in Handbook of Personality Psy- Gunderson JG: Borderline Personality Disorder: A
chology. Edited by Hogan R, Johnson J. San Clinical Guide. Washington, DC, American
Diego, CA, Academic Press, 1997, pp 269290 Psychiatric Publishing, 2001
32 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

Harter S: The Construction of the Self: A Develop- Psychopathology. Edited by Lenzenweger M,

mental Perspective. New York, Guilford, 1999 Dworkin R. Washington, DC, American Psy-
Herpertz SC, Dietrich TM, Wenning B, et al: Evi- chological Association, 1998, pp 93122
dence of abnormal amygdala functioning in Linehan MM: Cognitive-Behavioral Treatment of
borderline personality disorder: a functional Borderline Personality Disorder. New York,
MRI study. Biol Psychiatry 50: 292298, 2001 Guilford, 1993a
Horowitz M: Introduction to Psychodynamics: A Linehan MM: Skills-Training Manual for Treat-
Synthesis. New York, Basic Books, 1988 ment of Borderline Personality Disorder. New
Horowitz MJ: Cognitive Psychodynamics: From York, Guilford, 1993b
Conflict to Character. New York, Wiley, 1998 Livesley WJ, Bromley DB: Person Perception in
Kendler K, Walsh D: Schizotypal personality dis- Childhood and Adolescence. London, Wiley,
order in parents and the risk for schizophrenia 1973
in siblings. Schizophr Bull 21:4752, 1995 Livesley WJ, Jang KZ, Jackson DN, et al: Genetic
Kernberg OF: Borderline Conditions and Patho- and environmental contributions to dimensions
logical Narcissism. New York, Jason Aronson, of personality disorder. Am J Psychiatry
1975a 150:18261831, 1993
Kernberg OF: Transference and countertransfer- Livesley WJ, Jang KL, Vernon PA: Phenotypic and
ence in the treatment of borderline patients. genetic structure of traits delineating person-
J Natl Assoc Priv Psychiatr Hosp 7:1424, 1975b ality disorder. Arch Gen Psychiatry 55:941
Kernberg OF: Severe Personality Disorders: Psy- 948, 1998
chotherapeutic Strategies. New Haven, CT, Livesley WJ, Jang K, Vernon P: Genetic basis of
Yale University Press, 1984 personality structure, in Handbook of Psy-
Kernberg O: A psychoanalytic theory of personal- chology: Personality and Social Psychology,
ity disorders, in Major Theories of Personality Vol 5. Edited by Millon T, Lerner M. New
Disorder. Edited by Clarkin J, Lenzenweger York, Wiley, 2003, pp 5983
M. New York, Guilford, 1996, pp 106140 Lynam DR, Widiger TA: Using the five-factor
Kernberg O: Pathological narcissism and narcis- model to represent the DSM-IV personality
sistic personality disorder: theoretical back- disorders: an expert consensus approach.
ground and diagnostic classification, in Disor- J Abnorm Psychol 110:401412, 2001
ders of Narcissism: Diagnostic, Clinical, and Lyons M, Toomey R, Faraone S, et al: Correlates of
Empirical Implications. Edited by Ronning- psychosis proneness in relatives of schizo-
stam E. Washington, DC, American Psychiat- phrenic patients. Journal of Abnormal Psy-
ric Association, 1998, pp 2951 chology and Psychotherapy: Theory, Research
Kiesler D: The 1982 interpersonal circle: a taxon- and Practice 104:390394, 1995
omy for complementarity in human transac- Main M: Recent studies in attachment: overview,
tions. Psychol Rev 90:185214, 1983 with selected implications for clinical work, in
Kohut H: The Analysis of the Self: A Systematic Attachment Theory: Social, Developmental,
Approach to the Treatment of Narcissistic Per- and Clinical Perspectives. Edited by Goldberg
sonality Disorders. New York, International S, Muir R, Kerr J. Hillsdale, NJ, Analytic Press,
Universities Press, 1971 1995, pp 407474
Kohut H: The Restoration of the Self. New York, Masterson J: Psychotherapy of the Borderline
International Universities Press, 1977 Adult: A Developmental Approach. New
Kraepelin E: Dementia Praecox and Paraphrenia York, Brunner/Mazel, 1976
(1896). Chicago, IL, Chicago Medical Book Co, McCrae R, Costa P: Personality in Adulthood.
1919 New York, Guilford, 1990
Kristensen H, Torgersen S: MCMI-II personality McCrae R, Costa PL: Personality trait structure as a
traits and symptom traits in parents of chil- human universal. Am Psychol 52:509516, 1997
dren with selective mutism: a case-control McWilliams N: Relationship, subjectivity, and in-
study. J Abnorm Psychol 110:648652, 2001 ference in diagnosis, in Making Diagnosis
Krueger RF: The structure of common mental dis- Meaningful: Enhancing Evaluation and Treat-
orders. Arch Gen Psychiatry 56:921926, 1999 ment of Psychological Disorders. Edited by
Leary T: Interpersonal Diagnosis of Personality: A Barron JW. Washington, DC, American Psy-
Functional Theory and Methodology for Per- chological Association, 1998, pp 197226
sonality Evaluation. Oxford, England, Ronald Millon T: Modern Psychopathology: A Biosocial
Press, 1957 Approach to Maladaptive Learning and Func-
Lenzenweger M: Schizotypy and schizotypic psy- tioning. Philadelphia, PA, WB Saunders, 1969
chopathology, in Origins and Development of Millon T: Toward a New Psychology. New York,
Schizophrenia: Advances in Experimental Wiley, 1990
Theories of Personality and Personality Disorders 33

Millon T, Davis R: Conceptions of personality dis- Reich JH: Familiality of DSM-III dramatic and
orders: historical perspectives, the DSMs, and anxious personality clusters. J Nerv Ment Dis
future directions, in The DSM-IV Personality 177:96100, 1989
Disorders: Diagnosis and Treatment of Mental Reich W: Character Analysis, 3rd Edition (1933).
Disorders. Edited by Livesley WJ. New York, New York, Simon and Schuster, 1978
Guilford, 1995, pp 328 Ross S, Lutz C, Bailley S: Positive and negative
Millon T, Davis R: An evolutionary theory of person- symptoms of schizotypy and the five-factor
ality disorders, in Major Theories of Personality model: a domain and facet level analysis.
Disorder. Edited by Clarkin J, Lenzenweger M. J Pers Assess 79:5372, 2002
New York, Guilford, 1996, pp 221346 Schaefer E: Configurational analysis of childrens
Millon T, Davis R: The MCMI-III: present and fu- reports of parent behavior. J Consult Psychol
ture directions. J Pers Assess 68:6985, 1997 29:552557, 1965
Millon T, Grossman SD: Sociocultural factors, in Siever LJ, Davis KL: A psychobiological perspec-
The American Psychiatric Publishing Textbook tive on the personality disorders. Am J Psychi-
of Personality Disorders. Edited by Oldham JM, atry 148:16471658, 1991
Skodol AE, Bender DS. Washington, DC, Amer- Siever LJ, Torgersen S, Gunderson JG, et al: The
ican Psychiatric Publishing, 2005, pp 223235 borderline diagnosis, III: identifying endophe-
Mischel W: Toward a cognitive social learning re- notypes for genetic studies. Biol Psychiatry
conceptualization of personality. Psychol Rev 51:964968, 2002
39:351364, 1973 Skodol AE, Siever LJ, Livesley WJ, et al: The bor-
Mischel W: On the interface of cognition and per- derline diagnosis II: biology, genetics, and clin-
sonality: beyond the person-situation debate. ical course. Biol Psychiatry 51:951963, 2002
Am Psychol 34:740754, 1979 Spitzer RL, Endicott J, Gibbon M: Crossing the
Mischel W: Personality dispositions revisited and border into borderline personality and border-
revised: a view after three decades, in Hand- line schizophrenia. Arch Gen Psychiatry
book of Personality: Theory and Research. Ed- 36:1724, 1979
ited by Pervin L. New York, Guilford, 1990, Sullivan HS: The Interpersonal Theory of Psychi-
pp 111134 atry. New York, Norton, 1953
Mischel W, Shoda Y: A cognitive-affective system Torgersen S: Genetic and nosological aspects of
theory of personality: reconceptualizing situa- schizotypal and borderline personality disor-
tions, dispositions, dynamics, and invariance ders. Arch Gen Psychiatry 41:546554, 1984
in personality structure. Psychol Rev 102:246 Torgersen S, Lygren S, Oien PA, et al: A twin study
268, 1995 of personality disorders. Compr Psychiatry
Mitchell SA: Relational Concepts in Psychoanaly- 41:416425, 2000
sis: An Integration. Cambridge, MA, Harvard Trull T, Widiger T, Burr R: A structured interview
University Press, 1988 for the assessment of the Five-Factor Model of
Morey LC, Gunderson JG, Quigley BD, et al: The personality: facet-level relations to the Axis II
representation of borderline, avoidant, obses- personality disorders. J Pers 69:175198, 2001
sive-compulsive, and schizotypal personality Turkheimer E, Waldron M: Nonshared environ-
disorders by the five-factor model. J Personal ment: a theoretical, methodological, and quan-
Disord 16:215234, 2002 titative review. Psychol Bull 126:78108, 2000
New A, Siever L: Neurobiology and genetics of Turkheimer E, Haley A, Waldron M, et al: Socio-
borderline personality disorder. Psychiatric economic status modifies heritability of IQ in
Annals of Clinical Psychiatry 32:329336, 2002 young children. Psychol Sci 14:623628, 2003
Nigg JT, Goldsmith H: Genetics of personality dis- Tyrer P, Johnson T: Establishing the severity of per-
orders: perspectives from personality and psy- sonality disorder. Am J Psychiatry 153:1593
chopathology research. Psychol Bull 115:346 1597, 1996
380, 1994 Wachtel P: Psychoanalysis and Behavior Therapy.
Plomin R, Caspi A: Behavioral Genetics and Per- New York, Basic Books, 1977
sonality. New York, Guilford, 1999 Wachtel P: Psychoanalysis, Behavior Therapy, and
Pretzer JL, Beck AT: A cognitive theory of person- the Relational World. Washington, DC, Amer-
ality disorders, in Major Theories of Personal- ican Psychological Association, 1997
ity Disorder. Edited by Clarkin J, Lenzenweger Westen D: The relations among narcissism, ego-
M. New York, Guilford, 1996, pp 36105 centrism, self-concept, and self-esteem: exper-
Redl F, Wineman D: Children Who Hate: The Dis- imental, clinical and theoretical consider-
organization and Breakdown of Behavior ations. Psychoanalysis and Contemporary
Controls. Glencoe, IL, Free Press, 1951 Thought 13:183239, 1990a
34 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

Westen D: Towards a revised theory of borderline Widiger T: Personality disorders in the 21st cen-
object relations: contributions of empirical re- tury. J Personal Disord 14:316, 2000
search. Int J Psychoanal 71:661693, 1990b Widiger T, Costa P: Personality and personality
Westen D: Cognitive-behavioral interventions in disorders. Journal of Abnormal Psychology
the psychoanalytic psychotherapy of border- and Psychotherapy: Theory, Research, and
line personality disorders. Clin Psychol Rev Practice 103:7891, 1994
11:211230, 1991a Widiger T, Frances A: Towards a dimensional
Westen D: Social cognition and object relations. model for the personality disorders, in Person-
Psychol Bull 109:429455, 1991b ality Disorders and the Five-Factor Model of
Westen D: Toward an integrative model of affect Personality. Edited by Costa P, Widiger T.
regulation: applications to social-psychologi- Washington, DC, American Psychological As-
cal research. J Pers 62:641667, 1994 sociation, 1994, pp 1939
Westen D: A clinical-empirical model of personal- Widiger T, Trull TJ, Clarkin JF, et al: A description
ity: life after the Mischelian ice age and the of the DSM-IV personality disorders with the
NEO-lithic era. J Pers 63:495524, 1995 five-factor model of personality, in Personality
Westen D: A model and a method for uncovering Disorders and the Five-Factor Model of Per-
the nomothetic from the idiographic: an alterna- sonality, 2nd Edition. Edited by Costa PT Jr,
tive to the five-factor model? J Res Pers 30:400 Widiger TA. Washington, DC, American Psy-
413, 1996 chological Association, 2002, pp 8999
Westen D: Case formulation and personality diag- Wolf E: Treating the Self: Elements of Clinical Self-
nosis: two processes or one? In Making Diag- Psychology. New York, Guilford, 1988
nosis Meaningful: Enhancing Evaluation and Young J: Cognitive Therapy for Personality Disor-
Treatment of Psychological Disorders. Edited ders: A Schema-Focused Approach. Sarasota,
by Barron JW. Washington, DC, American Psy- FL, Professional Resource Exchange, 1990
chological Association, 1998, pp 111138 Young J, Gluhoski V: Schema-focused diagnosis
Westen D, Muderrisoglu S: Assessing personality for personality disorders, in Handbook of Re-
disorders using a systematic clinical interview: lational Diagnosis and Dysfunctional Family
evaluation of an alternative to structured inter- Patterns. Edited by Kaslow F. Oxford, En-
views. J Personal Disord 17:351369, 2003 gland, Wiley, 1996, pp 300321
Westen D, Shedler J: Revising and assessing Axis Young J, Lindemann M: An integrative schema-fo-
II, part I: developing a clinically and empiri- cused model for personality disorders, in Clin-
cally valid assessment method. Am J Psychia- ical Advances in Cognitive Psychotherapy:
try 156:258272, 1999a Theory and Application. Edited by Leahy R,
Westen D, Shedler J: Revising and assessing Axis Dowd T. New York, Springer, 2002, pp 93109
II, part II: toward an empirically based and Young J, Klosko J, Weishaar M: Schema therapy
clinically useful classification of personality for borderline personality disorder, in Schema
disorders. Am J Psychiatry 156:273285, 1999b Therapy: A Practitioners Guide. New York,
Guilford, 2003
Part II
Clinical Evaluation
This page intentionally left blank
Manifestations, Clinical
Diagnosis, and Comorbidity
Andrew E. Skodol, M.D.

A personality disorder is defined in DSM-IV- and includes paranoid, schizoid, and schizo-
TR as an enduring pattern of inner experi- typal personality disorders. Cluster B, the
ence and behavior that deviates markedly dramatic, emotional, or erratic cluster, in-
from the expectations of the individuals cul- cludes antisocial, borderline, histrionic, and
ture, is pervasive and inflexible, has an onset narcissistic personality disorders. Cluster C,
in adolescence or early adulthood, is stable the anxious and fearful cluster, includes
over time, and leads to distress or impair- avoidant, dependent, and obsessive-compul-
ment (American Psychiatric Association sive personality disorders. DSM-IV-TR also
2000, p. 685). Personality disorders are re- provides for a residual category of personality
ported on Axis II of the DSM-IV-TR multiax- disorder not otherwise specified (PDNOS).
ial system to ensure that consideration is This category is to be used when a patient
given to their presence in all patient evalua- meets the general criteria for a personality dis-
tions, even when Axis I disorder psychopa- order and has features of several different
thology is present and prominent. types but does not meet criteria for any spe-
DSM-IV-TR includes criteria for the diag- cific personality disorder (i.e., mixed per-
nosis of 10 specific personality disorders, ar- sonality disorder) or is considered to have a
ranged into three clusters based on descrip- personality disorder not included in the offi-
tive similarities. Cluster A is commonly cial classification (e.g., self-defeating or de-
referred to as the odd or eccentric cluster pressive personality disorders).

Sections of this chapter have been modified with permission from Skodol AE: Problems in Differential Di-
agnosis: From DSM-III to DSM-III-R in Clinical Practice. Washington, DC, American Psychiatric Press, 1989

38 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

This chapter considers the manifesta- with other people and about themselves.
tions, problems in differential diagnosis, and Most of the DSM-IV-TR diagnostic criteria for
patterns of comorbidity of the DSM-IV-TR paranoid personality disorder reflect a distur-
personality disorders. Included in this chap- bance in cognition, characterized by perva-
ter are descriptions of the clinical characteris- sive distrust and suspiciousness of others. Pa-
tics of the 10 DSM-IV-TR personality disor- tients with paranoid personality disorder
ders; discussions of problems in interviewing suspect that others are exploiting, harming, or
the patient with a suspected personality dis- deceiving them; doubt the loyalty or trust-
order in state versus trait discrimination, trait worthiness of others; read hidden, demean-
versus disorder distinctions, categorical ver- ing, or threatening meanings into benign re-
sus alternative classificatory approaches to marks or events; and perceive attacks on their
personality disorder diagnosis, and diagno- character or reputation. Among the major
sis based on inferential judgments; and an symptoms of schizotypal personality disorder
overview of personality disorder comorbid- are characteristic cognitive and perceptual
ity. Despite limitations in the DSM approach, distortions, such as ideas of reference; odd be-
personality disorders diagnosed by this sys- liefs and magical thinking (e.g., superstitious-
tem have been shown since the 1980s to have ness, belief in clairvoyance or telepathy);
considerable clinical utility in predicting bodily illusions; and suspiciousness and para-
functional impairment over and above that noia similar to that observed in patients with
associated with comorbid Axis I disorders, paranoid personality disorder.
extensive and intensive utilization of treat- Patients with borderline personality dis-
ment resources, and in many cases, adverse order (BPD) may also experience transient
outcomes. paranoid ideation when under stress, but the
characteristic cognitive manifestations of
borderline patients are dramatic shifts in
DEFINING FEATURES OF their views toward people with whom they
are intensely emotionally involved. These
PERSONALITY DISORDERS shifts emanate from disturbances in mental
representations of self and others (Bender
Patterns of Inner
and Skodol 2007) and result in their overide-
Experience and Behavior alizing others at one point and then devalu-
ating them at another point, when they feel
The general diagnostic criteria for a person-
disappointed, neglected, or uncared for. This
ality disorder in DSM-IV-TR (see Table 11)
phenomenon is commonly referred to as
indicate that a pattern of inner experience
splitting. Patients with narcissistic person-
and behavior is manifest by characteristic
ality disorder exhibit a grandiose sense of
patterns of 1) cognition (i.e., ways of perceiv-
self; have fantasies of unlimited success,
ing and interpreting self, other people, and
power, brilliance, beauty, or ideal love; and
events); 2) affectivity (i.e., the range, inten-
sity, lability, and appropriateness of emo- believe that they are special or unique.
tional response); 3) interpersonal function- Patients with avoidant personality disor-
ing; and 4) impulse control. Patients with der have excessively negative opinions of
themselves, in contrast to patients with nar-
personality disorders are expected to have
cissistic personality disorder. They see them-
manifestations in at least two of these areas.
selves as inept, unappealing, and inferior,
and they constantly perceive that they are be-
Cognitive Features ing criticized or rejected. Patients with depen-
Personality disorders commonly affect the dent personality disorder also lack self-confi-
ways patients think about their relationships dence and believe that they are unable to
Manifestations, Clinical Diagnosis, and Comorbidity 39

make decisions or to take care of themselves. ety in social situations; those with dependent
Patients with obsessive-compulsive person- personality disorder are preoccupied by anx-
ality disorder (OCPD) are perfectionistic and iety over the prospects of separation from car-
rigid in their thinking and are often preoccu- egivers and the need to be independent.
pied with details, rules, lists, and order.
Interpersonal Features
Affective Features
Interpersonal problems are probably the
Some patients with personality disorders are most typical of personality disorders (Ben-
emotionally constricted, whereas others are jamin 1996; Gunderson 2007; Hill et al. 2008;
excessively emotional. Among the con- Kiesler 1996). Other mental disorders are
stricted types are patients with schizoid per- characterized by prominent cognitive or af-
sonality disorder, who experience little plea- fective features or by problems with impulse
sure in life, appear indifferent to praise or control. All personality disorders, however,
criticism, and are generally emotionally cold, also have interpersonal manifestations that
detached, and unexpressive. Patients with can be described along the two orthogonal
schizotypal personality disorder also often poles of the so-called interpersonal circum-
have constricted or inappropriate affect, al- plex: dominance versus submission and af-
though they can exhibit anxiety in relation to filiation versus detachment.
their paranoid fears. Patients with OCPD Personality disorders characterized by a
have considerable difficulty expressing lov- need for or a tendency toward dominance in
ing feelings toward others, and when they do interpersonal relationships include antisocial,
express affection, they do so in a highly con- histrionic, narcissistic, and obsessive-compul-
trolled or stilted manner. sive. ASPD patients deceive and intimidate
Among the most emotionally expressive others for personal gain. Patients with histri-
patients with personality disorders are those onic and narcissistic personality disorders
with borderline and histrionic personality need to be the center of attention and require
disorders. Patients with BPD are emotionally excessive admiration, respectively. OCPD pa-
labile and react very strongly, particularly in tients need to control others and have them
interpersonal contexts, with a variety of in- submit to their ways of doing things. On the
tensely dysphoric emotions, such as depres- submissive side are patients with avoidant
sion, anxiety, or irritability. They are also and dependent personality disorders. Patients
prone to inappropriate, intense outbursts of with avoidant personality disorder are inhib-
anger and are often preoccupied with fears of ited in interpersonal relationships because
being abandoned by those they are attached they are afraid of being shamed or ridiculed.
to and reliant upon. Patients with histrionic Patients with dependent personality disorder
personality disorder often display rapidly will not disagree with important others for
shifting emotions that seem to be dramatic fear of losing their support or approval and
and exaggerated but are shallow in compari- will actually do things that are unpleasant, de-
son to the intense emotional expression seen meaning, or self-defeating in order to receive
in BPD. Patients with antisocial personality nurturance from them. BPD patients may al-
disorder (ASPD) characteristically have prob- ternate between submissiveness and domi-
lems with irritability and aggressive feelings nance, seeming to become deeply involved
toward others, which are expressed in the and dependent only to turn manipulative and
context of threat or intimidation. Patients demanding when their needs are not met.
with narcissistic personality disorder display In the domain of affiliation versus detach-
arrogant, haughty attitudes and have no em- ment, patients with histrionic, narcissistic,
pathy for other people. Patients with avoidant and dependent personality disorders have
personality disorder are dominated by anxi- the greatest degrees of affiliative behavior,
40 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

whereas patients with paranoid, schizoid, addition, patients with BPD engage in recur-
schizotypal, avoidant, and obsessive-com- rent suicidal threats, gestures, or attempts
pulsive personality disorders are the most de- and in self-mutilating behavior such as cut-
tached. Patients with histrionic, narcissistic, ting or burning. Finally, patients with BPD
and dependent personality disorders are pro- have problems with anger management,
social because of their needs for attention, ad- have frequent temper outbursts, and at times
miration, and support, respectively. Patients may even engage in physical fights.
with paranoid personality disorder do not In contrast, patients with avoidant per-
trust others enough to become deeply in- sonality disorder are generally inhibited, es-
volved; patients with schizotypal personality pecially in relation to people, and are re-
disorder have few friends or confidants, in luctant to take risks or to undertake new
part from a lack of trust and in part as a result activities. Patients with dependent personal-
of poor communication and inadequate relat- ity disorder cannot even make decisions and
edness. Patients with avoidant personality do not take initiative to start things. Patients
disorder are socially isolated because of their with OCPD are overly conscientious and
feelings of inadequacy and their fears of rejec- scrupulous about morality, ethics, and val-
tion, whereas those with schizoid personality ues; they cannot bring themselves to throw
disorder neither desire nor enjoy relation- away even worthless objects and are miserly.
ships. Patients with OCPD opt for work and The DSM-IV-TR personality disorder
productivity over friendships and interper- clusters, specific personality disorder types,
sonal activity because they feel more in con- and their principal defining clinical features
trol in the former than the latter. Patients with are presented in Table 31.
BPD again can vacillate between being overly
attached and dependent on someone (often Pervasiveness and Inflexibility
one who is not the best match) and being iso-
lated, distant, and aloof. For a personality disorder to be present, the
disturbances reviewed earlier have to be
Problems With Impulse Control manifest frequently over a wide range of be-
haviors, feelings, and perceptions and in
Problems with impulse control can also be many different contexts. In DSM-IV-TR, at-
viewed as extremes on a continuum. Person- tempts are made to stress the pervasiveness
ality disorders characterized by a lack of im- of the behaviors caused by personality disor-
pulse control include ASPD and BPD. Disor- ders. Added to the basic definition of each
ders involving problems with overcontrol personality disorder, serving as the stem
include avoidant, dependent, and obsessive- to which individual features apply, is the
compulsive personality disorders. ASPD is a phrase present in a variety of contexts. For
prototype of a personality disorder charac- example, the essential features of paranoid
terized by impulsivity. Patients with ASPD personality disorder in DSM-IV-TR, preced-
break laws, exploit others, fail to plan ahead, ing the specific criteria, begin: A pervasive
get into fights, ignore commitments and ob- distrust and suspiciousness of others such
ligations, and exhibit generally reckless be- that their motives are interpreted as malevo-
haviors without regard to consequences, lent, beginning by early adulthood and
such as speeding, driving while intoxicated, present in a variety of contexts, as indicated
having impulsive sex, or abusing drugs. Pa- by four (or more) of the following (Ameri-
tients with BPD also show many problems can Psychiatric Association 2000, p. 694).
with impulse control, including impulsive Similarly, for dependent personality disor-
spending, indiscriminate sex, substance der, the criteria are preceded by the descrip-
abuse, reckless driving, and binge eating. In tion: A pervasive and excessive need to be
Table 31. DSM-IV-TR personality clusters, specific types, and their defining clinical features

Cluster Type Characteristic features

A Odd or eccentric
Paranoid Pervasive distrust and suspiciousness of others such that their motives are interpreted as malevolent

Manifestations, Clinical Diagnosis, and Comorbidity

Schizoid Pervasive pattern of detachment from social relationships and restricted range of expression of emotions
in interpersonal settings
Schizotypal Pervasive pattern of social and interpersonal deficits marked by acute discomfort with, and reduced
capacity for, close relationships as well as by cognitive or perceptual distortions and eccentricities of
B Dramatic, emotional, or erratic
Antisocial History of conduct disorder before age 15; pervasive pattern of disregard for and violation of the rights
of others; current age at least 18
Borderline Pervasive pattern of instability of interpersonal relationships, self-image, and affects, and marked
Histrionic Pervasive pattern of excessive emotionality and attention seeking
Narcissistic Pervasive pattern of grandiosity (in fantasy or behavior), need for admiration, and lack of empathy
C Anxious or fearful
Avoidant Pervasive pattern of social inhibition, feelings of inadequacy, and hypersensitivity to negative
Dependent Pervasive and excessive need to be taken care of that leads to submissive and clinging behavior and fears
of separation
Obsessive- Pervasive pattern of preoccupation with orderliness, perfectionism, and mental and interpersonal
compulsive control at the expense of flexibility, openness, and efficiency

Source. Adapted from American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision. Washington, DC,
American Psychiatric Association, 2000, p. 685. Used with permission. Copyright 2000 American Psychiatric Association.

42 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

taken care of that leads to submissive and havior (BPD), may be evidenced much less
clinging behavior and fears of separation, be- frequently than traits such as views self as
ginning by early adulthood and present in a socially inept, personally unappealing or infe-
variety of contexts, as indicated by five (or rior to others (avoidant personality disor-
more) of the following (American Psychiat- der). How stable individual manifestations of
ric Association 2000, p. 725). personality disorders actually are and what
Inflexibility is a feature that helps to dis- the stable components of personality disor-
tinguish personality traits or styles and per- ders are have become areas of active empirical
sonality disorders. Inflexibility is indicated research. It may be that personality psychopa-
by a narrow repertoire of responses that are thology waxes and wanes depending on the
repeated even when the situation calls for an circumstances of a persons life (see Chapter 4,
alternative behavior or in the face of clear ev- Course and Outcome).
idence that a behavior is inappropriate or not
working. For example, an obsessive-compul- Distress or Impairment in
sive person rigidly adheres to rules and orga- Functioning
nization even in recreation and loses enjoy-
ment as a consequence. An avoidant person Another important aspect of personality dis-
is so fearful of being scrutinized or criticized, orders that distinguishes them from traits or
even in group situations in which he or she styles is that personality disorders lead to
could hardly be the focus of such attention, distress or impairment in functioning. By
that life becomes painfully lonely. their nature, some personality disorders may
not be accompanied by obvious subjective
distress on the part of the patient. Examples
Onset and Clinical Course would include schizoid personality disorder,
Personality and personality disorders have in which a patient is ostensibly satisfied with
traditionally been assumed to reflect stable his or her social isolation and does not seem
descriptions of a person, at least after a certain to need or desire the companionship of oth-
age. Thus, the patterns of inner experience ers, and ASPD, in which the patient has utter
and behaviors described earlier are called disdain and disregard for social norms and
enduring. Personality disorder is also de- will not experience distress unless his activi-
scribed as of long duration, with an onset ties are thwarted. On the other side of the
that can be traced back at least to adolescence coin are patients with BPD, who are likely to
or early adulthood (American Psychiatric experience and express considerable dis-
Association 2000, p. 686). These concepts per- tress, especially when disappointed in a sig-
sist as integral to the definition of personality nificant other, or patients with avoidant per-
disorder despite a large body of empirical ev- sonality disorder, who, in contrast with
idence that suggests that personality disorder schizoid patients, are usually very uncom-
psychopathology is not as stable as the DSM fortable and unhappy with their lack of close
definition would indicate. Longitudinal stud- friends and companions.
ies indicate that personality disorders tend to All personality disorders are maladap-
improve over time, at least from the point of tive, however, and are accompanied by func-
view of their overt clinical signs and symp- tional problems in school or at work, in social
toms (Grilo et al. 2004; Johnson et al. 2000; relationships, or at leisure. The requirement
Lenzenweger 1999; Zanarini et al. 2006). Fur- for impairment in psychosocial functioning is
thermore, personality disorder criteria sets codified in DSM-IV-TR in its criterion C of the
consist of combinations of pathological per- general diagnostic criteria for a personality
sonality traits and symptomatic behaviors disorder, which states that the enduring pat-
(McGlashan et al. 2005; Zanarini et al. 2007). tern [of inner experience and behavior, i.e.,
Some behaviors, such as self-mutilating be- personality] leads to clinically significant dis-
Manifestations, Clinical Diagnosis, and Comorbidity 43

tress or impairment in social, occupational, or even beyond apparent improvement in per-

other important areas of functioning (Amer- sonality disorder psychopathology itself
ican Psychiatric Association 2000, p. 689). (Seivewright et al. 2004; Skodol et al. 2005).
A number of studies have compared pa- The persistence of impairment is under-
tients with personality disorders to patients standable if one considers that personality
with no personality disorder or with Axis I disorder psychopathology has usually been
disorders and have found that patients with long-standing and, therefore, has disrupted a
personality disorders were more likely to be persons work and social development over
functionally impaired (Skodol and Gunder- a period of time (Roberts et al. 2003). The
son 2008). Specifically, they are more likely to scars or residua of personality disorder pa-
be separated, divorced, or never married and thology take time to heal or be overcome.
to have had more unemployment, frequent With time (and treatment), however, im-
job changes, or periods of disability. It is inter- provements in functioning can occur.
esting that only rarely have patients with per-
sonality disorders been found to be less well
educated. Fewer studies have examined APPROACHES TO
quality of functioning, but in those that have,
poorer social functioning or interpersonal re-
lationships and poorer work functioning or Interviewing a patient to assess for a possible
occupational achievement and satisfaction personality disorder presents certain chal-
have been found among patients with per- lenges that are somewhat unique. Thus, the
sonality disorders than with other disorders. interviewer is likely to need to rely on a vari-
When patients with different personality dis- ety of techniques for gathering information to
orders were compared with each other on arrive at a clinical diagnosis, including obser-
levels of functional impairment, those with vation and interaction with the patient, direct
severe personality disorders such as schizo- questioning, and interviewing informants.
typal and borderline were found to have sig-
nificantly more impairment at work, in social Observation and Interaction
relationships, and at leisure than patients
with less severe personality disorders, such One problem in evaluating a patient for a per-
as OCPD, or with an impairing Axis I disor- sonality disorder arises from the fact that
der, such as major depressiv e disorder most people are not able to view their own
(MDD) without personality disorder. Patients personality objectively (Zimmerman 1994).
with avoidant personality disorder had inter- Because personality is, by definition, the way
mediate levels of impairment. Even the less a person sees, relates to, and thinks about
impaired patients with personality disorders himself or herself and the environment, a per-
(e.g., OCPD), however, had moderate to se- sons assessment of his or her own personal-
vere impairment in at least one area of func- ity must be colored by it. The expression of
tioning (or a Global Assessment of Function- Axis I psychopathology may also be colored
ing rating of 60 or less) (Skodol et al. 2002). by Axis II personality stylefor example,
The finding that significant impairment may symptoms exaggerated by the histrionic or
be in only one area suggests that patients minimized by the compulsive personality
with personality disorders differ not only in but the symptoms of Axis I disorders are usu-
the degree of associated functional impair- ally more clearly alien to the patient and more
ment but also in the breadth of impairment easily identified as problematic. People usu-
across functional domains. ally learn about their own problem behavior
Another important aspect of the impair- and their patterns of interaction with others
ment in functioning in patients with person- through the reactions or observations of other
ality disorders is that it tends to be persistent people in their environments.
44 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

Traditionally, clinicians have not con- or contextthat is, the evaluation itself. Al-
ducted the same kind of interview in assess- though the interaction of patient and clini-
ing patients suspected of having a personal- cian can be a useful and objective observa-
ity disturbance as they do with persons tion, caution should be used in interpreting
suspected of having, for example, a mood or its significance, and attempts must be made
an anxiety disorder. Rather than directly to integrate this information into a broader
questioning the patient about characteristics overall picture of patient functioning.
of his or her personality, the clinician, assum-
ing that the patient cannot accurately de- Direct Questioning
scribe these traits, looks for patterns in the
way the patient describes social relations and In psychiatric research, a portion of the poor
work functioning. These two areas usually reliability of personality disorder diagnosis
give the clearest picture of personality style in has been assumed to be due to the variance in
general and personality problems specifi- information resulting from unsystematic as-
cally. Clinicians have also relied heavily on sessment of personality traits. Therefore, ef-
their observations of how patients interact forts have been made to develop various
with them during an evaluation interview or structured methods for assessing personality
in treatment as manifestations of their pa- disorders (McDermut and Zimmerman 2008)
tients personalities (Westen 1997). comparable with those that have been suc-
These approaches have the advantage of cessful in reducing information variance in
circumventing the lack of objectivity patients assessing Axis I disorders (Kobak et al. 2008).
might have about their personalities, but they These methods include both 1) self-report
also create problems. The clinician usually measures such as the Personality Diagnostic
comes away with a global impression of the Questionnaire4 (Hyler 1994), the Millon
patients personality but frequently is not Clinical Multiaxial InventoryIII (Millon et
aware of many of that patients specific per- al. 1997), and the Minnesota Multiphasic Per-
sonality characteristics because he or she has sonality Inventory2 (Somwaru and Ben-
not made a systematic assessment of the signs Porath 1995); and 2) clinical interviews such
and symptoms of the wide range of personal- as the Structured Interview for DSM-IV Per-
ity disorders (Blashfield and Herkov 1996; sonality (Pfohl et al. 1997), the International
Morey and Ochoa 1989; Zimmerman and Personality Disorder Examination (Loranger
Mattia 1999). In routine clinical practice, clini- 1999), the Structured Clinical Interview for
cians tend to use the nonspecific DSM-IV-TR DSM-IV Axis II Personality Disorders (First
diagnosis of PDNOS when they believe that a et al. 1997), the Diagnostic Interview for
patient meets the general criteria for a per- DSM-IV Personality Disorders (Zanarini et
sonality disorder, because they often do not al. 1996), and the Personality Disorder Inter-
have enough information to make a specific viewIV (Widiger et al. 1995).
diagnosis (Verheul and Widiger 2004). Alter- The interviews are based on the general
natively, clinicians will diagnose personality premise that the patient can be asked specific
disorders hierarchically: once a patient is seen questions that will indicate the presence or
as having one (usually severe) personality absence of each of the criteria of each of the
disorder, the clinician will not assess whether 10 DSM-IV-TR personality disorder types.
traits of other personality disorders are pres- The self-report instruments are generally
ent (Adler et al. 1990; Herkov and Blashfield considered to require a follow-up interview
1995). because of a very high rate of apparently
Reliance on interaction with the clinician false-positive responses, but data from stud-
for personality diagnosis runs the risk of gen- ies comparing self-report measures with clin-
eralizing a mode of interpersonal relating ical interviews suggest that the former aid in
that may be limited to a particular situation identification of personality disturbances
Manifestations, Clinical Diagnosis, and Comorbidity 45

(Hyler et al. 1990, 1992). Thus, the clinician a report about the patient with his or her own
can keep in mind that patients do not neces- personality style. In fact, the correspondence
sarily deny negative personality attributes: in between patient self-assessments of person-
fact, the evidence suggests that they may ality disorder psychopathology and infor-
even overreport traits that clinicians might mant assessments has been generally found
not think are very important, and that pa- to be modest at best (Klonsky et al. 2002).
tients can, if asked, consistently describe a Agreement on pathological personality
wide range of personality traits to multiple traits, temperament, and interpersonal prob-
interviewers. A self-report inventory might lems appears to be somewhat better than on
be an efficient way to help focus a clinical in- DSM personality disorders. Informants usu-
terview on a narrower range of personality ally report more personality psychopathol-
disorder psychopathology. A semistructured ogy than patients. Self/informant agreement
interview is useful clinically when the results on personality disorders is highest for Clus-
of an assessment might be subject to close ter B disorders (excluding narcissistic per-
scrutiny, such as in child custody, disability, sonality disorder), lower for Clusters A and
or forensic evaluations (Widiger and Coker C, and lowest for traits related to narcissism
2002). and entitlement, as might be expected. So the
clinician must make a judgment about the
objectivity of the informant and use this as a
Interviewing Informants
part, but not a sufficient part, of the overall
data on which to base a personality disorder
Frequently, a patient with a personality dis-
diagnosis (Zimmerman et al. 1988). Which
order consults a mental health professional
source, the patient or the informant, provides
for evaluation or treatment because another
information that is more useful for clinical
person has found his or her behavior prob-
purposes, such as choosing a treatment or
lematic. This person may be a boss, spouse,
predicting outcome (e.g., Klein 2003), is yet
boyfriend or girlfriend, teacher, parent, or
representative of a social agency. Indeed, to be definitively determined.
some people with personality disorders do
not even recognize the problematic aspects
of their manner of relating or perceiving ex- PROBLEMS IN CLINICAL ASSESSMENT
cept as it has a negative effect on someone
with whom they interact. Assessing Pervasiveness
Because of these blind spots that peo-
ple with personality disorders may have, the The pervasiveness of personality distur-
use of a third-party informant in the evalua- bance can be difficult to determine. When a
tion can be useful (Zimmerman et al. 1986). clinician inquires if a person often has a
In some treatment settings, such as a private particular experience, a patient will fre-
individual psychotherapy practice, it may be quently reply sometimes, which then has
considered counterproductive or contraindi- to be judged for clinical significance. What
cated to include a third party, but in many in- constitutes a necessary frequency for a par-
patient and outpatient settings, certainly ticular trait or behavior (Widiger 2002) and
during the evaluation process, it may be ap- in how many different contexts or with how
propriate and desirable to see some person many different people the trait or behavior
close to the patient to corroborate both the needs to be expressed has not been well
patients report and ones own clinical im- worked out. Clinicians are forced to rely on
pressions. their own judgment, keeping in mind also
Of course, there is no reason to assume that maladaptivity and inflexibility are hall-
that the informant is bias-free or not coloring marks of pathological traits.
46 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

For the clinician interviewing a patient prompted the person to seek treatment, the
with a possible personality disorder, data consequences in terms of the decision to seek
about the many areas of functioning, the in- treatment, the current level of stress, and any
terpersonal relationships with people inter- actual Axis I psychopathology, if present.
acting in different social roles with the pa- The DSM-IV-TR multiaxial system is of
tient, and the nature of the patientclinician considerable aid in the assessment of these
relationship should be integrated into a com- problems because of its separation of Axis I
prehensive assessment of pervasiveness. Too disorders from Axis II disorders and its indi-
often, clinicians place disproportionate im- vidual axes for physical disorders and psy-
portance on a patients functioning at a par- chosocial stressors. A multiaxial system
ticular job or with a particular boss or signif- forces clinicians to think about the effects of
icant other person. aspects of patients current state on long-term
patterns of behavior, but it does not make the
State Versus Trait distinctions for them.
It is not clear from the diagnostic criteria
An issue that cuts across all personality disor- of DSM-IV-TR how long a pattern of person-
der diagnoses and presents practical prob- ality disturbance needs to be present, or
lems in differential diagnosis is the distinc- when it should become evident, for a person-
tion between clinical state and personality ality disorder to be diagnosed. Earlier itera-
trait. Personality is presumed to be an endur- tions of the DSM stated that patients were
ing aspect of a person, yet assessment of per- usually age 18 years or older when personal-
sonality ordinarily takes place cross-section- ity disorder was diagnosed because it can be
allythat is, over a brief interval in time. argued that, up to that age, a personality pat-
Thus, the clinician is challenged to separate tern could neither have been manifest long
out long-term dispositions of the patient from enough nor have become significantly en-
other more immediate or situationally deter- trenched to be considered a stable constella-
mined characteristics. This task is more com- tion of behavior. DSM-IV-TR states, however,
plicated by the fact that the patient often that some manifestations of personality dis-
comes for evaluation when there is some par- order are usually recognizable by adoles-
ticularly acute problem, which may be a so- cence or earlier and that personality dis-
cial or job-related crisis or the onset of an Axis orders can be diagnosed in persons younger
I disorder (Shea 1997). In either case, the situ- than age 18 years who have manifestated
ation in which the patient is being evaluated symptoms for at least 1 year. Longitudinal
is frequently a state that is not completely research has shown that personality disorder
characteristic of the patients life over the symptoms evident in childhood or early ad-
longer run. olescence may not persist into adult life
(Johnson et al. 2000). Longitudinal research
Assessing an Enduring Pattern has also shown that there is continuity be-
tween certain disorders of childhood and ad-
DSM-IV-TR indicates that personality disor- olescence and personality disorders in early
ders are of long duration and are not better adulthood (Kasen et al. 1999, 2001). Thus, a
accounted for as a manifestation or conse- young boy with oppositional defiant or at-
quence of another mental disorder (Ameri- tention-deficit/hyperactivity disorder in
can Psychiatric Association 2000, p. 689). childhood may go on to develop conduct dis-
Making these determinations in practice is order as an adolescent, which can progress to
not easy. First of all, an accurate assessment full-blown ASPD in adulthood (Bernstein et
requires recognition of current state. An as- al. 1996; Lewinsohn et al. 1997; Rey et al.
sessment of current state, in turn, includes 1995; Zoccolillo et al. 1992). ASPD is the only
knowledge of the circumstances that have diagnosis not given before age 18; an adoles-
Manifestations, Clinical Diagnosis, and Comorbidity 47

cent exhibiting significant antisocial behav- pression, for example, may cause a person to
ior before age 18 years is diagnosed with con- seem excessively dependent, avoidant, or
duct disorder. self-defeating. Cyclothymia or bipolar disor-
Regarding the course of a personality dis- der (not otherwise specified; bipolar II) may
order, DSM-IV-TR states that personality dis- lead to periods of grandiosity, impulsivity,
orders are relatively stable over time, al- poor judgment, and depression that might be
though certain of them (e.g., ASPD and BPD) confused with manifestations of narcissistic
may become somewhat attenuated with age, or borderline personality disorders.
whereas others may not or may, in fact, be- The clinician must be aware of the Axis I
come more pronounced (e.g., obsessive-com- psychopathology and attempt to assess Axis
pulsive and schizotypal personality disor- II independently. This assessment can be at-
ders). As mentioned earlier and discussed in tempted in one of two ways. First, the clini-
greater detail in Chapter 4, Course and Out- cian can ask about aspects of personality func-
come, this degree of stability may not neces- tioning at times when the patient is not
sarily pertain to all of the features of all DSM- experiencing Axis I symptoms. This approach
IV-TR personality disorders equally. is feasible when the Axis I disorder is of recent
To assess stability retrospectively, the cli- onset and short duration or, if more chronic, if
nician must ask questions about periods of a the course of the disorder has been character-
persons life that are of various degrees of re- ized by relatively clear-cut episodes with
moteness from the current situation. Retro- complete remission and symptom-free peri-
spective reporting is subject to distortion, ods of long duration. When the Axis I disor-
however, and the only sure way of demon- der is chronic and unremitting, then the Axis I
strating stability over time is, therefore, to do psychopathology and personality function-
prospective follow-up evaluations. Thus ing blend together to an extent that makes dif-
from a practical, clinical point of view, per- ferentiating between them clearly artificial.
sonality disorder diagnoses made cross- A second approach to distinguishing
sectionally and on the basis of retrospectively signs of Axis I pathology from signs of Axis II
collected data would be tentative or provi- personality is longitudinal and would defer
sional pending confirmation by longitudinal an Axis II diagnosis pending the outcome of a
evaluation. On an inpatient service, a period trial of treatment for the Axis I disorder. This
of intense observation by many professionals strategy may be the preferred approach in the
from diverse perspectives may suffice to es- case of a long-standing and chronic Axis I dis-
tablish a pattern over time (Skodol et al. 1988, order, like cyclothymia, that has never been
1991). In a typical outpatient setting in which previously recognized or treated. Although
encounters with the patients are much less one always runs the risk of a partial response
frequent, more time may be required. Ideally, to treatment and some residual symptoms,
features of a personality disorder should be this tactic may bring the clinician as close,
evident over years, but it is not practical to practically speaking, as he or she will get to
wait inordinate amounts of time before observing the patients baseline functioning.
coming to a diagnostic conclusion. The following case is adapted from Skodol
Assessing the Effect of an Axis I Disorder
Case Example
An Axis I disorder can complicate the diagno-
sis of a personality disorder in several ways
A 24-year-old unemployed man sought
(Widiger and Sanderson 1995; Zimmerman psychiatric hospitalization because of a
1994). An Axis I disorder may cause changes serious problem with depression. The
in a persons behavior or attitudes that can ap- man reported that he had felt mildly,
pear to be signs of a personality disorder. De- but continuously, depressed since the
48 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

age of 16. When he reached his twenties, him, and that he now carried a chip on
he had begun to have more severe bouts his sho ulder because he had been
that made him suicidal and unable to burned by others so often. He admit-
function. ted that he was not blame-free in rela-
During the most recent episode, be- tionships because he had also used peo-
ginning about 6 months previously, he ple, especially members of his family.
had quit his job as a taxi driver and iso- The patient felt that he was not im-
lated himself from his friends. He spent proving in his outpatient treatment of
his time lying around and eating a lot the last 3 years. His reason for seeking
and, in fact, had gained 60 pounds. He hospitalization, in addition to the fact
had difficulty falling asleep, felt fa- that he continually thought of suicide
tigued all day long, could not concen- and was frightened he might actually
trate, felt worthless (Theres no pur- succeed in killing himself, was that he
pose to my life) and guilty (I missed felt totally lost in his life, without di-
my chances; Ive put my family through rection, goals, or knowing what mat-
hell), and had taken an overdose of tered to him. He said he felt hollow.
sleeping pills. If they cut me open after I was dead,
The man received a semistructured he said, theyd probably find out I was
interview assessment of Axis II psycho- all shriveled up inside.
pathology. In describing his personality, This mans description of his per-
he said that he once thought of himself sonality, the ways in which he charac-
as lively and good-natured, but that teristically thought about himself, saw
over the past 4 or 5 years, he felt he had others and his relationships to them, and
changed. He said that he was very sen- behaved, actually met DSM-IV-TR crite-
sitive to criticism, afraid to get involved ria for avoidant, dependent, paranoid,
with people, fearful of new places and and borderline personality disorders. He
experiences, convinced he was making was hospitalized for long-term treat-
a fool of himself, and afraid of losing ment, which was available at the time. In
control. He felt very dependent on oth- addition to receiving individual, psy-
ers for decision making and for initia- choanalytically oriented psychotherapy
tive. He said that he was so needy of sessions and participating in a variety of
others that they could do anything to therapeutic groups, he was given fluox-
him and he would take it. He felt etine, up to 80 mg/day, for treatment of
helpless when alone, was sure he would Axis I MDD and dysthymia.
end up alone and in the streets, and Six months after admission, the pa-
was constantly looking to others, espe- tient reported that he felt significantly
cially family members, for comfort and less depressed. Measured in terms of the
reassurance. Hamilton Rating Scale for Depression,
The man also thought that people the initial severity of his depression was
took advantage of him now and that he 30, and his posttreatment score was 10.
let them because he never stood up for A repeat semistructured assessment of
his own self-interest. He felt like a total his personality functioning revealed
failure with no redeeming virtues. He that he no longer met DSM-IV-TR crite-
said he either deliberately passed up op- ria for any personality disorder, al-
portunities to improve his situation be- though he continued to exhibit some de-
cause he felt I dont deserve any better pendent traits.
or else undermined himself without
thinking by failing to follow through, Another example of the way in which
for example, on a job interview. He be- Axis I and II disorders interact to obscure dif-
lieved that no one could really be trus- ferential diagnosis is the case of apparent Axis
ted, that old friends probably talked
II psychopathology that, in fact, is the pro-
about him behind his back (They think
Im a slob), that he could not open up drome of an Axis I disorder. Distinguishing
with new people because they too Cluster A personality disorders, such as para-
would eventually turn on him and reject noid, schizoid, and schizotypal, from the
Manifestations, Clinical Diagnosis, and Comorbidity 49

early signs of Axis I disorders in the schizo- For the patient with a diagnosis of schizo-
phrenia and other psychotic disorders class typal personality disorder, the occurrence of a
can be particularly difficult. If a clinician is 1-month-long psychotic episode almost cer-
evaluating a patient early in the course of the tainly means the disturbance will meet the
initial onset of a psychotic disorder, he or she criteria for schizophrenia, the symptoms of
may be confronted with changes in the person schizotypal personality disorder counting
toward increasing suspiciousness, social as prodromal symptoms toward the 6-month
withdrawal, eccentricity, or reduced function- duration requirement. Under these circum-
ing. Because the diagnosis of psychotic disor- stances, the diagnosis of schizophrenia, with
ders, including schizophrenia, requires that its pervasive effects on cognition, perception,
the patient have an episode of active psycho- functional ability, and so on, is sufficient, and
sis with delusions and hallucinations, it is not a diagnosis of schizotypal personality disor-
possible to diagnose this prodrome as a psy- der is redundant. When the patient becomes
chotic disorder. In fact, until the full-blown nonpsychotic again, he or she would be con-
disorder is present, the clinician cannot be cer- sidered to have residual schizophrenia in-
tain if it is, indeed, a prodrome. stead of schizotypal personality disorder.
If a change in behavior is of recent onset,
then it does not meet the stability criteria for Personality Traits Versus
a personality disorder. In such cases, the cli-
Personality Disorders
nician is forced to diagnose an unspecified
mental disorder (nonpsychotic; DSM-IV-TR Another difficult distinction is between per-
code 300.9). If, however, the pattern of suspi- sonality traits or styles and personality disor-
ciousness or social withdrawal with or with- ders. All patientsall people for that mat-
out eccentricities has been well established, it tercan be described in terms of distinctive
may legitimately be a personality disorder patterns of personality, but all do not neces-
and be diagnosed as such. sarily warrant a diagnosis of personality dis-
If the clinician follows such a patient over order. This error is particularly common
time and the patient develops a full-fledged among inexperienced evaluators. The im-
psychotic disorder, the personality distur- portant features that distinguish pathologi-
bance is no longer adequate for a complete di- cal personality traits from normal traits are
agnosis because none of the Axis II disorders their inflexibility and maladaptiveness, as
includes frankly psychotic symptoms. This discussed earlier.
fairly obvious point is frequently overlooked DSM-IV-TR recognizes that it is impor-
in practice. All of the personality disorders tant to describe personality style as well as to
that have counterpart psychotic disorders on diagnose personality disorder on Axis II.
Axis I have milder symptoms in which reality Therefore, instructions are included to list
testing is, at least in part, intact. For instance, personality features on Axis II even when a
a patient with paranoid personality disorder personality disorder is absent, or to include
may have referential ideas but not frank delu- them as modifiers of one or more diagnosed
sions of reference, and a patient with schizo- personality disorders (e.g., BPD with histri-
typal personality disorder may have illusions onic features). In practice, however, this op-
but not hallucinations. A possible exception tion has been seldom utilized (Skodol et al.
is BPD, in which brief psychotic experiences 1984), even though research has shown that,
(lasting minutes to an hour or two at most) in addition to the approximately 50% of
are included in the diagnostic criteria. In all clinic patients who meet criteria for a person-
cases, however, when the patient becomes ality disorder, another 35% warrant informa-
psychotic for even a day or two, an additional tion descriptive of their personality styles on
Axis I diagnosis is necessary. Axis II (Kass et al. 1985). The overlap among
50 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

the features of personality disorders also be- tially was agreeable to both parties. She
comes very evident when emphasis is placed craved affection and attention and fell
on the assessment of traits of all personality deeply in love with these men. How-
ever, she eventually became overtly
disorders, even when one is predominant.
self-centered, demanding, and manipu-
The following case example (adapted lative, and the man would break off the
from Skodol 1989) describes a patient with an relationship. After breaking up, she
Axis I disorder whose ongoing treatment would almost immediately start claim-
was very much affected by Axis II personal- ing that the particular man was going
ity traits, none of which met criteria for a per- nowhere, was not for her, and would
sonality disorder. not be missed. In between these rela-
tionships, she often had periods in
which she engaged in a succession of
Case Example
one-night stands, having sex with a
half-dozen partners in a month. Alter-
A 25-year-old, single female receptionist natively, she would frequent rock clubs
was referred for outpatient therapy fol- and bars, in-spots, as she called them,
lowing hospitalization for her first merely on the chance of meeting some-
manic episode. The patient had attended one who would introduce her to the
college for 1 year but dropped out in or- glamorous world she dreamed of.
der to go into advertising. Over the The patient had no female friends
next 5 years, she had held a series of re- other than her sister. She could see little
ceptionist, secretarial, and sales jobs, use for such friendships. She preferred
each of which she quit because she spending her time shopping for stylish
wasnt getting ahead in the world. She clothes or watching television alone at
lived in an apartment on the north side home. She liked to dress fashionably
of Chicago, by herself, that her parents and seductively but often felt that she
had furnished for her. She ate all of her was too fat or that her hair was the
meals, however, at her mothers house wrong color. She had trouble control-
and claimed not even to have a box of ling her weight and would periodically
crackers in her cupboard. Between her go on eating binges for a few days that
jobs, her parents paid her rent. might result in a 10-pound weight gain.
Her career problems stemmed She read popular novels but had very
from the fact that, although she felt few other interests. She admitted she
quite ordinary and without talent for was bored much of the time but would
the most part, she had fantasies of a ca- not admit that cultural or athletic pur-
reer as a movie star or high fashion suits were other than a waste of time.
model. She took acting classes and sing- This patient was referred for outpa-
ing lessons but had never had even a tient follow-up without an Axis II per-
small role in a play or show. What she sonality disorder diagnosis. In fact, her
desired was not so much the careers long-term functioning failed to meet
themselves but the glamour associated DSM-IV-TR criteria for any specific type
with them. Although she wanted to of personality disorder. On the other
move in the circles of the beautiful hand, she almost met the criteria for
people, she was certain that she had several, especially BPD: the patient
nothing to offer them. She sometimes showed signs of impulsivity (overeat-
referred to herself as nothing but a shell ing, sexual promiscuity), intense inter-
and scorned herself because of it. She personal relationships (manipulative,
was unable to picture herself working overidealization/devaluation), identity
her way up along any realistic career disturbance, and chronic feelings of
line, feeling both that it would take too emptiness. She did not, however, dis-
long and that she would probably fail. play intense anger, intolerance of being
She had had three close relation- alone, physically self-damaging behav-
ships with men that were characterized ior, stress-related paranoia or dissocia-
by an intense interdependency that ini- tion, or affective instability indepen-
Manifestations, Clinical Diagnosis, and Comorbidity 51

dent of her mood disorder. Similarly, passivity, emotionality, emphasis on work

she had symptoms of histrionic person- and productivity, and unusual beliefs and rit-
ality disorder: she was inappropriately uals are culturally sanctioned. Only when
sexually seductive and used her physi-
such behaviors are clearly in excess or discor-
cal appearance to draw attention to her-
self, but she was not emotionally over-
dant with the standards of a persons cultural
dramatic. She had shallow expression milieu would the diagnosis of a personality
of emotions and was uncomfortable disorder be considered. Certain sociocultural
when she was not the center of atten- contexts may lend themselves to eliciting and
tion, but was not overly suggestible. reinforcing behaviors that might be mistaken
She also had some features of narcissis- for personality disorder psychopathology.
tic, avoidant, and dependent personal- Members of minority groups, immigrants, or
ity disorders. The attention paid to per-
refugees, for example, might appear overly
sonality traits in her outpatient clinic
evaluation conveyed a vivid picture of guarded or mistrustful, avoidant, or hostile in
the patients complicated personality response to experiences of discrimination,
pathology, which became the focus of language barriers, or problems in accultura-
her subsequent therapy. tion (Alarcon 2005).

Effects of Gender, Culture, and Age
Although personality disorders are usually
Gender not diagnosed prior to age 18 years, certain
thoughts, feelings, and behaviors suggestive
Although definitive estimates about the sex of personality psychopathology may be ap-
ratio of personality disorders cannot be made parent in childhood. For example, depen-
because ideal epidemiological studies do not dency, social anxiety and hypersensitivity,
exist, some personality disorders are believed disruptive behavior, or identity problems
to be more common in clinical settings among may be developmentally expected. Follow-
men and others among women. Personality up studies of children have shown decreases
disorders listed in DSM-IV-TR as occurring in such behaviors over time (Johnson et al.
more often among men are paranoid, schiz- 2000), although children with elevated rates
oid, schizotypal, antisocial, narcissistic, and of personality disordertype signs and symp-
obsessive-compulsive personality disorders. toms do appear to be at higher risk for both
Those occurring more often in women are Axis I and Axis II disorders in young adult-
borderline, histrionic, and dependent person- hood (Johnson et al. 1999; Kasen et al. 1999).
ality disorders. Avoidant is said to be equally Thus, some childhood problems may not
common in men and women. Apparently el- turn out to be transitory, and personality dis-
evated sex ratios that do not reflect true prev- order may be viewed developmentally as a
alence rates can be the result of sampling or failure to mature out of certain age-appropri-
diagnostic biases in clinical settings (Widiger ate or phase-specific feelings or behaviors. A
1998). True differences may be due to biolog- developmental perspective on personality
ical factors such as hormones, social factors disorders is presented more fully in Chapter
such as child-rearing practices, and their in- 7, Developmental Issues.
teractions (Morey et al. 2005).
Other Aspects of
Culture Personality Functioning
Apparent manifestations of personality dis- A problem with the DSM conceptualizations
orders must be considered in the context of a of personality disorders is that the individual
patients cultural reference group and the de- categories do not correspond well with exist-
gree to which behaviors such as diffidence, ing treatment approaches. Thus, whether a
52 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

clinician is a psychodynamically oriented includes a Defensive Functioning Scale and a

therapist, a cognitive-behavioral therapist, Glossary of Specific Defense Mechanisms
or a psychopharmacologist, information in and Coping Styles. The 27 defense mecha-
addition to that necessary for a DSM person- nisms defined in this glossary are acting-out,
ality disorder diagnosis is needed to formu- affiliation, altruism, anticipation, autistic
late a treatment plan. Usually, this additional fantasy, denial, devaluation, displacement,
information is based on the theory of why a dissociation, help-rejecting complaining, hu-
patient has a personality disorder and/or the mor, idealization, intellectualization, isola-
mechanisms responsible for perpetuating tion of affect, omnipotence, passive aggres-
the dysfunctional patterns. sion, projection, projective identification,
rationalization, reaction formation, repres-
Conflicts, Ego Functions, Object sion, self-assertion, self-observation, split-
Relations, and Defense Mechanisms ting, sublimation, suppression, and undoing.
Some defense mechanisms, such as projec-
Psychodynamically oriented clinicians have tion, splitting, or acting-out, are always mal-
expressed dissatisfaction with the DSM sys- adaptive, whereas others, such as sublima-
tem of axes, including Axis II, since its incep- tion or humor, are adaptive. Patients with
tion. The DSM multiaxial system fails, in personality disorders have characteristic
their opinion, to discriminate between pa- predominant defensive patterns. Thus pa-
tients according to clinical variables impor- tients with paranoid personality disorder use
tant for planning treatment with psychody- denial and projection, those with BPD typi-
namic psychotherapy (Karasu and Skodol cally rely on acting-out and splitting (among
1980). Thus, they may be more interested in others), and those with OCPD use isolation
exploring conflicts between wishes, fears, of affect and undoing (Perry and Bond 2005).
and moral standards; ego functions such as Clinicians may note current defenses or cop-
impulse control or affect regulation; or self ing styles as well as a patients predominant
and other (object) representations based on current defense level using the Defensive
early attachment experiences than on the Functioning Scale.
signs and symptoms of personality disor-
ders. Elaborations of psychodynamic theo- Coping Styles
ries of personality disorders can be found in
Chapter 2, Theories of Personality and Per- Although defense mechanisms in DSM-IV-
sonality Disorders, and Chapter 10, Psy- TR are said to include coping styles, the liter-
choanalysis and Psychodynamic Psycho- ature on coping discusses styles not included
therapy, along with discussions of relevant in the DSM list. Coping refers to specific
clinical variables. thoughts and behaviors that a person uses to
Several groups of researchers (Bond and manage the internal and external demands
Vaillant 1986; Perry and Cooper 1989; Vail- of situations appraised as stressful (Folkman
lant et al. 1986) have been able to document and Moskowitz 2004; Lazarus and Folkman
empirically the clinical utility of categorizing 1984; Pearlin and Schooler 1978). Coping in-
a patients defensive functioning. Defense volves cognitive, behavioral, and emotional
mechanisms are automatic psychological pro- responses and may or may not be consistent
cesses that protect people against anxiety across stressful situations or functional roles.
and against awareness of internal or external Two major broad styles of coping are prob-
stressors or dangers. Although this work was lem-focused coping and emotion-focused
considered too early in its development to coping. Problem-focused coping refers to ef-
justify including a separate official axis based forts to resolve a threatening problem or di-
on it, Appendix B in DSM-IV-TR (Criteria minish its impact by taking direct action.
Sets and Axes Provided for Further Study) Emotion-focused coping refers to efforts to re-
Manifestations, Clinical Diagnosis, and Comorbidity 53

duce the negative emotions aroused in re- and judgment necessary to make many of the
sponse to a threat by changing the way the diagnoses. Numerous critics have noted that it
threat is attended to or interpreted. Meaning- is easy to disagree about symptoms such as af-
focused and social coping are other observed fective instability, self-dramatization, shallow
coping strategies. Coping has traditionally emotional expression, exaggerated fears, or
been assessed by retrospective self-report feelings of inadequacyall symptoms of
measures (e.g., the Coping Responses Inven- DSM-IV-TR personality disorders. Only the
tory [Moos 1993], the Ways of Coping Ques- antisocial criteria, among the personality dis-
tionnaire [Folkman and Lazarus 1988], and orders, have historically yielded acceptable
the COPE Inventory [Carver et al. 1989]) levels of reliability, and those criteria have em-
and more recently by ecological momentary phasized overtly criminal and delinquent acts.
assessment (real-time) techniques (Stone et These observations led several investiga-
al. 1998); but the major types of coping, such tors to attempt to determine sets of behaviors
as problem solving, seeking support, dis- that might serve to identify types of person-
tancing and distracting, accepting responsi- ality disorder. Although any one behavior
bility, positive reappraisal, or self-blame, can might not be sufficient to indicate a particu-
also be assessed by clinical interview. lar personality trait, multiple behavioral in-
dicators considered together would increase
Cognitive Schemas confidence in recognizing the trait.
Behaviors that typify a particular person-
Cognitive therapists want to characterize
ality style have been referred to as prototypi-
patients with personality disorders accord-
cal. Livesley (1986) developed a set of proto-
ing to patients dysfunctional cognitive sche-
typical behaviors for the DSM-III (American
mas (core beliefs by which they process
Psychiatric Association 1980) personality dis-
information) or their automatic thoughts,
orders and compared them with prototypical
interpersonal strategies, and cognitive dis-
traits. He found that highly prototypical be-
tortions. Again, particular personality disor-
haviors could be derived from corresponding
ders tend to have particular core beliefs. For
traits. For example, with regard to the con-
example, patients with BPD frequently have
cepts of social awkwardness and withdrawal
beliefs such as I am needy and weak or I
of the schizoid personality disorder, Livesley
am helpless if left on my own, whereas pa-
found that behaviors such as does not speak
tients with OCPD believe It is important to
unless spoken to, does not initiate social
do a perfect job on everything or People
contacts, and rarely reveals self to others
should do things my way (Beck et al. 2004).
were uniformly rated as highly prototypical.
In contrast to beliefs, which map onto per-
Corresponding to the overly dramatic and
sonality disorders specifically, schemas are
emotional traits of the histrionic personality
broader themes regarding the self and rela-
disorder were behaviors such as expressed
tionships with others and can cut across per-
feelings in an exaggerated way, considered
sonality disorder categories. For example, a
a minor problem catastrophic, and flirted
schema of impaired limits can encompass
with several members of the opposite sex.
the entitlement of narcissistic personality
Behaviors such as has routine schedules and
disorder as well as the lack of self-control of
is upset by deviations, overreacted to criti-
ASPD or BPD (Young and Klosko 2005).
cism, and spent considerable time on the
minutest details corresponded to the con-
Objective Behaviors trolled, perfectionist traits of OCPD.
Versus Inferential Traits DSM-IV-TR makes strides in translating
the characteristic traits of the personality dis-
Another difficulty in diagnosing personality orders into explicit behaviors. The criteria for
disorders stems from the degree of inference each personality disorder begin with the def-
54 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

inition of the overall style or set of traits, fol- describe a patients psychopathology and
lowed by a listing of ways this might be ex- virtually ensures that patients will receive
pressed. In some instances, for example, for more than one diagnosis. In addition to the
dependent personality disorder, the criteria co-occurrence of personality disorders with
are quite behavioral. For dependent person- Axis I disorders, it is also common for pa-
ality disorder, a pervasive and excessive tients to receive more than one personality
need to be taken care of that leads to submis- disorder diagnosis to fully describe their per-
sive and clinging behavior and fears of sepa- sonality problems (Lilienfeld et al. 1994; Old-
ration is indicated by such items as has dif- ham et al. 1992). In the sections that follow,
ficulty making everyday decisions without major patterns of personality disorder co-
an excessive amount of advice and reassur- morbidity will be described.
ance from others and needs others to as-
sume responsibility for most major areas of Co-Occurrence of Personality
his or her life (American Psychiatric Associ- Disorders and Axis I Disorders
ation 2000, p. 725). For other disorders, such
as OCPD, an example of the behavior is There are a number of explanations for the
given along with the trait. For OCPD, perfec- high rates of co-occurrence of personality dis-
tionism is indicated by the following crite- orders and Axis I disorders (Lyons et al.
rion: shows perfectionism that interferes 1997). Co-occurring disorders may share a
with task completion (e.g., is unable to com- common etiology and be different pheno-
plete a project because his or her own overly typic expressions of a common causal factor
strict standards are not met) (American Psy- or factors. They may also be linked by etiol-
chiatric Association 2000, p. 729). ogy or pathological mechanism, but one dis-
order may be a milder version of the other on
a spectrum of severity of pathology or im-
COMORBIDITY pairment. One disorder may precede and in-
crease the risk for the occurrence of another
Since the introduction of a multiaxial system disorder, making a person more vulnera-
for recording diagnoses in DSM-III, which ble to developing the second disorder. A
provided for the diagnosis of personality dis- second disorder may arise after a first as a
orders on an axis (II) separate from the ma- complication or residual phenomenon or
jority of other mental disorders, it has be- scar. People with certain personality disor-
come apparent that most patients with ders and related Axis I disorders may share
personality disorders also meet criteria for common psychobiological substrates that
other disorders. Rates have ranged from regulate cognitive or affective processes or
about two-thirds to almost 100% (Dolan- impulse control. The Axis I disorders may be
Sewell et al. 2001). The co-occurrence of Axis the direct symptomatic expression of dys-
I and Axis II disorders has often been re- functions in these systems, whereas personal-
ferred to as comorbidity, although our current ity disorders may reflect coping mechanisms
understanding of the fundamental nature of and more general personality predispositions
most mental disorders is insufficient to jus- arising from the same systems (Siever and
tify the use of the term according to its formal Davis 1991). This more comprehensive model
definition, which requires that a comorbid of disorder co-occurrence integrates aspects
disorder be distinct from the index disease of the common cause, spectrum, and vulner-
or condition (Feinstein 1970). The DSM sys- ability hypotheses.
tem, with its tendency to split as opposed Axis I/Axis II co-occurrence may be
to lump psychopathology via its many and viewed from the perspectives of the course of
expanding lists of disorders, encourages the a persons lifetime or the current presenting
diagnosis of multiple putative disorders to illness. Lifetime rates will obviously be
Manifestations, Clinical Diagnosis, and Comorbidity 55

higher. Patients with personality disorders is most strongly associated with substance
who are seeking treatment also tend to have use disorders in clinical and general popula-
elevated rates of Axis I disorder co-occur- tion samples (Compton et al. 2007; Grant et
rence, because the development or exacerba- al. 2004; Hasin et al. 2007; Kessler et al. 1997;
tion of an Axis I disorder is often the reason a Morgenstern et al. 1997; see also Chapter 19,
personality disorder patient comes for clini- Substance Abuse). This association sup-
cal attention (Shea 1997). For disorder co- ports an underlying dimension of impulsiv-
occurrence to be significant from a scientific ity or externalization (acting-out and being at
perspective, rates must be elevated above odds with mainstream goals and values)
those expected by chance, based on the rates shared by these disorders (Krueger et al.
of occurrence of the individual disorders in a 1998, 2002, 2007).
given clinical setting or population. From a Cluster C personality disorders, espe-
treatment perspective, any co-occurrence cially avoidant and dependent personality
may be significant. disorders, are linked by theory and phenom-
The personality disorders of Cluster A enology to anxiety disorders (Tyrer et al.
paranoid, schizoid, and schizotypalare 1997). Several studies have demonstrated
linked by theory and phenomenology to Axis high rates of co-occurrence of avoidant per-
I psychotic disorders such as delusional dis- sonality disorder with MDD, agoraphobia,
order, schizophreniform disorder, or schizo- social phobia, and obsessive-compulsive dis-
phrenia. Few studies have actually docu- order (Herbert et al. 1992; Oldham et al. 1995;
mented these associations, however, possibly Skodol et al. 1995). The co-occurrence rates
because of problems in being able to differen- between avoidant personality disorder and
tiate between clinical presentations of attenu- social phobia (particularly the generalized
ated and full-blown psychotic symptoms that type) have been so high in some studies that
warrant two diagnoses instead of just one. investigators have argued that they are the
Oldham et al. (1995) found elevated odds of a same disorder. Several studies have indi-
current psychotic disorder in patients with cated that dependent personality disorder
Cluster A personality disorders but also co-occurs with a wide variety of Axis I disor-
found elevated odds for Clusters B and C per- ders, consistent with the notion of excessive
sonality disorders as well, suggesting less dependency as a nonspecific maladaptive
disorder specificity than might be expected. behavior pattern that may result from coping
In contrast, Cluster B personality disor- with other chronic mental disorders (Skodol
ders, especially BPD, which is linked by the- et al. 1996). OCPD may be specifically linked
ory and phenomenology to Axis I mood and to obsessive-compulsive disorder; however,
impulse control disorders, have repeatedly an association between them has only incon-
been shown to have high rates of co-occur- sistently been found.
ring MDD and other mood disorders, sub- Following the co-occurrence of Axis I and
stance use disorders, and bulimia nervosa Axis II disorders is more than an intellectual
(Oldham et al. 1995; Skodol et al. 1993, 1999; exercise. The presence of an Axis I disorder in
Zanarini et al. 1989, 1998). Taking into ac- a patient with a personality disorder may sug-
count co-occurrence expected by chance gest a more specific treatment approach, ei-
alone, however, neither Oldham et al. (1995) ther with pharmacological agents, psycho-
nor McGlashan et al. (2000) substantiated the therapy, or self-help groups (as in the case of
relationship between BPD and MDD. In ad- substance use disorders), that will favorably
dition, several studies have shown signifi- affect outcome in these patients. Conversely,
cantly elevated rates of anxiety disorders, in- the presence of personality disorder in a pa-
cluding panic disorder and posttraumatic tient with an Axis I disorder often indicates
stress disorder, in patients with BPD (Mc- greater and more widespread levels of impair-
Glashan et al. 2000; Skodol et al. 1995). ASPD ment (Jackson and Burgess 2002; Skodol et al.
56 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

2002), more chronicity (Grilo et al. 2005; Hart criteria for more than one disorder, clinicians
et al. 2001), and an overall poorer response to have found the practice of diagnosing multi-
treatment requiring more intensive and pro- ple disorders conceptually difficult and
longed care (Reich and Vasile 1993; Shea et al. therefore seldom attempt such diagnoses.
1992). Twenty years later, co-occurring Axis I Prior to DSM-III-R (American Psychiatric
and Axis II disorders present a higher risk for Association 1987), part of the problem had
negative prognoses than do disorders on ei- been that most of the personality disorders
ther axis alone (Crawford et al. 2008). were defined as classical categories (Cantor et
al. 1980)that is, ones in which all members
Co-Occurrence of Personality clearly share certain identifying features.
Disorders With Other Classical categories imply a clear demar-
cation between members and nonmembers,
Personality Disorders
but natural phenomena rarely fit neatly into
When thorough assessments of the full range such categories.
of Axis II disorders are conducted, as in re-
search studies employing semistructured in-
terviews, approximately half of patients re- CATEGORICAL VERSUS
ceive more than one personality disorder
diagnosis. Patterns of co-occurrence of per- DIMENSIONAL ASSESSMENTS
sonality disorders generally follow the DSM OF PERSONALITY
cluster structure (i.e, schizotypal personality
disorder occurs more frequently with para- Traditionally, in much of the psychological
noid and schizoid personality disorders than literature, personality has been described and
with personality disorders outside Cluster measured along certain dimensions (Frances
A). These patterns are consistent with factor- 1982). Dimensions of personality frequently
analytic studies that support the clustering of are continuous with opposite traits at either
personality disorders in DSM (Kass et al. end of a spectrum, such as dominant-sub-
1985; Sanislow et al. 2002). Some personality missive or hostile-friendly. People can then
disorders, however, particularly those in vary in the extent to which each of the traits
Cluster C, show associations with personal- describes them. Dimensional models of per-
ity disorders from other clusters. Dependent sonality diagnosis appear to be more flexible
personality disorder commonly occurs in pa- and specific than categorical models when
tients with BPD, which makes clinical sense the phenomenology lacks clear-cut bound-
because patients with BPD can display re- aries between normal and abnormal and be-
gressive, clinging, and dependent behavior tween different constellations of maladaptive
in interpersonal relationships. Some person- traits, as seems true of personality distur-
ality disorders rarely co-occur. OCPD and bance (Widiger et al. 1987). Scaled rating sys-
ASPD would be an exceedingly rare combi- tems have been devised to transform Axis II
nation, because the careful planning and disorders into dimensions (Kass et al. 1985;
work orientation of OCPD are the antithesis Oldham and Skodol 2000), but they are not
of the impulsivity and irresponsibility of representative of dimensional approaches
ASPD. currently in wide use. Dimensional models of
Elevated rates of personality disorder co- personality disorders are being seriously con-
occurrence raise questions about the appro- sidered for DSM-V (see Chapter 20, Future
priate application of DSM-IV-TR categories to Directions: Toward DSM-V.
phenomenology that rarely appears to have Some personality disorder researchers
discrete boundaries. Although DSM-IV-TR advocate a prototype matching approach to
clearly stipulates that for many patients, per- the diagnosis of personality disorders rather
sonality disturbance would frequently meet than the current DSM procedure, which con-
Manifestations, Clinical Diagnosis, and Comorbidity 57

tinues to involve making present/absent American Psychiatric Association: Diagnostic and

judgments about individual criteria (Shedler Statistical Manual of Mental Disorders, 3rd
Edition. Washington, DC, American Psychiat-
and Westen 2004; Westen et al. 2006). They
ric Association, 1980
would replace the diagnostic criteria sets American Psychiatric Association: Diagnostic and
with descriptions of various personality dis- Statistical Manual of Mental Disorders, 3rd
order prototypes in paragraph form and ask Edition, Revised. Washington, DC, American
clinicians to rate the degree of similarity be- Psychiatric Association, 1987
tween the prototypes and the patient under- American Psychiatric Association: Diagnostic and
Statistical Manual of Mental Disorders, 4th
going evaluation. They argue that a proto- Edition, Text Revision. Washington, DC,
type matching approach allows the clinician American Psychiatric Association, 2000
to consider individual criteria in the context Andreoli A, Gressot G, Aapro N, et al: Personality
of the whole personality disorder descrip- disorders as a predictor of outcome. J Personal
tion, such that no single criterion can make Disord 3:307321, 1989
Beck AT, Freeman A, Davis DD, et al: Cognitive
or break the diagnosis. They also argue that
Therapy of Personality Disorders, 2nd Edi-
a prototype matching approach is closer to tion. New York, Guilford, 2003
the way clinicians make personality disorder Bender DS, Skodol AE: Borderline personality as a
diagnoses in actual practice. self-other representational disturbance. J Per-
sonal Disord 21:500517, 2007
Benjamin LS: Interpersonal Diagnosis and Treat-
ment of Personality Disorders, 2nd Edition.
CONCLUSION New York, Guilford, 1996
Bernstein DP, Cohen P, Skodol AE, et al: Child-
Although considerable dissatisfaction has hood antecedents of adolescent personality
disorders. Am J Psychiatry 153:907913, 1996
been expressed over the DSM approach to
Blashfield RK, Herkov MJ: Investigating clinician
these disorders and a major overhaul has adherence to diagnosis by criteria: a replica-
been recommended by many researchers tion of Morey and Ochoa (1989). J Personal
and clinicians in the field (Clark et al. 1997; Disord 10:219228, 1996
Shedler and Westen 2004; Widiger 1991, Bond MP, Vaillant JS: An empirical study of the re-
lationship between diagnosis and defense
1993), the DSM approach remains the official
style. Arch Gen Psychiatry 43:285288, 1986
standard for diagnosing personality disor- Cantor N, Smith EE, French RS, et al: Psychiatric
der psychopathology. Work on DSM-V has diagnosis as prototype categorization. J Ab-
recently begun, but its publication is not an- norm Psychol 89:181193, 1980
ticipated until at least 2012. Therefore, even if Carver CS, Scheier MF, Weintraub JK: Assessing
coping strategies: a theoretically based ap-
a dimensional approach to personality disor-
proach. J Pers Soc Psychol 56:267283, 1989
ders were to replace the categorical approach Casey PR, Tyrer P: Personality disorder and psy-
in DSM-V, these changes would not be im- chiatric illness in general practice. Br J Psychi-
plemented for several years. atry 156:261265, 1990
Clark LA, Livesley WJ, Morey L: Special feature:
personality disorder assessment: the chal-
lenge of construct validity. J Personal Disord
REFERENCES 11:205231, 1997
Compton WM, Thomas YF, Stinson FS, et al: Prev-
Adler DA, Drake RE, Teague GB: Clinicians prac- alence, correlates, disability, and comorbidity
tices in personality assessment: does gender of DSM-IV drug abuse and dependence in the
influence the use of DSM-III Axis II? Compr United States: results from the National Epide-
Psychiatry 31:125133, 1990 miologic Survey on Alcohol and Related Con-
Alarcon RD: Cross-cultural issues, in American ditions. Arch Gen Psychiatry 64:566576, 2007
Psychiatric Publishing Textbook of Personal- Crawford TN, Cohen P, First MB, et al: Comorbid
ity Disorders. Edited by Oldham JM, Skodol Axis I and Axis II disorders in early adoles-
AE, Bender DS. Washington, DC, American cence: outcomes 20 years later. Arch Gen Psy-
Psychiatric Publishing, 2005, pp 561578 chiatry 65:641648, 2008
58 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

Dolan-Sewell RT, Krueger RF, Shea MT: Co-occur- Hill J, Pilkonis P, Morse J, et al: Social domain dys-
rence with syndrome disorders, in Handbook function and disorganization in borderline
of Personality Disorders: Theory, Research, personality disorder. Psychol Med 38:135146,
and Treatment. Edited by Livesley WJ. New 2008
York, Guilford, 2001, pp 84104 Hyler SE: Personality Diagnostic Questionnaire
Drake RE, Vaillant GE: A validity study of Axis II 4 (PDQ-4). New York, New York State Psychi-
of DSM-III. Am J Psychiatry 142:553558, 1985 atric Institute, 1994
Feinstein AR: The pre-therapeutic classification of Hyler SE, Skodol AE, Kellman D, et al: Validity of
comorbidity in chronic disease. J Chronic Dis the Personality Diagnostic Questionnaire:
23:455468, 1970 comparison with two structured interviews.
First M, Gibbon M, Spitzer RL, et al: Users Guide Am J Psychiatry 147:10431048, 1990
for the Structured Clinical Interview for DSM- Hyler SE, Skodol AE, Oldham JM, et al: Validity of
IV Axis II Personality Disorders. Washington, the Personality Diagnostic QuestionnaireRe-
DC, American Psychiatric Press, 1997 vised (PDQ-R): a replication in an outpatient
Folkman S, Lazarus RS: The Ways of Coping sample. Compr Psychiatry 33:7377, 1992
Questionnaire. Palo Alto, CA, Consulting Psy- Jackson HJ, Burgess PM: Personality disorders in
chologists Press, 1988 the community: results from the Australian
Folkman S, Moskowitz JT: Coping: pitfalls and National Survey of Mental Health and Well-
promise. Ann Rev Psychol 55:745774, 2004 Being, part II. Relationships between person-
Frances A: Categorical and dimensional systems ality disorder, Axis I mental disorders, and
of personality diagn osis: a comparison. physical conditions with disability and health
Compr Psychiatry 23:516527, 1982 consultations. Soc Psychiatry Psychiatr Epide-
Grant BF, Stinson FS, Dawson DA, et al: Co-occur- miol 37:251260, 2002
rence of 12-month alcohol and drug use disor- Johnson JG, Cohen P, Skodol AE, et al: Personality
ders and personality disorders in the United disorders in adolescence and risk of major
States: results from the National Epidemio- mental disorders and suicidality during adult-
logic Survey on Alcohol and Related Condi- hood. Arch Gen Psychiatry 56:805811, 1999
tions. Arch Gen Psychiatry 61:361368, 2004 Johnson JG, Cohen P, Kasen S, et al: Age-related
Grilo CM, Sanislow CA, Gunderson JG, et al: Two- change in personality disorder trait levels be-
year stability and change of schizotypal, bor- tween early adolescence and adulthood: a
derline, avoidant, and obsessive-compulsive community-based longitudinal investigation.
personality disorders. J Consult Clin Psychol Acta Psychiatr Scand 102:265275, 2000
72:767775, 2004 Karasu TB, Skodol AE: Fourth axis for DSM-III:
Grilo CM, Sanislow CA, Shea MT: Two-year pro- psychodynamic evaluation. Am J Psychiatry
spective naturalistic study of remission from 137:607610, 1980
major depressive disorder as a function of per- Kasen S, Cohen P, Skodol AE, et al: Influence of
sonality disorder comorbidity. J Consult Clin child and adolescent psychiatric disorders on
Psychol 73:7885, 2005 young adult personality disorder. Am J Psy-
Gunderson JG: Disturbed relationships as a phe- chiatry 156:15291535, 1999
notype for borderline personality disorder. Kasen S, Cohen P, Skodol AE, et al: Childhood de-
Am J Psychiatry 164:16371640, 2007 pression and adult personality disorder: alter-
Hart AB, Craighead WE, Craighead LW: Predict- native pathways of continuity. Arch Gen Psy-
ing recurrence of major depressive disorder in chiatry 58:231236, 2001
young adults: a prospective study. J Abnorm Kass F, Skodol AE, Charles E, et al: Scaled ratings
Psychol 110:633643, 2001 of DSM-III personality disorders. Am J Psychi-
Hasin DS, Stinson FS, Ogburn E, et al: Prevalence, atry 142:627630, 1985
correlates, disability, and comorbidity of DSM- Kessler RC, Crum RM, Warner LA, et al: Lifetime
IV alcohol abuse and dependence in the co-occurrence of DSM-III-R alcohol abuse and
United States: results from the National Epide- dependence with other psychiatric disorders
miologic Survey on Alcohol and Related Con- in the National Comorbidity Survey. Arch Gen
ditions. Arch Gen Psychiatry 64:830842, 2007 Psychiatry 54:313321, 1997
Herbert JD, Hope DA, Bellack AS: Validity of the Kiesler DJ: Contemporary Interpersonal Theory
distinction between generalized social phobia and Research: Personality, Psychopathology,
and avoidant personality disorder. J Abnorm and Psychotherapy. New York, Wiley, 1996
Psychol 101:332339, 1992 Klein DN: Patients versus informants reports of
Herkov MJ, Blashfield RK: Clinicians diagnoses personality disorders in predicting 7-year
of personality disorder: evidence of a hierar- outcome in outpatients with depressive disor-
chical structure. J Pers Assess 65:313321, 1995 ders. Psychol Assess 15: 216222, 2003
Manifestations, Clinical Diagnosis, and Comorbidity 59

Klonsky ED, Oltmanns TF, Turkheimer E: Infor- McGlashan TH, Grilo CM, Sanislow CA, et al:
mant-reports of personality disorder: relation Two-year prevalence and stability of individ-
to self-reports and future directions. Clin Psy- ual DSM-IV criteria for schizotypal, border-
chol Sci Pract 9:300311, 2002 line, avoidant and obsessive-compulsive per-
Kobak KA, Skodol AE, Bender DS: Diagnostic sonality disorders: toward a hybrid model of
measures for adults, in Handbook of Psychiat- Axis II disorders. Am J Psychiatry 162:883
ric Measures, 2nd Edition. Edited by Rush AJ, 889, 2005
First MB, Blacker D. Washington, DC, Ameri- Millon T, Millon C, Davis R: MCMI-III Manual,
can Psychiatric Publishing, 2008, pp 3560 2nd Edition. Minneapolis, MN, National
Krueger RF, Caspi A, Moffitt TE, et al: The struc- Computer Systems, 1997
ture and stability of common mental disorders Moos RH: Coping Responses Inventory. Odessa,
(DSM-III-R): a longitudinal-epidemiological FL, Psychological Assessment Resources, 1993
study. J Abnorm Psychol 106:216227, 1998 Morey LC, Ochoa ES: An investigation of adher-
Krueger RF, Hicks BM, Patrick CJ, et al: Etiologic ence to diagnostic criteria. J Personal Disord
connections among substance dependence, 3:180192, 1989
antisocial behavior, and personality: modeling Morey LC, Alexander GM, Boggs C: Gender, in
the externalizing spectrum. J Abnorm Psychol American Psychiatric Publishing Textbook of
111:411424, 2002 Personality Disorders. Edited by Oldham JM,
Krueger RF, Markon KE, Patrick CJ, et al: Linking Skodol AE, Bender DS. Washington, DC,
antisocial behavior, substance use, and per- A merica n Psychiat ric Publishin g, 2 00 5,
sonality: an integrative quantitative model of pp 541559
the adult externalizing spectrum. J Abnorm Morgenstern J, Langenbucher J, Labouvie E, et al:
Psychol 116:645666, 2007 The comorbidity of alcoholism and personal-
Lazarus RS, Folkman S: Stress, Appraisal, and ity disorders in a clinical population: preva-
Coping. New York, Springer, 1984 lence rates and relation to alcohol typology
Lenzenweger MF: Stability and change in person- variables. J Abnorm Psychol 106:7484, 1997
ality disorder features: the Longitudinal Study Noyes R Jr, Reich J, Christiansen J, et al: Outcome
of Personality Disorders. Arch Gen Psychiatry of panic disorder: relationship to diagnostic
56:10091015, 1999 subtypes and comorbidity. Arch Gen Psychia-
Lewinsohn PM, Rohde P, Seeley JR, et al: Axis II try 47:809818, 1990
psychopathology as a function of Axis I disor- Oldham JM, Skodol AE: Charting the future of
der in childhood and adolescents. J Am Acad Axis II. J Personal Disord 14:1729, 2000
Child Adolesc Psychiatry 36:17521759, 1997 Oldham JM, Skodol AE, Kellman HD, et al: Diag-
Lilienfeld SO, Waldman ID, Israel AC: A critical nosis of DSM-III-R personality disorders by
examination of the use of the term comorbid- two structured interviews: patterns of comor-
ity in psychopathology research. Clin Psy- bidity. Am J Psychiatry 149:213220, 1992
chol Sci Pract 1:7183, 1994 Oldham JM, Skodol AE, Kellman HD, et al: Co-
Livesley WJ: Trait and behavioral prototypes of morbidity of Axis I and Axis II disorders. Am
personality disorder. Am J Psychiatry 143:728 J Psychiatry 152:571578, 1995
732, 1986 Pearlin LI, Schooler C: The structure of coping.
Loranger AW: International Personality Disorder J Health Soc Behav 19:221, 1978
Examination (IPDE). Odessa, FL, Psychologi- Perry JC, Bond M: Defensive functioning, in Amer-
cal Assessment Resources, 1999 ican Psychiatric Publishing Textbook of Person-
Lyons MJ, Tyrer P, Gunderson J, et al: Special fea- ality Disorders. Edited by Oldham JM, Skodol
ture: heuristic models of comorbidity of Axis I AE, Bender DS. Washington, DC, American
and Axis II disorders. J Personal Disord Psychiatric Publishing, 2005, pp 523540
11:260269, 1997 Perry JC, Cooper SH: An empirical study of de-
McDermut W, Zimmerman M: Personality disor- fense mechanisms, I: clinical interview and life
ders, personality traits, and defense mecha- vignette ratings. Arch Gen Psychiatry 46:444
nisms, in Handbook of Psychiatric Measures, 452, 1989
2nd Edition. Edited by Rush AJ, First MB, Pfohl B, Stangl D, Zimmerman M: The implications
Blacker D. American Psychiatric Publishing, of DSM-III personality disorders with major
2008, pp 687729 depression. J Affect Disord 7:309318, 1984
McGlashan TH, Grilo CM, Skodol AE, et al: The Pfohl B, Blum N, Zimmerman M: Structured In-
Collaborative Longitudinal Personality Disor- terview for DSM-IV Personality. Washington,
ders Study: Axis I/II and II/II diagnostic co- DC, American Psychiatric Press, 1997
occurrence. Acta Psychiatr Scand 102:256264, Pope HG Jr, Jonas JM, Hudson JI, et al: The valid-
2000 ity of DSM-III borderline personality disorder:
60 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

a phenomenologic, family history, treatment Skodol AE, Williams JBW, Spitzer RL, et al: Identi-
response, and long-term follow-up study. fying common errors in the use of DSM-III
Arch Gen Psychiatry 40:2330, 1983 through diagnostic supervision. Hosp Com-
Reich J, Vasile RG: Effect of personality disorders munity Psychiatry 35:251255, 1984
on the treatment outcome of Axis I conditions: Skodol AE, Rosnick L, Kellman D, et al: Validating
an update. J Nerv Ment Dis 181:475484, 1993 structured DSM-III-R personality disorder as-
Reich J, Yates W, Nduaguba M: Prevalence of sessments with longitudinal data. Am J Psy-
DSM-III personality disorders in the commu- chiatry 145:12971299, 1988
nity. Soc Psychiatry Psychiatr Epidemiol Skodol AE, Oldham JM, Rosnick L, et al: Diagno-
24:1216, 1989 sis of DSM-III-R personality disorders: a com-
Rey JM, Morris-Yates A, Singh M, et al: Continu- parison of two structured interviews. Int J
ities between psychiatric disorders in adoles- Methods Psychiatr Res 1:1326, 1991
cents and personality disorders in young Skodol AE, Oldham JM, Hyler SE, et al: Comor-
adults. Am J Psychiatry 152:895900, 1995 bidity of DSM-III-R eating disorders and per-
Roberts BW, Caspi A, Moffitt TE: Work experi- sonality disorders. Int J Eating Disorders
ences and personality development in young 14:403416, 1993
adulthood. J Pers Soc Psychol 84:582593, 2003 Skodol AE, Oldham JM, Hyler SE, et al: Patterns of
Sanislow CA, Morey LC, Grilo CM, et al: Confir- anxiety and personality disorder comorbidity.
matory factor analysis of DSM-IV borderline, J Psychiatr Res 29:361374, 1995
schizotypal, avoidant, and obsessive-compul- Skodol AE, Gallaher PE, Oldham JM: Excessive de-
sive personality disorders: findings from the pendency and depression: is the relationship
Collaborative Longitudinal Personality Study. specific? J Nerv Ment Dis 184:165171, 1996
Acta Psychiatr Scand 105:2836, 2002 Skodol AE, Oldham JM, Gallaher PE: Axis II co-
Seivewright H, Tyrer P, Johnson T: Persistent so- morbidity of substance use disorders among
cial dysfunction in anxious and depressed pa- patients referred for treatment of personality
tients with personality disorder. Acta Psychi- disorders. Am J Psychiatry 156:733738, 1999
atr Scand 109:104109, 2004 Skodol AE, Gunderson JG, McGlashan TH, et al:
Shea MT: Assessment of change in personality Function al impairment in patients with
disorders, in Measuring Changes in Mood, schizotypal, borderline, avoidant, or obses-
Anxiety, and Personality Disorders: Toward a sive-compulsive personality disorder. Am J
Core Battery. Edited by Strupp HH, Horowitz Psychiatry 159:276283, 2002
LM, Lambert MJ. Washington, DC, American Skodol AE, Pagano MP, Bender DS, et al: Stability
Psychological Association, 1997, pp 389400 of functional impairment in patients with
Shea MT, Pilkonis PA, Beckham E, et al: Personal- schizotypal, borderline, avoidant, or obses-
ity disorder and treatment outcome in the sive-compulsive personality disorder over
NIMH Treatment of Depression Collaborative two years. Psychol Med 35:443451, 2005
Research Program. Am J Psychiatry 147:711 Soloff PH, Ulrich RF: Diagnostic interview for
718, 1990 borderline patients: a replication study. Arch
Shea MT, Widiger TA, Klein MH: Comorbidity of Gen Psychiatry 38:686692, 1981
personality disorders and depression: impli- Somwaru DP, Ben-Porath YS: Development and
cations for treatment. J Consult Clin Psychol reliability of MMPI-2 based personality disor-
60:857868, 1992 der scales. Paper presented at the 30th annual
Shedler J, Westen D: Refining personality disorder Workshop and Symposium on Recent Devel-
diagnosis: integrating science and practice. opments in the Use of the MMPI-2 and MMPI-
Am J Psychiatry 161:13501365, 2004 A, St. Petersburg, FL, April, 1995
Siever LJ, Davis KL: A psychobiological perspec- Stone AA, Schwartz JE, Neale JM, et al: A compar-
tive on the personality disorders. Am J Psychi- ison of coping assessed by ecological momen-
atry 148:16471658, 1991 tary assessment and retrospective recall. J Pers
Skodol AE: Problems in Differential Diagnosis: Soc Psychol 74:16701680, 1998
From DSM-III to DSM-III-R in Clinical Prac- Torgersen S: Genetic and nosological aspects of
tice. Washington, DC, American Psychiatric schizotypal and borderline personality dis-
Press, 1989 orders: a twin study. Arch Gen Psychiatry
Skodol AE, Gunderson JG: Personality disorders, 41:546554, 1984
in The American Psychiatric Publishing Text- Turner SM, Beidel DC, Borden JW, et al: Social
book of Psychiatry, 5th Edition. Edited by phobia: Axis I and II correlates. J Abnorm Psy-
Hales RE, Yudofsky SC, Gabbard GO. Wash- chol 100:102106, 1991
ington, DC, American Psychiatric Punblish- Tyrer P, Gunderson J, Lyons MJ, et al: Special fea-
ing, 2008, pp 829851 ture: extent of comorbidity between mental
Manifestations, Clinical Diagnosis, and Comorbidity 61

state and personality disorders. J Personal Dis- Disorders. Odessa, FL, Psychological Assess-
ord 11:242259, 1997 ment Resources, 1995
Vaillant GE, Bond M, Vaillant CO: An empirically Young J, Klosko J: Schema therapy, in American
validated hierarchy of defense mechanisms. Psychiatric Publishing Textbook of Personal-
Arch Gen Psychiatry 43:786794, 1986 ity Disorders. Edited by Oldham JM, Skodol
Verheul R, Widiger TA: A meta-analysis of the AE, Bender DS. Washington, DC, American
prevalence and usage of personality disorder Psychiatric Publishing, 2005, pp 289306
not otherwise specified (PDNOS). J Personal Zanarini MC, Gunderson JG, Frankenburg FR: Axis
Disord 18:309319, 2004 I phenomenology of borderline personality dis-
Westen D: Divergences between clinical and re- order. Compr Psychiatry 30:149156, 1989
search methods for assessing personality disor- Zanarini MC, Frankenburg FR, Sickel AE, et al: Di-
ders: implications for research and the evolution agnostic Interview for DSM-IV Personality Dis-
of Axis II. Am J Psychiatry 154:895903, 1997 orders. Belmont, MA, McLean Hospital, 1996
Westen DS, Shedler J, Bradley R: A prototype ap- Zanarini MC, Frankenburg FR, Dubo ED, et al:
proach to personality disorder diagnosis. Am Axis I comorbidity of borderline personality
J Psychiatry 163:846856, 2006 disorder. Am J Psychiatry 155:17331739, 1998
Widiger TA: Personality disorder dimensional Zanarini MC, Frankenburg FR, Hennen J, et al:
models proposed for DSM-V. J Personal Dis- Prediction of the 10-year course of borderline
ord 5:386398, 1991 personality disorder. Am J Psychiatry 163:827
Widiger TA: The DSM-III-R categorical personal- 832, 2006
ity disorder diagnoses: a critique and an alter- Zanarini MC, Frankenburg FR, Reich DB, et al:
native. Psychol Inq 4:7590, 1993 The subsyndromal phenomenology of border-
Widiger TA: Sex biases in the diagnosis of person- line personality disorder: a 10-year follow-up
ality disorders. J Personal Disord 12:95118, study. Am J Psychiatry 164:929935, 2007
1998 Zimmerman M: Diagnosing personality disor-
Widiger TA: Personality disorders, in Handbook ders: a review of issues and research methods.
of Assessment and Treatment Planning for Arch Gen Psychiatry 51:225245, 1994
Psychological Disorders. Edited by Antony Zimmerman M, Coryell W: DSM-III personality
MM, Barlow DH. New York, Guilford, 2002, disorder diagnoses in a nonpatient sample:
pp 453480 demographic correlates and comorbidity.
Widiger TA, Coker LA: Assessing personality dis- Arch Gen Psychiatry 46:682689, 1989
orders, in Clinical Personality Assessment: Zimmerman M, Mattia JI: Differences between
Practical Approaches, 2nd Edition. Edited by clinical and research practices in diagnosing
Butcher JN. New York, Oxford University borderline personality disorder. Am J Psychia-
Press, 2002, pp 407434 try 156:15701574, 1999
Widiger TA, Sanderson CJ: Assessing personality Zimmerman M, Pfohl B, Stangl D, et al: Assess-
disorders, in Clinical Personality Assessment: ment of DSM-III personality disorders: the im-
Practical Approaches. Edited by Butcher JN. portance of interviewing an informant. J Clin
New York, Oxford University Press, 1995, Psychiatry 47:261263, 1986
pp 380394 Zimmerman M, Pfohl B, Coryell W, et al: Diagnos-
Widiger TA, Trull TJ, Hurt SW, et al: A multidimen- ing personality disorder in depressed patients:
sional scaling of the DSM-III personality disor- a comparison of patient and informant inter-
ders. Arch Gen Psychiatry 44:557563, 1987 views. Arch Gen Psychiatry 45:733737, 1988
Widiger TA, Mangine S, Corbitt EM, et al: Person- Zoccolillo M, Pickles A, Quinton D, et al: The out-
ality Disorder InterviewIV: A Semistructured come of conduct disorder: implications for de-
Interview for the Assessment of Personality fining adult personality disorder and conduct
disorder. Psychol Med 22:971986, 1992
This page intentionally left blank
Course and Outcome
Carlos M. Grilo, Ph.D.
Thomas H. McGlashan, M.D.

The diagnostic construct of personality dis- tion systems, DSM-IV-TR and ICD-10 (World
order has evolved considerably over the past Health Organization 1992), although the two
few decades (see Skodol 1997 for a detailed systems differ somewhat in their classification
ontogeny of the DSM system; and see Chapter and definitions for personality disorders and
1, Personality Disorders: Recent History and thus demonstrate only moderate convergence
the DSM System, for a historical overview). for some diagnoses (Ottosson et al. 2002). The
Substantial changes have occurred in both the extent of stability of personality disorders re-
number and types of specific personality dis- mains uncertain (Shea and Yen 2003; Tyrer
order diagnoses over time, as well as in the and Simonsen 2003). This chapter provides an
admixture of criteria (Sanislow and Mc- overview of the course and outcome of per-
Glashan 1998) representing possible manifes- sonality disorders and synthesizes the empiri-
tations of personality disorders (e.g., DSM-IV- cal literature on the stability of personality dis-
TR [American Psychiatric Association 2000] orders.
specifies that the enduring pattern can be
manifested by problems in at least two of the
following areas: cognition, affectivity, in- STABILITY AS THE CENTRAL TENET OF
terpersonal functioning, or impulse control).
One central tenetthat a personality disorder
reflects a persistent, pervasive, enduring, and The concept of stability has remained a cen-
stable patternhas not changed. The concept tral tenet of personality disorders throughout
of stability is salient in both major classifica- the various editions of DSM, dating back to

64 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

the first edition (American Psychiatric Asso- David and Pilkonis 1996; Perry 1993; Ruegg
ciation 1952). In what some experts have re- and Frances 1995; Stone 1993; Zimmerman
ferred to as a bold step (Tyrer and Simon- 1994). These reviews, although varied, have
sen 2003), personality disorders were placed agreed on the pervasiveness of methodolog-
on a separate axis (Axis II) of the multiaxial ical problems that characterize much of the
DSM-III (American Psychiatric Association literature and thereby preclude any firm con-
1980). DSM-III stated that the separation to clusions regarding the nature of the stability
Axis II was intended, in part, to encourage cli- of personality disorders. The reviews, how-
nicians to assess the possible presence of dis- ever, have also generally agreed that avail-
orders that are frequently overlooked when able research raises questions regarding
attention is directed to the usually more florid many aspects of the construct validity of per-
Axis I disorder. Conceptually, this separa- sonality disorders (Zimmerman 1994), in-
tion reflected the putative stability of person- cluding their hypothesized high degree of
ality disorders relative to the episodically un- stability (Grilo and McGlashan 1999).
stable course of Axis I psychiatric disorders The few early (pre-DSM-III era) studies
(Grilo et al. 1998; Skodol 1997). of the course of personality disorders re-
ported findings that borderline (Carpenter
and Gunderson 1977; Grinker et al. 1968) and
antisocial (Maddocks 1970; Robbins et al.
1977) personality disorders were highly sta-
RESEARCH STUDIES ON STABILITY ble. Carpenter and Gunderson (1977), for ex-
ample, reported that the impairment in func-
First, we provide a brief review of the empiri-
tioning observed for borderline personality
cal literature through the end of the twentieth
disorder (BPD) was comparable with that ob-
century. This period can be thought of as in-
served for patients with schizophrenia over a
cluding the first generation (mostly clinical-
5-year period. As previously noted (Grilo et
descriptive accounts) and the second genera-
al. 1998), the dominant clinical approach to
tion (the emerging findings based on attempts
assessing personality disorder diagnoses
at greater standardization of diagnoses and
based partly on treatment refractoriness nat-
assessment methods) of research efforts on
urally raises the question of whether these
personality disorders. Second, we provide a
findings simply reflect a tautology.
brief overview of methodological problems
The separation of personality disorders to
and conceptual gaps that characterize this lit-
Axis II in DSM-III contributed to increased re-
erature and that must be considered when in-
search attention to these clinical problems
terpreting ongoing research and designing fu-
(Blashfield and McElroy 1987). The develop-
ture studies. Third, we summarize emerging
ment and utilization of a number of struc-
findings from ongoing longitudinal studies
tured and standardized approaches to clini-
that have shed light on a number of key issues
cal interviewing and diagnosis during the
about the course of personality disorders.
1980s represented notable advances (Zim-
merman 1994). The greater attention paid to
defining the criteria required for diagnosis in
OVERVIEW OF THE LITERATURE the classification systems and by researchers
THROUGH 1999 during the development of standardized in-
terviews greatly facilitated research efforts in
A number of previous reviews have been this field.
published addressing aspects of the course In our previous reviews of the DSM-III
and outcome of personality disorders (Grilo and DSM-III-R (American Psychiatric Associ-
and McGlashan 1999; Grilo et al. 1998; Mc- ation 1987) studies, we concluded that the
Course and Outcome 65

available research suggested that personality

disorders demonstrate only moderate stabil- CONCEPTUAL AND
ity and that, although personality disorders METHODOLOGICAL QUESTIONS
are generally associated with negative out-
comes, they can improve over time and can
benefit from specific treatments (Grilo and
Previous reviews of personality disorders
McGlashan 1999, p. 157). In our 1998 review
have raised many methodological problems.
(Grilo et al. 1998), we noted that the 20 se-
Common limitations highlighted include
lected studies of DSM-III-R criteria generally
small sample sizes; concerns about nonstand-
found low to moderate stability of any per-
ardized assessments, interrater reliability,
sonality disorder over relatively short follow-
blindness to baseline characteristics, and nar-
up periods (6 to 24 months). For example, the
row assessments; failure to consider alterna-
major studies that employed diagnostic inter-
tive (e.g., dimensional) models of personality
views reported kappa coefficients for the pres-
disorder; reliance on only two assessments
ence of any personality disorder of 0.32
typically over short follow-up periods; insuf-
(Johnson et al. 1997), 0.40 (Ferro et al. 1998),
ficient attention to the nature and effects of
0.50 (Loranger et al. 1994), and 0.55 (Loranger
co-occurring Axis I and Axis II diagnoses;
et al. 1991). Especially noteworthy is that the
and inattention to treatment effects. Diag-
stability coefficients for specific personality
noses other than ASPD and BPD have re-
disorder diagnoses (in the few cases in which
ceived little attention. Particularly striking is
they could be calculated given the sample
the absence of relevant comparison or con-
sizes) were generally lower. In addition, fol-
trol groups in the longitudinal literature. We
low-up studies of adolescents diagnosed with
comment briefly on a few of these issues.
personality disorders also reported modest
stability; for example, Mattanah et al. (1995)
reported a 50% rate of stability for any person-
ality disorder at 2-year follow-up. More re- Reliability of assessments represents a central
cently, Grilo et al. (2001) also found modest issue for any study of course and outcome.
stability in dimensional personality disorder The creation of standardized instruments for
scores in this adolescent follow-up study. collecting data was a major development of
Squires-Wheeler et al. (1992), as part of the the 1980s (Loranger et al. 1991; Zimmerman
New York State high-risk offspring study, re- 1994). Such instruments, however, were less-
ported low stability for schizotypal personal- than-perfect assessment methods and have
ity disorder and features, although the stabil- been criticized for a variety of reasons (Wes-
ity was higher for the offspring of patients ten 1997; Westen and Shedler 1999). It is crit-
with schizophrenia than for those with mood ical to keep in mind that interrater reliability
disorders or control subjects. and testretest reliability represent the limits
Subsequently, we (Grilo and McGlashan (or ceiling) for estimating the stability of a
1999) reviewed nine reports of longitudinal construct.
findings for personality disorder diagnoses Previous reviews (Grilo and McGlashan
published in 1997 and 1998. In terms of spe- 1999; Zanarini et al. 2000; Zimmerman 1994)
cific diagnoses, the studies generally reported of reliabilities for Axis II diagnostic inter-
moderate stability (kappa approximately 0.5) views have generally reported median inter-
for BPD and antisocial personality disorder rater reliabilities of roughly 0.70 and short-
(ASPD). These reports, like most of the previ- interval, testretest reliabilities of 0.50 for
ous literature, had small sample sizes and in- diagnoses. These reliabilities compare favor-
frequently followed more than one personal- ably with those generally reported for diag-
ity disorder. nostic instruments for Axis I psychiatric dis-
66 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

orders. Both int errat er and testre test unrelated to state-trait effects, reflects
reliability coefficients tend to be higher for some combination of the following: regres-
dimensional scores than for categorical diag- sion to the mean, error in either or both the
noses of personality disorders. Another find- baseline and repeated assessments, and
ing of note is that even when experts admin- overreporting by patients at hospital admis-
ister diagnostic interviews, the degree of sion and underreporting during retest at dis-
convergence or agreement produced by two charge (Loranger et al. 1991; Shea and Yen
different interviews administered only 2003). These phenomena were discussed fur-
1 week apart is limited (Oldham et al. 1992). ther by Gunderson et al. (2000).

Reliability and Change Categorical Versus

Dimensional Approaches
Testretest reliability is also relevant for ad-
dressing, in part, the well-known problem of Long-standing debate regarding the concep-
regression to the mean in repeated mea- tual and empirical advantages to dimen-
sures studies (Nesselroade et al. 1980). It has sional models of personality disorders
been argued that the multiwave or repeated (Frances 1982; Livesley et al. 1992; Loranger
measures approach lessens the effects of re- et al. 1994; Widiger 1992) has accompanied
gression to the mean (Lenzenweger 1999). the DSM categorical classification system.
This argument may be true in terms of the Overall, longitudinal studies of personality
obvious decreases in severity with time (i.e., disorder have reported moderate levels of
very symptomatic participants meeting eligi- stability for dimensional scores for most per-
bility at study entry are likely to show some sonality disorders, with the stability coeffi-
improvement because, by definition, they cients tending to be higher than for categori-
are already reporting high levels of symp- cal or diagnostic stability (Ferro et al. 1998;
toms). However, other effects need to be con- Johnson et al. 1997; Klein and Shih 1998; Lo-
sidered whenever assessments are repeated ranger et al. 1991, 1994).
within a study. As cogently noted by Shea
and Yen (2003), repeated measures studies of Comorbidity
both Axis II (Loranger et al. 1991) and Axis I
(Robins 1985) disorders have found hints Most studies have ascertained participants
that participants systematically report or en- who meet criteria for multiple Axis I and Axis
dorse fewer problems during repeated inter- II diagnoses. This problem of diagnostic over-
views to reduce interview time. For example, lap, or comorbidity, represents a well-known,
Loranger et al. (1991), in his testretest study long-standing major challenge (Berkson 1946)
of the Personality Disorder Examination in- in working with clinical samples. One expert
terview (Loranger 1988) conducted between and critic of DSM (Tyrer 2001), in speaking of
1 and 26 weeks after baseline, documented the spectre of comorbidity, noted that the
significant decreases in personality disorder main reason for abandoning the present clas-
criteria for all but two of the DSM-III-R diag- sification is summed up in one word, comor-
noses. Recall that the Personality Disorder bidity. Comorbidity is the nosologists night-
Examination, which requires skilled and mare; it shouts, you have failed (p. 82). We
trained research clinicians, has a required suggest, however, that such clinical realities
minimum duration stipulation of 5 years for (multiple presenting problems that are espe-
determining persistence and pervasiveness cially characteristic of treatment-seeking pa-
of the criteria being assessed. Thus, the mag- tients) represent not only potential confounds
nitude of changes observed during such a but also potential opportunities to under-
short period of time, which was shown to be stand personality and dysfunctions of person-
Course and Outcome 67

ality better. Comorbidity begs the question: during a 2-year follow-up, even after control-
what are the fundamental personality dimen- ling for previous psychiatric history (John-
sions and disorders of personality, and how son et al. 1997).
do their courses influence (and conversely,
how are their courses affected by) the pres-
Comorbidity and Continuity Models
ence and course of Axis I psychiatric disor-
ders? A variation of the comorbidity concept is that
certain disorders may be associated with one
Continuity another in a number of possible ways over
time. A variety of models have been pro-
A related issue pertaining to course concerns posed for the possible relationships between
longitudinal comorbidities (Kendell and Axis II and Axis I disorders (Dolan-Sewell et
Clarkin 1992) or continuities. An obvious al. 2001; Lyons et al. 1997; Tyrer et al. 1997).
example is that conduct disorder during ado- These include, for example, the predisposi-
lescence is required for the diagnosis of tion or vulnerability model, the complication
ASPD to be given to adults. This definitional or scar model, the pathoplasty or exacerba-
isomorphism is one likely reason for the con- tion model, and various spectrum models.
sistently strong associations between conduct We emphasize that these models do not nec-
disorder and later ASPD in the literature. essarily assume categorical entities. Indeed,
This association is, however, more than an ar- an especially influential spectrum model
tifactual relationship, because longitudinal proposed by Siever and Davis (1991) posits
research has clearly documented that chil- four psychobiological dimensions to account
dren and adolescents with behavior disor- for Axis II and Axis I psychopathology. The
ders have substantially elevated risk for anti- Cloninger et al. (1993) psychobiological
social behavior during adulthood (Robins model of temperament and character repre-
1966). More generally, studies with diverse sents another valuable approach that consid-
recruitment and ascertainment methods re- ers dimensions across personality and psy-
ported that disruptive behavior disorders chopathology. More broadly, Krueger noted
during the adolescent years prospectively that although most research has focused on
predicted personality disorders during pairs of constructs (i.e., Axis II and Axis I as-
young adulthood (Bernstein et al. 1996; sociations) (Krueger 1999; Krueger and Tack-
Lewinsohn et al. 1997; Myers et al. 1998; Rey ett 2003), it seems important to examine the
et al. 1995). The Yale Psychiatric Institute fol- multivariate structure of the personality-
low-up study found that personality disorder psychopathology domain (Krueger and
diagnoses in adolescent inpatients prospec- Tackett 2003, p. 109).
tively predicted greater drug use problems
but not global functioning (Levy et al. 1999). Age (Early Onset)
The importance of considering comor-
bidity is underscored in the findings of the A related point, stressed by Widiger (2003),
longitudinal study by Lewinsohn et al. is that personality disorders need to be more
(1997). They found that the apparent longitu- clearly conceptualized and carefully charac-
dinal continuity noted for disruptive behav- terized as having an early onset. However,
ioral disorders during adolescence and sub- the validity of personality disorders in ado-
sequent ASPD in adulthood was accounted lescents remains controversial (Krueger and
for, in part, by Axis I psychiatric comorbidity. Carlson 2001). It can be argued, for example,
A longitudinal study of young adult men that determining early onset of personality
found that personality disorders predicted disorders is impossible because adolescence
the subsequent onset of psychiatric disorders is a period of profound changes and flux in
68 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

personality and identity. A critical review of Age and the Aging Process
the longitudinal literature on personality
traits throughout the life span revealed that Another age issue concerns the aging process
personality traits are less stable during child- itself. Considerable research suggests that
hood and adolescence than they are through- personality remains relatively stable thor-
out adulthood (Roberts and DelVecchio ough adulthood (Heatherton and Wein-
2000). Roberts and DelVecchios (2000) meta- berger 1994; Roberts and DelVecchio 2000)
analysis of data from 152 longitudinal stud- and is highly stable after age 50 (Roberts and
ies of personality traits revealed that rank- DelVecchio 2000). Little is known, however,
order consistency for personality traits in- about personality disorders in older persons
creased steadily throughout the life span; (Abrams et al. 1998). The 12-year follow-up of
testretest correlations (over 6.7-year time in- personality disorders that was part of the
tervals) increased from 0.31 (during child- Nottingham Study of Neurotic Disorder
hood) to 0.54 (during college), to 0.64 (age 30 (Seivewright et al. 2002) documented sub-
years), to a high of 0.74 (ages 5070 years). stantial changes in personality disorder trait
Nonetheless, if childhood precursors of scores based on blind administration of a
personality disorders could be identified (as semistructured interview. Seivewright et al.
in the case of conduct disorder for ASPD), (2002) reported that Cluster B personality dis-
they could become part of the diagnostic cri- order diagnoses (ASPD, histrionic) showed
teria and thus create some degree of longitu- significant improvements, whereas Cluster A
dinal continuity in the diagnostic system. and Cluster C diagnoses appeared to worsen
Myers et al. (1998), for example, found that with age. Although the Seivewright et al.
early onset (before 10 years of age) of con- (2002) findings are limited somewhat by the
duct disorder problems predicted subse- two-point cross-sectional assessment (little is
quent ASPD. More generally, temperamental known about the intervening period), Tyrer
vulnerabilities or precursors to personality et al. (1983) previously reported good reliabil-
disorders have been posited as central in a ity (weighted kappa of 0.64) for this diagnos-
variety of models of personality disorders tic interview over a 3-year testretest period.
(Cloninger et al. 1993; Siever and Davis These findings echo somewhat the results of
1991). Specific temperamental features evi- the seminal Chestnut Lodge follow-up stud-
dent in childhood have been noted to be pre- ies (McGlashan 1986a, 1986b) that suggested
cursors for diverse personality disorders distinctions between BPD and schizotypal
(Paris 2003; Rettew et al. 2003; Wolff et al. personality disorders, decreases in impulsiv-
1991) as well as for differences in interper- ity and interpersonal instability with age, and
sonal functioning (Newman et al. 1997) in increased avoidance with age. There are
adulthood. For example, studies have noted other reports of diminished impulsivity with
early od d and w ithd raw n patterns for increasing age in BPD (Paris and Zweig-
schizotypal personality disorder in adults Frank 2001; Stevenson et al. 2003), although
(Wolff et al. 1991) and shyness for avoidant this type of reduction was not observed in a
personality disorder (Rettew et al. 2003). recent prospective analysis of individual BPD
Speaking more generally, although the criteria (McGlashan et al. 2005).
degree of stability for personality traits is The reader is referred to Judd and Mc-
higher throughout adulthood than through- Glashan (2003) for detailed accounts of four
out childhood and adolescence (Roberts and specific cases that elucidate the course and
DelVecchio 2000), longitudinal analyses of outcome of BPD. These detailed case studies,
personality data have revealed that the tran- based on rich clinical material available
sition from adolescence to adulthood is through the Chestnut Lodge study, demon-
characterized by greater personality conti- strate the considerable heterogeneity in the
nuity than change (Roberts et al. 2001). course of BPD.
Course and Outcome 69

Summary and Implications of the conceptual and methodological issues

noted earlier. These studies utilized multiple
To resolve these complex issues, comple- and standardized assessment methods, care-
mentary research efforts are required, with fully considered training and reliability,
large samples of both clinical and commu- andperhaps most notablymultiwave re-
nity populations. It is clear that prospective peated assessments that are essential for de-
longitudinal studies with repeated assess- termining longitudinal change. They have
ments over time are needed to understand employed, to varying degrees, multiple as-
the course of personality disorders. Such sessment methods and have considered per-
studies must consider (and cut across) differ- sonality and its disorders (personality disor-
ent developmental eras, broad domains of ders) as well as Axis I psychiatric disorders.
functioning, and multimodal approaches to Collectively, these studies have provided
personality and disorders of personality. valuable insights into the complexities of
These approaches have, in fact, been per- personality (traits and disorders) and its vi-
formed with personality traits (Roberts et al. cissitudes over time.
2001) and with other forms of psychiatric
problems and have yielded invaluable in-
sights. Notable are the contributions of the
Longitudinal Study of
National Institutes of Health (NIH)funded Personality Disorders
multisite efforts on depression (Collabora-
The Longitudinal Study of Personality Disor-
tive Depression Study; Katz et al. 1979) and
ders (Lenzenweger 1999; Lenzenweger et al.
anxiety (Harvard/Brown Anxiety Research
1997) assessed 250 participants drawn from
Project; Keller 1991).
Cornell University at three points over a 4-
year period. It utilized a semistructured di-
agnostic interview (International Personality
REVIEW OF RECENT EMPIRICAL Disorder Examination; Loranger et al. 1994)
ADVANCES AND UNDERSTANDING and a self-report measure (Millon Clinical
OF STABILITY Multiaxial InventoryII; Millon 1987) to ob-
tain complementary information on person-
Of particular relevance for this review are ality. Of the 250 participants, 129 met criteria
three prospective studies on the longitudinal for at least one personality disorder and 121
course of adult personality disorders funded did not meet any personality disorder diag-
by the NIH during the 1990s. These studies nosis. Dimensional scores for the personality
included the Longitudinal Study of Person- disorders were characterized by significant
ality Disorders (Lenzenweger 1999), the levels of stability on both the interview and
McLean Study of Adult Development (Za- self-report measures. Stability coefficients
narini et al. 2003), and the multisite Collabo- for the total number of personality disorder
rative Longitudinal Personality Disorders features ranged from 0.61 to 0.70. Cluster B
Study (CLPS; Gunderson et al. 2000). The personality disorders had the highest stabil-
NIH also funded a community-based pro- ity coefficients, and Cluster A personality
spective longitudinal study of personality, disorders had the lowest. Personality disor-
psychopathology, and functioning of chil- der dimensions showed significant declines
dren/adolescents and their mothers (Chil- over time, and the decline was more rapid
dren in the Community study; Brook et al. for the personality disorder group than for
2002) that began in 1975. These four studies the nonpersonality disorder group. Axis I
are especially noteworthy in that they, to psychiatric disorders (diagnosed in 63% of
varying degrees, partly correct for a number personality disorder subjects and 26% of
70 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

nonpersonality disorder subjects) did not by year 4, and 74% by year 6. Recurrences
significantly influence changes in personal- were rare and were reported for only 6% of
ity disorder dimensions over time. those patients who achieved a remission. The
The Longitudinal Study of Personality authors concluded that symptomatic im-
Disorders BPD findings are generally consis- provement is both common and stable, even
tent (although the three-point assessment is among the most disturbed borderline pa-
an important incremental contribution) with tients, and that the symptomatic prognosis
those previously reported by Trull et al. for most, but not all, severely ill borderline
(1997, 1998) in a prospective study of BPD patients is better than previously recog-
features using two different assessment in- nized (Zanarini et al. 2003; p. 274). Zanarini
struments administered to a college student et al. (2004) reported that Axis I psychiatric
sample assessed twice over a 2-year period. disorders are less common over time in pa-
The Longitudinal Study of Personality tients with BPD and particularly among
Disorders (Lenzenweger 1999), however, is those who remit from BPD.
limited by its relatively homogeneous study
group of college students, its narrow devel- Collaborative Longitudinal
opmental time frame, and most importantly Personality Disorders Study
the insufficient frequency of any personality
disorder diagnosis at a categorical (diagnos- The CLPS (Gunderson et al. 2000; McGlashan
tic) level to allow analysis of a clinical entity. et al. 2000; Skodol et al. 2005b) is an ongoing
Lenzenweger (1999) noted the need for re- prospective, longitudinal, repeated measures
peated measures longitudinal data from clin- study designed to examine the course and
ically based personality disorder samples to outcome of patients meeting DSM-IV (Amer-
address the question of the course and stabil- ican Psychiatric Association 1994) criteria for
ity of dysfunctions of personality. one of four personality disorders: schizo-
typal, borderline, avoidant, and obsessive-
McLean Study of compulsive. The CLPS includes a compari-
Adult Development son group of patients with major depressive
disorder (MDD) without any personality dis-
The McLean Study of Adult Development order. This comparison group was selected
(Zanarini et al. 2003, 2005) is an ongoing pro- because of its episodic and fluctuating course
spective, longitudinal study comparing the (thought to distinguish Axis I from Axis II)
course and outcome of hospitalized patients and because MDD has been carefully studied
with BPD with those of patients with other in similar longitudinal designs (e.g., Collabo-
personality disorders. It utilizes repeated as- rative Depression Study; [Katz et al. 1979; So-
sessments performed every 2 years (Zanarini lomon et al. 1997]). The CLPS has employed
et al. 2003) and has reported outcomes multimodal assessments (Gunderson et al.
through 6 years (Zanarini et al. 2003) and 10 2000; Zanarini et al. 2000) to prospectively
years (Zanarini et al. 2006) of follow-up. Za- follow and capture different aspects of the
narini et al. (2003) assessed personality disor- fluctuating nature of personality disorders
ders in 362 inpatients (290 with BPD and 72 and dimensions (both interviewer-based and
with other personality disorders) using two self-report representing different conceptual
semistructured diagnostic interviews and models) (Morey et al. 2007), Axis I psychiatric
administered assessments to characterize disorders and symptoms (Grilo et al. 2005,
Axis I psychiatric disorders, psychosocial 2007), various domains of psychosocial func-
functioning domains, and treatment utiliza- tioning (Markowitz et al. 2007; Skodol et al.
tion. Of the patients diagnosed with BPD, re- 2005a, 2005d), and treatment utilization
mission was observed for 35% by year 2, 49% (Bender et al. 2007).
Course and Outcome 71

To date, the CLPS has reported on differ- can be done given the parallel assessment
ent concepts of categorical and dimensional instrumentation) reveals that personality
stability of four personality disorders over 12 disorders demonstrate greater stability than
months (Shea et al. 2002) and 24 months Axis I mood and anxiety disorders (as hy-
(Grilo et al. 2004) using prospective data ob- pothesized) but show less diagnostic (cate-
tained for 668 patients recruited from diverse gorical) stability than conceptualized. Per-
settings at four universities. Based on the tra- haps noteworthy is that the longitudinal
ditional testretest approach, blind repeated studies for both mood and anxiety disorders
administration of a semistructured interview documented much greater chronicity (much
conducted 24 months after baseline revealed lower remission rates) than previously
remission rates (based solely on falling be- known.
low DSM-IV diagnostic thresholds) ranging Returning to the CLPS findings (Grilo et
from 50% (avoidant personality disorder) to al. 2004), Figure 41 reveals that although
61% (schizotypal personality disorder). Grilo personality disorders were more stable than
et al. (2004) applied lifetable survival analy- MDD, a substantial number of remissions
ses to prospective data obtained using an as- occurred during the 24 months of follow-up.
sessment methodology modeled after the Using the arbitrarily selected 2-month defini-
Collaborative Depression Study (Keller et al. tion (2 months with two or fewer criteria)
1982) and the Longitudinal Interval Follow- adopted from the MDD field (Keller et al.
Up Evaluation (Keller et al. 1987) methodol- 1982; Solomon et al. 1997), remission rates
ogy. These findings are summarized in Fig- range from 33% (schizotypal personality dis-
ures 41 and 42. order) to 55% (obsessive-compulsive per-
Figure 41 shows the times to remission sonality disorder). Figure 42 shows the
for the four personality disorder groups and comparable remission rates if a very strin-
for the MDD comparison group, which were gent definition of 12 consecutive months
calculated based on parallel definitions of with two or fewer criteria is adopted. As can
two con secutiv e months with minimal be seen, the remission rates using the 12-
symptoms (Grilo et al. 2004). As can be seen, month definition range from 23% (schizo-
the MDD group had a significantly higher re- typal personality disorder) to 38% (obses-
mission rate than the personality disorder sive-compulsive personality disorder). Grilo
groups. This study represents the first empir- et al. (2004) concluded that these four per-
ical demonstration of the central tenet that sonality disorders show substantial im-
personality disorders are characterized by provements in symptomatology over a 2-
greater degree of stability than the hypothe- year period even when a stringent definition
sized episodic course of Axis I psychiatric is used.
disorders (Grilo et al. 1998; Shea and Yen The CLPS also provided complementary
2003). analyses using dimensional approaches for
The reader is referred to Shea and Yen 12-month (Shea et al. 2002) and 24-month
(2003) for a broader discussion of this issue. (Grilo et al. 2004) follow-ups. Grilo et al.
These researchers, who have played roles in (2004) documented a significant decrease in
the CLPS as well as the longitudinal studies the mean proportion of criteria met in each of
of depression (Collaborative Depression the personality disorder groups over time,
Study) and anxiety (Harvard/Brown Anxi- which is suggestive of decreased severity.
ety Research Project), provide an overview of However, when the relative stability of indi-
the central findings that pertain to the issue vidual differences was examined across the
of stability as a distinction between Axis II multiwave assessments (baseline and 6-, 12-,
and Axis I diagnoses (Shea and Yen 2003). and 24-month time points), a high level of
Briefly, comparison across the studies (which consistency was observed as evidenced by
E s s e n t i a l s o f P e r so n a l i t y D i s o r d e r s
Proportion not remitted

Time from intake (months)

Figure 41. Time to remission for four personality disorder groups and major depressive disorder (MDD) comparison group using a 2-month
criterion in Collaborative Longitudinal Personality Study.
AVPD=avoidant personality disorder; BPD =borderline personality disorder; OCPD= obsessive-compulsive personality disorder; STPD= schizotypal personality
Source. From Grilo CM, Shea MT, Sanislow CA, et al: Two-Year Stability and Change in Schizotypal, Borderline, Avoidant, and Obsessive-Compulsive Personality
Disorders. Journal of Consulting and Clinical Psychology 72:767775, 2004. Reprinted with permission.
Course and Outcome 73

correlation coefficients ranging from 0.53 to with certain psychiatric disorders (MDD and
0.67 for proportion of criteria met between posttraumatic stress disorder), whereas
baseline and 24 months. Grilo et al. (2004) avoidant personality disorder was signifi-
concluded that patients with personality dis- cantly associated with two anxiety disorders
order are consistent in terms of their rank or- (social phobia and obsessive-compulsive
der of personality disorder criteria (i.e., that disorder). While these findings were consis-
individual differences in personality disor- tent with predictions based on the Siever and
der features are stable), although they may Davis (1991) model, other personality disor-
fluctuate in the severity or number of person- ders (schizotypal and obsessive-compulsive)
ality disorder features over time. It is worth did not demonstrate significant longitudinal
noting that the range of the stability coeffi- associations. Gunderson et al. (2004) fol-
cients was quite similar to that documented lowed up on the Shea et al. (2004) findings re-
by the Longitudinal Study of Personality garding changes in BPD and MDD by per-
Disorders (Lenzenweger 1999) for a nonclin- forming a more fine-grained analysis of
ical sample. specific changes in the two disorders using
In contrast to their symptomatic im- 3 years of longitudinal data. Changes (im-
provement, however, patients with personal- provements) in BPD severity preceded im-
ity disorders show less significant and more provements in MDD but not vice versa (Gun-
gradual improvement in their functioning, derson et al. 2004).
particularly in social relationships (Skodol et Another report (Warner et al. 2004) exam-
al. 2005c). In addition, depressed patients ined whether personality traits are stable in
with personality disorders show longer time patients with personality disorders and
to remission from MDD (Grilo et al. 2005) tested the hypothesis that the stability of
and have greater risks for recurrences and for these personality disorders is due in part to
new onsets of MDD (Gunderson et al. 2008). the stability in these traits (Lynam and Widi-
Personality disorders, however, were unre- ger 2001). A series of latent longitudinal
lated to the course of some Axis I psychiatric models tests whether changes in specific
disorders, such as eating disorders (Grilo et traits prospectively predicted changes in rel-
al. 2007). Because personality psychopathol- evant personality disorders. Warner et al.
ogy usually begins in adolescence or early (2004) documented significant cross-lagged
adulthood, the potential for delays in occu- relationships between changes in specific
pational and interpersonal development is traits and subsequent (later) changes for
greatand even after symptomatic im- schizotypal, borderline, and avoidant per-
provement, it might take time to overcome sonality disorders but not for obsessive-
deficits and make up the necessary ground to compulsive personality disorder.
achieve normal functioning. Developmen- McGlashan et al. (2005) examined the in-
tal issues for patients with personality disor- dividual criteria for schizotypal, borderline,
ders are discussed in more detail in Chapter avoidant, and obsessive-compulsive person-
7, Developmental Issues. ality disorders and how they changed over a
Several reports from the CLPS are also 2-year period. The individual criteria for
relevant here given the issue of longitudinal these four personality disorders showed var-
comorbidities and continuities. Shea et al. ied patterns of stability and change over
(2004) examined the time-varying (longitu- time. Overall, within personality disorders,
dinal) associations between personality dis- the relatively fixed (least changeable) criteria
orders and psychiatric disorders, in part were generally more traitlike (and attitudi-
guided by the Siever and Davis (1991) cross- nal), whereas the more fluctuating criteria
cutting psychobiological dimension model. were generally behavioral (or reactive). Mc-
BPD demonstrated significant associations Glashan et al. (2005) posited that perhaps
E s s e n t i a l s o f P e r so n a l i t y D i s o r d e r s
Proportion not remitted

Time from intake (months)

Figure 42. Time to remission for four personality disorder groups using a 12-month criterion in Collaborative Longitudinal Personality Study.
AVPD=avoidant personality disorder; BPD =borderline personality disorder; OCPD= obsessive-compulsive personality disorder; STPD= schizotypal personality
Source. From Grilo CM, Shea MT, Sanislow CA, et al: Two-Year Stability and Change in Schizotypal, Borderline, Avoidant, and Obsessive-Compulsive Personality
Disorders. Journal of Consulting and Clinical Psychology 72:767775, 2004. Reprinted with permission.
Course and Outcome 75

personality disorders are hybrids of traits bances predict personality disorders in ado-
and symptomatic behaviors and that it is the lescents and that personality disorders during
interaction of these over time that helps to adolescence, in addition to demonstrating sig-
define the observable diagnostic stability. nificant levels of continuity into adulthood,
Collectively, along with the recent CLPS ef- also predict psychiatric disorders, suicidality,
forts, these findings suggest that personality and violent and criminal behavior during
disorder traits are stable over time and across young adulthood (Crawford et al. 2008;
developmental eras and may generate intra- Johnson et al. 2000a, 2000b, 2005). Collec-
and interpersonal conflicts that result in be- tively, these findings support the continuity
haviors symptomatic of personality disor- and persistence of personality disturbances,
ders (which are less stable over time). although their developmental pathways are
not yet understood.
Children in the Community Study
The Children in the Community study (Brook CONCLUSION
et al. 1995, 2002; Cohen et al. 2005) is an espe-
cially impressive longitudinal effort that has We have reviewed the literature regarding
already provided a wealth of information the course and stability of personality disor-
about the course of personality and behav- ders. We once again conclude that personal-
ioral traits, psychiatric problems, substance ity disorders demonstrate only moderate sta-
abuse, and adversities. It is an ongoing pro- bility and that they can improve over time.
spective study of nearly 1,000 families with This conclusion is offered with less caution
children ages 110 years originally recruited than during our previous reviews (Grilo and
in 1975 in New York State using a random McGlashan 1999), given some notable ad-
sampling procedure. The study has em- vances in research. We also conclude that
ployed repeated multimodal assessments when personality disorders are considered
and has followed over 700 participants since dimensionally, the degree of stability is sub-
childhood and through the development eras stantial. Emerging work has suggested that
of childhood, adolescence, and early adult- personality disorder traits, although deviant,
hood. are stable over time and across developmen-
This landmark study, which has reported tal eras and may generate intra- and interper-
20-year outcomes (Crawford et al. 2008), has sonal conflicts that result in personality dis-
provided data that speak to the critical issues ordersymptomatic behaviors (which are
of longitudinal comorbidities and continu- less stable over time). Future research in per-
ities. In a series of papers, the collaborating sonality disorders is necessary to dissect and
researchers have documented important understand this trait-state interaction and
findings relevant to the issues raised in this track its vicissitudes across time and circum-
review but especially to the critical issues of stances.
continuity of risk and functioning across
developmental eras. These include docu-
mentation of the validity of certain forms of
dramatic-erratic personality disorders in ado- REFERENCES
lescents (Crawford et al. 2001a, 2001b); find-
ings of age-related changes in personality dis- Abrams RC, Spielman LA, Alexopoulos GS, et
order traits, including their moderate levels of al: Personality disorder symptoms and func-
tioning in elderly depressed patients. Am J
stability throughout adolescence and early
Geriatr Psychiatry 6:2430, 1998
adulthood (Johnson et al. 2000b); and indica- American Psychiatric Association: Diagnostic and
tions that early forms of behavioral distur- Statistical Manual of Mental Disorders. Wash-
76 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

ington, DC, American Psychiatric Association, mental pathways from early adolescence to
1952 adulthood. J Personal Disord 15:336350, 2001b
American Psychiatric Association: Diagnostic and Crawford TN, Cohen P, First MB, et al: Comorbid
Statistical Manual of Mental Disorders, 3rd Axis I and Axis II disorders in early adoles-
Edition. Washington, DC, American Psychiat- cence: outcomes 20 years later. Arch Gen Psy-
ric Association, 1980 chiatry 65:641648, 2008
American Psychiatric Association: Diagnostic and Dolan-Sewell RT, Krueger RF, Shea MT: Co-occur-
Statistical Manual of Mental Disorders, 3rd Edi- rence with syndrome disorders, in Handbook
tion, Revised. Washington, DC, American Psy- of Personality Disorders. Edited by Livesley
chiatric Association, 1987 WJ. New York, Guilford, 2001, pp 84104
American Psychiatric Association: Diagnostic and Ferro T, Klein DN, Schwartz JE, et al: 30-month
Statistical Manual of Mental Disorders, 4th stability of personality disorder diagnoses in
Edition. Washington, DC, American Psychiat- depressed outpatients. Am J Psychia try
ric Association, 1994 155:653659, 1998
American Psychiatric Association: Diagnostic and Frances A: Categorical and dimensional systems of
Statistical Manual of Mental Disorders, 4th personality diagnosis: a comparison. Compr
Edition, Text Revision. Washington, DC, Psychiatry 23:516527, 1982
American Psychiatric Association, 2000 Grilo CM, McGlashan TH: Stability and course of
Bender DS, Skodol AE, Dyck IR, et al: Ethnicity personality disorders. Curr Opin Psychiatry
and mental health treatment utilization by pa- 12:157162, 1999
tients with personality disorders. J Consult Grilo CM, McGlashan TH, Oldham JM: Course and
Clin Psychol 75:992999, 2007 stability of personality disorders. J Pract Psy-
Berkson J: Limitations of the application of four- chiatry Behav Health 4:6175, 1998
fold table analysis to hospital data. Biometrics Grilo CM, Becker DF, Edell WS, et al: Stability and
Bulletin 2:4753, 1946 change of personality disorder dimensions in
Bernstein DP, Cohen P, Skodol AE, et al: Child- adolescents followed up two years after psy-
hood antecedents of adolescent personality chiatric hospitalization. Compr Psychiatry
disorders. Am J Psychiatry 153:907913, 1996 42:364368, 2001
Blashfield RK, McElroy RA: The 1985 journal liter- Grilo CM, Sanislow CA, Gunderson JG, et al: Two-
ature on the personality disorders. Compr year stability and change of schizotypal, bor-
Psychiatry 28:536546, 1987 derline, avoidant, and obsessive-compulsive
Brook DW, Brook JS, Zhang C, et al: Drug use and personality disorders. J Consult Clin Psychol
the risk of major depressive disorder, alcohol 72:767775, 2004
dependence, and substance use disorders. Grilo CM, Sanislow CA, Shea MT, et al: Two-year
Arch Gen Psychiatry 59:10391044, 2002 prospective naturalistic study of remission
Brook JS, Whiteman M, Cohen P, et al: Longitudi- from major depressive disorder as a function
nally predicting late adolescent and young of personality disorder comorbidity. J Consult
adult drug use: childhood and adolescent pre- Clin Psychol 73:7885, 2005
cursors. J Am Acad Child Adolesc Psychiatry Grilo CM, Pagano ME, Skodol AE, et al: Natural
34:12301238, 1995 course of bulimia nervosa and of eating disor-
Carpenter WT, Gunderson JG: Five year follow- der not otherwise specified: 5-year prospec-
up comparison of borderline and schizo- tive study of remissions, relapses, and the ef-
phrenic patients. Compr Psychiatry 18:567 fects of personality disorder psychopathology.
571, 1977 J Clin Psychiatry 68:738746, 2007
Cloninger CR, Svrakic DM, Przybeck TR: A psy- Grinker RR, Werble B, Drye RC: The Borderline
chobiological model of temperament and char- Syndrome. New York, Basic Books, 1968
acter. Arch Gen Psychiatry 50:975990, 1993 Gunderson JG, Shea MT, Skodol AE, et al: The
Cohen P, Crawford TN, Johnson JG, et al: The Chil- Collaborative Longitudinal Personality Disor-
dren in the Community study of developmen- ders Study, I: development, aims, designs and
tal course of personality disorder. J Personal sample characteristics. J Personal Disord
Disord 19:466486, 2005 14:300315, 2000
Crawford TN, Cohen P, Brook JS: Dramatic-erratic Gunderson JG, Morey LC, Stout RL, et al: Major
personality disorder symptoms, I: continuity depressive disorder and borderline personal-
from early adolescence into adulthood. J Per- ity disorder revisited: longitudinal interac-
sonal Disord 15:319335, 2001a tions. J Clin Psychiatry 65:10491056, 2004
Crawford TN, Cohen P, Brook JS: Dramatic-erratic Gunderson JG, Stout RL, Sanislow CA, et al: New
personality disorder symptoms, II: develop- episodes and new onsets of major depression in
Course and Outcome 77

borderline and other personality disorders. Krueger RF, Tackett JL: Personality and psychopa-
J Affect Disord 111:4045, 20008 thology: working toward the bigger picture. J
Heatherton TF, Weinberger JL (eds): Can Person- Personal Disord 17:109128, 2003
ality Change? Washington, DC, American Lenzenweger MF: Stability and change in person-
Psychological Press, 1994 ality disorder features: the Longitudinal Study
Johnson JG, Williams JBW, Goetz RR, et al: Stabil- of Personality Disorders. Arch Gen Psychiatry
ity and change in personality disorder symp- 56:10091015, 1999
tomatology: findings from a longitudinal Lenzenweger MF, Loranger AW, Korfine L, et al:
study of HIV+ and HIV men. J Abnorm Psy- Detecting personality disorders in a nonclini-
chol 106:154158, 1997 cal population: application of a 2-stage proce-
Johnson JG, Cohen P, Kasen S, et al: Age-related dure for case identification. Arch Gen Psychi-
change in personality disorder trait levels be- atry 54:110133, 1997
tween early adolescence and adulthood: a Levy KN, Becker DF, Grilo CM, et al: Concurrent
community-based longitudinal investiga- and predictive validity of the personality dis-
tion. Acta Psychiatr Scand 102:265275, 2000a order diagnosis in adolescent inpatients. Am J
Johnson JG, Cohen P, Smailes E, et al: Adolescent Psychiatry 156:15221528, 1999
personality disorders associated with violence Lewinsohn PM, Rohde P, Seeley JR, et al: Axis II
and criminal behavior during adolescence and psychopathology as a function of Axis I disor-
early adulthood. Am J Psychiatry 157:1406 der in childhood and adolescents. J Am Acad
1412, 2000b Child Adolesc Psychiatry 36:17521759, 1997
Johnson JG, First MB, Cohen P, et al: Adverse out- Livesley WJ, Jackson DN, Schroeder ML: Factorial
comes associated with personality disorder structure of traits delineating personality disor-
not otherwise specified in a community sam- ders in clinical and general population sam-
ple. Am J Psychiatry 162:926932, 2005 ples. J Abnorm Psychol 101:432440, 1992
Judd PH, McGlashan TH: A Developmental Loranger AW. Personality Disorder Examination
Model of Borderline Personality Disorder: Un- (PDE) Manual. Yonkers, NY, DV Communica-
derstanding Variations in Course and Out- tions, 1988
come. Washington, DC, American Psychiatric Loranger AW, Lenzenweger MF, Gartner AF, et al:
Publishing, 2003 Trait-state artifacts and the diagnosis of per-
Katz MM, Secunda SK, Hirschfeld R, et al: NIMH sonality disorders. Arch Gen Psychiatry
Clinical Research Branch Collaborative Pro- 48:720728, 1991
gram on Psychobiology of Depression. Arch Loranger AW, Sartorius N, Andreoli A, et al: The
Gen Psychiatry 36:765771, 1979 International Personality Disorder Examina-
Keller MB, Shapiro RW, Lavori PW, et al: Recovery tion (IPDE): the World Health Organization/
in major depressive disorder: analysis with the Alcohol, Drug Abuse, and Mental Health Ad-
lifetable and regression models. Arch Gen ministration International Pilot Study of Per-
Psychiatry 39:905910, 1982 sonality Disorders. Arch Gen Psychiatry
Keller MB, Lavori PW, Friedman B, et al: The Lon- 51:215224, 1994
gitudinal Interval Follow-Up Evaluation. Lynam DR, Widiger TA: Using the five-factor
Arch Gen Psychiatry 44:540548, 1987 model to represent the DSM-IV personality
Keller MB: Social anxiety disorder clinical course disorders: an expert consensus approach. J
and outcome: review of Harvard/Brown Anx- Abnorm Psychol 110:401412, 2001
iety Research Program (HARP) findings. Lyons MJ, Tyrer P, Gunderson J, et al: Special fea-
J Clin Psychiatry 67(suppl):1419, 2006 ture: heuristic models of comorbidity of Axis I
Kendell PC, Clarkin JF: Introduction to special and Axis II disorders. J Personal Disord
section: comorbidity and treatment implica- 11:260269, 1997
tions. J Consult Clin Psychol 60:833834, 1992 Maddocks PD: A five year follow-up of untreated
Klein DN, Shih JH: Depressive personality: associ- psychopaths. Br J Psychiatry 116:511515, 1970
ations with DSM-III-R mood and personality Markowitz JC, Skodol AE, Petkova E, et al: Longi-
disorders and negative and positive affectiv- tudinal effects of personality disorders on psy-
ity, 30-month stability, and prediction of chosocial functioning of patients with major
course of axis I depressive disorders. J Ab- depressive disorder. J Clin Psychiatry 68:186
norm Psychol 107:319327, 1998 193, 2007
Krueger RF: The structure of common mental dis- Mattanah JJ, Becker DF, Levy KN, et al: Diagnostic
orders. Arch Gen Psychiatry 56:921926, 1999 stability in adolescents followed up 2 years af-
Krueger RF, Carlson SR: Personality disorders in ter hospitalization. Am J Psychiatry 152:889
children and adolescents. Curr Psychiatry Rep 894, 1995
3:4651, 2001
78 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

McDavid JD, Pilkonis PA: The stability of person- a retrospective study. J Am Acad Child Ado-
ality disorder diagnoses. J Personal Disord lesc Psychiatry 42:11221130, 2003
10:115, 1996 Rey JM, Morris-Yates A, Singh M, et al: Continu-
McGlashan TH: The Chestnut Lodge Follow-Up ities between psychiatric disorders in adoles-
Study, part III: long-term outcome of border- cents and personality disorders in young
line personalities. Arch Gen Psychiatry 42:20 adults. Am J Psychiatry 152:895900, 1995
30, 1986a Robbins E, Gentry KA, Munoz RA, et al: A con-
McGlashan TH: The Chestnut Lodge Follow-up trast of the three more common illnesses with
Study, part VI: schizotypal personality disor- the ten less common in a study and 18-month
der. Arch Gen Psychiatry 43:329334, 1986b follow-up of 314 psychiatric emergency room
McGlashan TH, Grilo CM, Skodol AE, et al: The patients, III: findings at follow-up. Arch Gen
Collaborative Longitudinal Personality Disor- Psychiatry 34:285291, 1977
ders Study: baseline Axis I/II and II/II diag- Roberts BW, DelVecchio WF: The rank-order con-
nostic co-occurrence. Acta Psychiatr Scand sistency of personality traits from childhood
102:256264, 2000 to old age: a quantitative review of longitudi-
McGlashan TH, Grilo CM, Sanislow CA, et al: nal studies. Psychol Bull 126:325, 2000
Two-year prevalence and stability of individ- Roberts BW, Caspi A, Moffitt TE: The kids are al-
ual DSM-IV criteria for schizotypal, border- right: growth and stability in personality devel-
line, avoidant, and obsessive-compulsive dis- opment from adolescence to adulthood. J Pers
orders: toward a hybrid model of Axis II Soc Psychol 81:670683, 2001
disorders. Am J Psychiatry 162:883889, 2005 Robins LN: Deviant Children Grown Up: A Socio-
Millon T: Millon Clinical Multiaxial InventoryIII. logical and Psychiatric Study of Sociopathic
Minneapolis, MN, National Computer Sys- Personality. Baltimore, MD, Williams & Wilkins,
tems, 1987 1966
Morey LC, Hopwood CJ, Gunderson JG, et al: Robins LN: Epidemiology: reflections on testing
Comparison of alternative models for person- the validity of psychiatric interviews. Arch
ality disorders. Psychol Med 37:983994, 2007 Gen Psychiatry 42:918924,1985
Myers MG, Stewart DG, Brown SA: Progression Ruegg R, Frances A: New research in personality
from conduct disorder to antisocial personality disorders. J Personal Disord 9:148, 1995
disorder following treatment for adolescent Sanislow CA, McGlashan TH: Treatment outcome
substance abuse. Am J Psychiatry 155:479485, of personality disorders. Can J Psychiatry
1998 43:237250, 1998
Nesselroade JR, Stigler SM, Baltes PB: Regression Seivewright H, Tyrer P, Johnson T: Change in per-
toward the mean and the study of change. sonality status in neurotic disorders. Lancet
Psychol Bull 88:622637, 1980 359:22532254, 2002
Newman DL, Caspi A, Moffitt TE, et al: Anteced- Shea MT, Yen S: Stability as a distinction between
ents of adult interpersonal functioning: effects Axis I and Axis II disorders. J Personal Disord
of individual differences in age three tempera- 17:373386, 2003
ment. Dev Psychol 33:206217, 1997 Shea MT, Stout RL, Gunderson JG, et al: Short-term
Oldham JM, Skodol AE, Kellman HD, et al: Diag- diagnostic stability of schizotypal, borderline,
nosis of DSM-III-R personality disorders by avoidant, and obsessive-compulsive personal-
two structured interviews: patterns of comor- ity disorders. Am J Psychiatry 159:20362041,
bidity. Am J Psychiatry 149: 213220, 1992 2002
Ottosson H, Ekselius L, Grann M, et al: Cross-sys- Shea MT, Stout RL, Yen S, et al: Associations in the
tem concordance of personality disorder diag- course of personality disorders and Axis I dis-
noses of DSM-IV and diagnostic criteria for re- orders over time. J Abnorm Psychol 113:449
search of ICD-10. J Personal Disord 16:283 508, 2004
292, 2002 Siever LJ, Davis KL: A psychobiological perspec-
Paris J: Personality disorders over time: precur- tive on the personality disorders. Am J Psychi-
sors, course and outcome. J Personal Disord atry 148:16471658, 1991
17:479488, 2003 Skodol AE: Classification, assessment, and differ-
Paris J, Zweig-Frank H: A twenty-seven year fol- ential diagnosis of personality disorders. J
low-up of borderline patients. Compr Psychi- Pract Psychiatry Behav Health 3:261274, 1997
atry 42:482487, 2001 Skodol AE, Grilo CM, Pagano ME, et al: Effects of
Perry JC: Longitudinal studies of personality dis- personality disorders on functioning and
orders. J Personal Disord 7 (suppl):6385, 1993 well-being in major depressive disorder. J Psy-
Rettew DC, Zanarini MC, Yen S, et al: Childhood chiatr Pract 11:363368, 2005a
antecedents of avoidant personality disorder:
Course and Outcome 79

Skodol AE, Gunderson JG, Shea MT, et al: The orders: implications for research and the evo-
Collaborative Longitudinal Personality Disor- lution of Axis II. Am J Psychiatry 154:895903,
ders Study (CLPS): overview and implica- 1997
tions. J Personal Disord 19:487504, 2005b Westen D, Shedler J: Revising and assessing Axis
Skodol AE, Oldham JM, Bender DS, et al: Dimen- II, part II: toward an empirically based and
sional representations of DSM-IV personality clinically useful classification of personality
disorders: relationships to functional impair- disorders. Am J Psychiatry 156: 273285, 1999
ment. Am J Psychiatry 162:19191925, 2005c Widiger TA: Categorical versus dimensional clas-
Skodol AE, Pagano ME, Bender DS, et al: Stability sification: implications from and for research.
of functional impairment in patients with J Personal Disord 6:287300, 1992
schizotypal, borderline, avoidant, or obses- Widiger TA: Personality disorder and Axis I psy-
sive-compulsive personality disorder over chopathology: the problematic boundary of
two years. Psychol Med 35:443451, 2005d Axis I and Axis II. J Personal Disord 17:90108,
Solomon DA, Keller MB, Leon AC, et al: Recovery 2003
from major depression: a 10-year prospective Wolff S, Townshend R, McGuire RJ, et al: Schizoid
follow-up across multiple episodes. Arch Gen personality in childhood and adult life, II:
Psychiatry 54:10011006, 1997 adult adjustment and the continuity with
Squires-Wheeler E, Skodol AE, Erlenmeyer-Kim- schizotypal personality disorder. Br J Psychia-
ling L: The assessment of schizotypal features try 159:620629, 1991
over two points in time. Schizophrenia Res World Health Organization: The ICD-10 Classifi-
6:7585, 1992 cation of Mental and Behavioural Disorders:
Stevenson J, Meares R, Comerford A: Diminished Clinical Description and Diagnostic Guide-
impulsivity in older patients with borderline lines. Geneva, World Health Organization,
personality disorder. Am J Psychiatry 160:165 1992
166, 2003 Zanarini MC, Skodol AE, Bender D, et al: The Col-
Stone MH: Long-term outcome in personality dis- laborative Longitudinal Personality Disorders
orders. Br J Psychiatry 162:299313, 1993 Study: reliability of Axis I and II diagnoses.
Trull TJ, Useda JD, Conforti K, et al: Borderline J Personal Disord 14:291299, 2000
personality disorder features in nonclinical Zanarini MC, Frankenburg FR, Hennen J, et al:
young adults, 2: two-year outcome. J Abnorm The longitudinal course of borderline psycho-
Psychol 106:307314, 1997 pathology: 6-year prospective follow-up of the
Trull TJ, Useda JD, Doan BT, et al: Two-year stabil- phenomenology of borderline personality dis-
ity of borderline personality measures. J Per- order. Am J Psychiatry 160:274283, 2003
sonal Disord 12:187197, 1998 Zanarini MC, Frankenburg FR, Hennen J, et al:
Tyrer P: Personality disorder. Br J Psychiatry Axis I comorbidity in patients with borderline
179:8184, 2001 personality disorder: 6-year follow-up and
Tyrer P, Simonsen E: Personality disorder in psy- prediction of time to remission. Am J Psychia-
chiatric practice. World Psychiatry 2:4144, try 161:21082114, 2004
2003 Zanarini MC, Frankenburg FR, Hennen J, et al:
Tyrer P, Strauss J, Cicchetti D: Temporal reliability The McLean Study of Adult (MSAD): over-
of personality in psychiatric patients. Psychol view and implications of the first six years of
Med 13:393398, 1983 prospective follow-up. J Personal Disord
Tyrer P, Gunderson J, Lyons M, et al: Special fea- 19:505523, 2005
ture: extent of comorbidity between mental Zanarini MC, Frankenburg FR, Hennen J, et al:
state and personality disorders. J Personal Dis- Prediction of the 10-year course of borderline
ord 11:242259, 1997 personality disorder. Am J Psychiatry 163:827
Warner MB, Morey LC, Finch JF, et al: The longitu- 832, 2006
dinal relationship of personality traits and dis- Zimmerman M: Diagnosing personality disor-
orders. J Abnorm Psychol 113:217227, 2004 ders: a review of issues and research methods.
Westen D: Divergences between clinical and re- Arch Gen Psychiatry 51:225245, 1994
search methods for assessing personality dis-
This page intentionally left blank
Part III
This page intentionally left blank
Prevalence, Sociodemographics,
and Functional Impairment
Svenn Torgersen, Ph.D.

F rom clinical work we get an impression of ciodemographic features may increase the
which personality disorders are more com- likelihood of a particular person to seek treat-
mon and which are rarer. However, people ment. These complexities mean that only
with some types of personality disorders population (epidemiological) studies can
may be more likely to seek treatment and ob- demonstrate the true relationship between
tain treatment compared with people with personality disorders and socioeconomic and
other types of personality disorders. Conse- sociodemographic variables, or any other
quently, if we are interested in how preva- variables such as traumas, disastrous events,
lent different personality disorders are in the upbringing, or partner relationships.
general population, we have to study repre-
sentative samples of the general population.
Epidemiological research does just that. PREVALENCE
Clinical work also gives us ideas about re-
lationships between socioeconomic and so- We know much about the prevalence of Axis
ciodemographic factors and personality dis- I disorders in the general population (Kring-
orders. However, in a clinical setting we only len et al. 2001). As to personality disorders,
meet those from an unfavorable environment however, less is known. Some studies have
who have developed a personality disorder. been performed, but few of them adequately
We do not meet those from an unfavorable represent the general population (Torgersen
environment who have not developed a dis- et al. 2001). In this chapter I review published
order. Furthermore, the combination of a studies that are closest to what one might call
specific personality disorder and specific so- an epidemiological population study. These

84 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

individual studies are discussed below (and medical or psychiatric history was per-
presented in Table 51) in view of different formed. The control probands, their spouses,
elements of epidemiology, beginning with a and first-degree relatives constituted the
discussion of sample selection for each study. sample.
The sample studied by Moldin et al.
Sample Selection (1994) consisted of parents and their off-
spring in two control groups used in the New
The sample studied by Zimmerman and York High Risk Project. One of the groups
Coryell (1989, 1990) included first-degree was recruited from two schools in the New
relatives of normal subjects (23%) and of psy- York metropolitan area. The other group
chiatric patients (mood disorders and schiz- came from the pool of a population sampling
ophrenia) as well as a smaller group of first- firm. The subjects were white, English-
degree relatives of nonpsychotic psychiatric speaking families screened for psychiatric
patients. Thus, even if this is a nonpatient disorder.
sample, it is not an average population sam- In the study by Klein et al. (1995), the sam-
ple. However, the prevalence of mania was ple comprised relatives of a control group
not higher than 2%, and the prevalence of screened for Axis I disorders in Stony Brook,
schizophrenia was not higher than 1%. Inter- New York. The interviews were partly con-
views were conducted in person (27%) and ducted in person and partly by telephone.
by telephone (73%). The Structured Inter- Lenzenweger et al. (1997) examined a
view for DSM-III Personality Disorders was sample consisting initially of 1,684 under-
applied (Stangl et al. 1985). The study took graduate students from Cornell University in
place in Iowa City. New York. They were screened by means of a
The sample reported on by Black et al. questionnaire; a sample of those expected
(1993) consisted of 120 relatives of 32 outpa- and those not expected to have a personality
tients with obsessive-compulsive disorder disorder was interviewed. The total number
and 127 relatives of a comparison group of subjects interviewed was 258. In this over-
screened for Axis I disorders. Strangely, no view (Table 51), I apply the actual numbers.
difference was found between the preva- The estimated prevalence for any personality
lence of personality disorders in the two rel- disorder is a little different.
ative groups. More than half of the sample The study by Torgersen et al. (2001) was
were siblings, a quarter were parents, and conducted in Oslo, the capital of Norway. A
the rest were children. A little more than half random sample of names of 3,590 citizens be-
were women. The mean age was 42 years. tween age 18 and 65 years was selected from
More than half were interviewed in person the National Register of Oslo. Some had
and the rest by telephone. moved out of town, some were impossible to
Maier et al. (1992) conducted one of the trace, and some were dead. Others refused to
few reported studies in which the sample is participate or postponed the interview be-
relatively representative of the general pop- yond the period of the study (18%). Of the
ulation. Control probands were selected by a original sample, 2,053 (57%) delivered inter-
marketing company to match patients older views of sufficient quality for the study. All
than age 20 on sex, age, residential area, and interviews were performed in person. The
educational level. The participants had to sampling procedure made it possible to iden-
have at least one living first-degree relative tify all causes of reduction in the sample from
who also had agreed to be interviewed. Oth- the initial to the final sample. There were al-
erwise, this sample represented the general most equal numbers of men and women.
population of a mixed urban/rural German The sampling procedure used by Sam-
residential area near Mainz. No screening for uels et al. (2002) was very complicated. Ini-
Prevalence, Sociodemographics, and Functional Impairment 85

tially, a sample of 3,481 adult household res- R (p. 554) was interviewed with the com-
idents in Baltimore was studied in the 1980s. plete IPDE.
About 10 years later, a subsample was se-
lected that included individuals previously Results
evaluated by psychiatrists or those who ap-
peared to have an Axis I diagnosis based on Table 51 presents the prevalences in the
the Diagnostic Interview Schedule. In addi- published studies discussed above, includ-
tion, a random sample was selected. A num- ing all personality disorders. So-called
ber of subjects could not be traced, refused, mixed personality disorders, defined by the
were too ill to participate, or were deceased. absence of one criterion for two or more per-
The remaining sample consisted of 742 indi- sonality disorders and not having the re-
viduals. Their ages varied between 34 and 94 quired number of criteria for any disorder,
years, and two-thirds were women. are excluded. The same is the case for per-
In the study by Grant et al. (2004), a very sonality disorders not otherwise specified
large sample (N=43,093; the so-called Na- (NOS), defined by having 10 or more criteria,
tional Epidemiologic Survey on Alcohol and but not a sufficient number of criteria ob-
Related Conditions [NESARC]), was inter- tained for any specific disorder. Unweighted
viewed. A measure named Alcohol Use Dis- prevalences (rather than weighted preva-
order and Associated Disabilities Interview lences based on questionable weighting pro-
SchedulIV (AUDADIS-IV) was applied. cedures) are presented, if possible, because
Borderline, schizotypal, and narcissistic per- the prevalences among those not reached
sonality disorders were not included. The re- cannot be known. The qualified, although
sults were weighted to represent the U. S. questionable, guesswork gives one an impres-
population age 18 years and older. sion of increased accuracy. A nonweighted
Crawford et al. (2005) studied prospec- rate is transparent and does not claim more
tively 597 community members from age 14 than it can stand for.
to 33 years. The reported percentages in Ta- In almost half of the studies, the preva-
ble 51 are based on the application of the lence of any disorder is between 7% and 13%.
Structured Clinical Interview for DSM-IV The median prevalence is above 10%. The
Axis II Personality Disorders (SCID-II) to re- median prevalence of specific personality
spondents at age 33 years. disorders is mostly between 1% and 2%.
Coid et al. (2006) interviewed 626 indi- Obsessive-compulsive personality disor-
viduals age 1674 years in the United King- der is the most frequently occurring personal-
dom. Table 51 reports the weighted preva- ity disorder (above 2%). Next come paranoid,
lences (unweighted prevalences for the avoidant, and passive-aggressive personality
specific disorders are not stated in the publi- disorder (1.7%); then borderline (1.6%), histri-
cation). The SCID-II screening questionnaire onic (1.5%), antisocial (1.1%), and schizoid
was applied. and schizotypal (0.9%); and finally narcissis-
Personality disorders were assessed in tic personality disorder (0.5%). The only two
the National Comorbidity Survey Repli- studies that included self-defeating and sa-
cation (NCS-R). A set of the International distic personality disorder reported an aver-
Personality Disorder Examination (IPDE) age of 0.4% and 0.1% respectively.
screening questions was applied to 5,692 in- The correlations between the different
dividuals age 18 years or older (Lenzen- studies are very low, around 0.17. The only
weger et al. 2007). A subsample of 214 indi- two studies that reflect the average preva-
viduals who screened positive for one or lences (calculated by intraclass correlation)
more of our outcome measures on the basis are those of Maier et al. (1992) and Torgersen
of the IPDE screening questions in the NCS- et al. (2001).
Table 51. Prevalences of personality disorders in twelve population studies

Zimmerman Black Maier Moldin Klein Lenzenweger Torgersen Samuels Grant Crawford Coid Lenzenweger
and Coryell et al. et al. et al. et al. et al. et al. et al. et al. et al. et al. et al.
1989 1993 1992 1994 1995 1997 2001 2002 2004 2005 2006 2007 Median

Place Iowa Iowa Mainz NYC New York New York Oslo Baltimore USA New York UK USA
-IV screening

E s s e n t i a l s o f P e r so n a l i t y D i s o r d e r s
Paranoid 0.9 1.6 1.8 0.0 1.8 0.4 2.2 0.7 4.4 5.1 5.1 2.3 1.7
Schizoid 0.9 0.0 0.4 0.0 0.9 0.4 1.6 0.7 3.1 1.7 1.7 4.9 0.9
Schizotypal 2.9 3.2 0.7 0.7 0.0 0.0 0.6 1.8 1.1 1.1 3.3 0.9
Antisocial 3.3 0.8 0.2 2.6 2.6 0.8 0.6 4.5 3.6 1.2 1.2 1.0 1.1
Borderline 1.7 3.2 1.1 2.0 1.8 0.0 0.7 1.2 3.9 3.9 1.6 1.6
Histrionic 3.0 3.2 1.3 0.3 1.8 1.9 1.9 0.4 1.8 0.9 0.9 0.0 1.5
Narcissistic 0.0 0.0 0.0 0.0 4.4 1.2 0.8 0.1 2.2 2.2 0.0 0.5
Avoidant 1.3 2.0 1.1 0.7 5.7 0.4 5.0 1.4 2.4 6.4 6.4 5.2 1.7
Dependent 1.8 1.6 1.6 1.0 0.4 0.4 1.5 0.3 0.5 0.8 0.8 0.6 0.7

Obsessive- 2.0 9.3 2.2 0.7 2.6 0.0 1.9 1.2 7.9 4.7 4.7 2.4 2.1
Passive- 3.3 10.5 1.8 1.7 1.8 0.0 1.6 1.7
Self-defeating 0.0 0.8 0.4
Sadistic 0.0 0.2 0.1
Table 51. Prevalences of personality disorders in twelve population studies (continued)

Zimmerman Black Maier Moldin Klein Lenzenweger Torgersen Samuels Grant Crawford Coid Lenzenweger
and Coryell et al. et al. et al. et al. et al. et al. et al. et al. et al. et al. et al.
1989 1993 1992 1994 1995 1997 2001 2002 2004 2005 2006 2007 Median

Prevalence, Sociodemographics, and Functional Impairment

Any personality 14.3 22.7 10.0 7.3 14.8 3.9 13.1 10.0 15.7 4.4 10.3 10.3
Number 797 247 452 303 229 258 2,053 742 43,093 597 626 214

Note. IPDE=International Personality Disorders Examination; NYC=New York City; PDE=Personality Disorder Examination; SCID-II=Structured Clinical Interview for DSM-
IV Axis II Personality Disorders; SIDP=The Structured Interview for DSM-III Personality Disorders; SIDP-R=The Structured Interview for DSM-III-R Personality Disorders.

88 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

Dependent and borderline personality are between 1% and 2%. The validity of the
disorders display the most stable preva- prevalences of those personality disorders
lences from study to study, while narcissistic that are less frequently occurring is question-
and passive-aggressive personality disor- able. The prevalence of any personality dis-
ders vary most from study to study (based on order is around 10%. The sum of the per-
spreading/mean ratio or t test). It is notewor- centages for the specific disorders is higher,
thy that these two disorders are not included around 15%, pointing to the fact that around
among the ICD-10 (World Health Organiza- one-third of those with one disorder have
tion 1992) or DSM-IV (American Psychiatric one more. (The distribution is skewed with
Association 1994) personality disorders. most having only one, some having two, and
Table 52 shows a comparison between few having many personality disorders,
the prevalences in a large outpatient clinic in hence 50% of individuals will not have more
Oslo (Alns and Torgersen 1988) and in the than one disorder.) Studies of patients give
general population of that city (Torgersen et us a distorted impression of the absolute and
al. 2001). The ratio between the prevalence in relative prevalences of personality disorders,
the clinic and that in the population is calcu- because those with dependent and border-
lated separately for women and men and in line much more often seek treatment.
the total sample. There are relatively small
differences in the range of the ratios between
women and men, even if the ratios are a little SOCIODEMOGRAPHIC CORRELATES
larger for the specific personality disorder
among men (not for any personality disor- Gender
ders). Those with dependent, borderline,
avoidant, and obsessive-compulsive person- Gender differences are common among
ality disorder are strongly overrepresented mental disorders. Women more often have
among the patients based on prevalence mood and anxiety disorders, and men more
rates in the general population, whereas often have substance-related disorders
those with antisocial, schizoid, and paranoid (Kringlen et al. 2001). For personality disor-
personality disorder are less common in the ders, women and men also differ.
clinical compared with the general popula- With regard to personality disorders,
tion. To have a borderline, avoidant, or Zimmerman and Coryell (1989) observed a
schizotypal personality disorder implies higher prevalence of any personality disor-
pain and dysfunction, as I discuss later in the ders among males as did Jackson and Bur-
chapter. One may speculate that those who gess (2000) for ICD-10 screening when re-
are dependent seek help, whereas obsessive- gression analysis was applied. However,
compulsive patients want to do something differences between genders were very
with their problems, even if they do not suf- small, and Torgersen et al. (2001) did not ob-
fer as much. In the other direction, those who serve any differences.
are antisocial do not want psychological help As to the personality disorder clusters,
and are also refused help. Schizoid individu- Samuels et al. (2002) and Torgersen et al.
als keep their distance, whereas paranoid (2001) reported that Cluster A (odd/eccen-
subjects do not believe in any cure. tric) and Cluster B (dramatic/emotional) per-
In conclusion, while the prevalences of sonality disorders or traits were more com-
the specific personality disorders vary strongly mon among men. Coid et al. (2006) found the
from study to study, the number of pub- same for Cluster B only.
lished studies makes it possible to draw Among the specific Cluster A disorders,
some conclusions today. At least in the both Torgersen et al. (2001) and Zimmerman
United States and Europe, the prevalences and Coryell (1990) found that schizoid per-
Table 52. Prevalences of personality disorders in the common population and among outpatients in Oslo, Norway

Females Males Total

Personality Torgersen et Alns and Ratio Torgersen et Alns and Ratio Torgersen et Alns and Ratio
disorder al. 2001 Torgersen (range) al. 2001 Torgersen (range) al. 2001 Torgersen (range)

Prevalence, Sociodemographics, and Functional Impairment

1988 1988 1988
Paranoid 2.2 3.9 1.8 (9) 2.3 7.6 3.3 (9) 2.2 5.0 2.3 (9)
Schizoid 1.1 0.0 0.0 (10) 2.2 5.4 2.5 (10) 1.6 1.7 1.1 (10)
Schizotypal 0.6 3.9 6.5 (6) 0.5 12.0 24.0 (3) 0.6 6.4 10.7 (4)
Antisocial 0.0 0.0 0.0 (10) 1.3 0.0 0.0 (11) 0.6 0.0 0.0 (11)
Borderline 0.9 17.0 18.9 (2) 0.4 9.8 24.5 (2) 0.7 14.8 21.1 (2)
Histrionic 2.5 15.0 6.0 (7) 1.2 10.9 9.1 (7) 1.9 13.8 7.3 (6)
Narcissistic 0.8 1.9 2.4 (8) 0.9 10.9 12.1 (6) 0.8 4.7 5.9 (8)
Avoidant 5.0 53.4 10.7 (3) 4.9 59.8 12.2 (5) 5.0 55.4 11.1 (3)
Dependent 2.0 47.6 23.8 (1) 0.9 45.7 50.8 (1) 1.5 47.0 31.3 (1)
Obsessive- 1.3 13.6 10.5 (4) 2.6 33.7 13.0 (4) 1.9 19.8 10.4 (5)
Passive- 0.9 6.3 7.0 (5) 2.2 18.5 8.4 (8) 1.6 10.1 6.3 (7)
Any 12.6 76.7 6.1 13.7 90.2 6.6 13.1 80.9 6.2

Number 1,142 206 911 92 2,053 298

90 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

sonality disorder or traits were more com- borderline traits. In patient samples, border-
mon among men. Zimmerman and Coryell line personality disorder is not more preva-
(1990) found this also for paranoid traits. lent among women than among men (Alns
Grant et al. (2004), however, observed that and Torgersen 1988; Fossati et al. 2003;
women more often had a paranoid personal- Golomb et al. 1995). In one study of patients,
ity disorder. Neither Zimmerman and Cor- borderline personality disorder was, in fact,
yell (1989, 1990) nor Torgersen et al. (2001) more common among men than among
observed any gender difference for schizo- women (Carter et al. 1999). In our unsystem-
typal personality disorder. Among the Clus- atic impression of people, we are more likely
ter B personality disorders, antisocial disor- to see borderline features in women than
der is much more common among men in men. That schizotypal personality disor-
(Torgersen et al. 2001; Zimmerman and Cor- der does not show any gender bias will more
yell 1989, 1990). Those with histrionic per- easily be recognized. A trend in the direction
sonality disorder or traits appear more often of men being more often schizoid, narcissis-
to be women (Torgersen et al. 2001; Zimmer- tic, and obsessive-compulsive and women
man and Coryell 1990). Narcissistic traits are being more often histrionic and avoidant is
found more often among men, and there are in accordance with common opinion.
no statistically significant gender differences
for borderline personality disorder or traits Age
(Torgersen et al. 2001; Zimmerman and
Coryell 1990). To diagnose a personality disorder in an indi-
Among the Cluster C (anxious/fearful) vidual younger than age 18 years, the features
personality disorders, dependent personality must have been present at least 1 year (Amer-
disorder is much more common among ican Psychiatric Association 2000). At the
women (Grant et al. 2004; Torgersen et al. same time, it is assumed that personality dis-
2001; Zimmerman and Coryell 1989, 1990), orders start early in life and are relatively sta-
and obsessive-compulsive personality disor- ble. For some personality disorders, espe-
der or traits are found more often among men cially the dramatic types, it is also assumed
(Torgersen et al. 2001; Zimmerman and that they are typical for young people. On the
Coryell 1989, 1990). Zimmerman and Coryell other hand, the older people are, the longer
(1989, 1990) and Grant et al. (2004) reported they have had to develop personality disor-
more avoidant personality disorder and traits ders, even though personality disorders may
among women. also disappear. Suicide and fatal accidents
Regarding personality disorders pro- also may happen more often among those
vided for further study (American Psych- with personality disorders than among other
iatric Association 2000), Torgersen et al. individuals. These facts will influence the rate
(2001)but not Zimmerman and Coryell of specific personality disorders in older age.
(1989, 1990)found that men more often had What does empirical research show?
passive-aggressive personality disorder. Tor- Zimmerman and Coryell (1989) observed
gersen and colleagues also found that women that individuals with personality disorders
more often presented with self-defeating were younger than those without. Jackson
traits, and men more often presented with sa- and Burgess (2000) found the same using a
distic traits. short ICD-10 screening instrument (IPDE
The most clear-cut results from the stud- screener). Torgersen et al. (2001), however,
ies are that men tend to be antisocial and observed the opposite. This can be explained
women tend to be dependent. These results by the high prevalence of introverted and low
are perhaps not surprising. However, more prevalence of impulsive personality traits in
surprising is a lack of gender difference for Norway as compared with the United States.
Prevalence, Sociodemographics, and Functional Impairment 91

As to the clusters of personality disor- and obsessive-compulsive personality disor-

ders, Torgersen et al. (2001) found that indi- ders are more common in younger age.
viduals with odd/eccentric personality dis- Zimmerman and Coryell (1989) found
orders were older, whereas Samuels et al. that individuals with passive-aggressive per-
(2002), Coid et al. (2006), and Lenzenweger et sonality disorder are typically of a younger
al. (2007) did not find any age variation. For age, and Torgersen et al. (2001) observed that
the dramatic/emotional cluster, Samuels et such traits were negatively correlated with
al. (2002), Coid et al. (2006), and Lenzen- age. The latter study also examined self-de-
weger et al. (2007) found a higher prevalence feating and sadistic traits and found that sa-
among the younger subjects, whereas Torg- distic traits were associated with younger age.
ersen et al. (2001) found that the dramatic/ To summarize, persons with antisocial,
emotional trait dimensions decreased with borderline, or histrionic personality disorder
age. As to the anxious/fearful cluster, neither seem to be younger. As for other disorders,
group observed any age trend. the studies are conflicting. The reason for the
Among the odd/eccentric personality age difference in disorders is that people be-
disorders, schizoid personality disorder or come less impulsive and overtly aggressive
traits seem to be associated with being older as they age. Agreeableness and conscien-
(Torgersen et al. 2001; Zimmerman and tiousness increase with age (Srivastava et al.
Coryell 1989, 1990). Several studies have 2003). Cluster B disorders are typically nega-
found that paranoid personality disorder is tively correlated with agreeableness and con-
unrelated to age (Torgersen et al. 2001; Zim- scientiousness (Saulsman and Page 2004).
merman and Coryell 1989, 1990). Zimmer-
man and Coryell (1989, 1990) observed that Marital Status
those with schizotypal personality disorder
were younger, and Torgersen et al. (2001) Most of the results concerning marital status
found that they were older. Grant et al. (2004) are from Zimmerman and Coryell (1989).
found that both schizoid and paranoid per- Some of the data from Torgersen et al. (2001)
sonality disorder decreased with age. have been calculated for this chapter to fit the
Among the dramatic/emotional person- tables in Zimmerman and Coryell (1989) (see
ality disorders, those with antisocial and bor- Table 53). Grant et al. (2004) also provide in-
derline personality disorder or traits are formation.
younger (Torgersen et al. 2001; Zimmerman As illustrated in Table 53, subjects with
and Coryell 1989, 1990), and Zimmerman personality disorder have more often been
and Coryell (1990) observed that those with separated or divorced compared with those
histrionic and narcissistic traits are younger without a personality disorder, and they are
as well. These results were not confirmed by more often divorced at the time of the inter-
Torgersen et al. (2001). Grant et al. (2004) ob- view (Zimmerman and Coryell 1989). They
served that histrionic and antisocial person- are less frequently married (Jackson and Bur-
ality disorders are more prevalent among gess 2000; Zimmerman and Coryell 1989),
younger respondents. and they are more often never married (Zim-
Zimmerman and Coryell (1989, 1990) did merman and Coryell 1989). If we include liv-
not find any age trend for any of the fearful ing nonmarried persons with a partner, sub-
disorders, whereas Torgersen et al. (2001) jects with personality disorder live more
observed that individuals with obsessive- often alone without a partner compared with
compulsive disorder and avoidant traits are those without a personality disorder in the
older. No difference was found for depen- general population (Torgersen et al. 2001).
dent personality disorders. Finally, Grant et However, as the risk of having a person-
al. (2004) reported that avoidant, dependent, ality disorder is related to gender and age,
Table 53. Marital status and personality disorders, calculated from Torgersen et al. (2001)
Single (never Ever Ever
Personality disorder Number married) (%) Married (%) Separateda (%) Divorceda (%) Widowed (%) separated b (%) divorcedc (%)
Paranoid 46 34.8 34.8 6.5 21.7c 2.2 15.8 36.7
Schizoid 32 56.3 31.3 0.0 6.3 6.3 20.0 28.6
Schizotypal 12 50.0 33.3 0.0 8.3 8.3 20.0 16.7
Antisocial 12 75.0d 8.3d 0.0 16.7 0.0 0.0 66.7

E s s e n t i a l s o f P e r so n a l i t y D i s o r d e r s
Borderline 14 57.1 35.7 7.1 0.0 0.0 20.0 16.7
Histrionic 39 46.2 35.9 0.0 17.9 0.0 0.0 47.6d
Narcissistic 17 35.6 52.9 0.0 5.9 5.9 10.0 9.1
Avoidant 102 45.1 36.3 1.0 14.7 2.9 7.5 28.6
Dependent 31 58.1d 25.8d 3.2 12.9 0.0 11.1 30.8
Obsessive-compulsive 39 41.6 43.6 0.0 10.3 5.1 5.6 21.7
Passive-aggressive 32 35.3 31.3 6.3 9.4 3.1 18.2 31.3
Self-defeating 17 35.3 17.6d 0.0 41.2e 5.9 25.0 63.6
Sadistic 4 50.0 56.0 0.0 0.0 0.0 0.0 0.0
Eccentric 80 45.6 33.8d 3.8 15.0 2.5 13.8 34.1
Dramatic 62 49.3 35.2 1.4 12.7 1.4 8.3 33.3
Fearful 189 45.5 36.5d 1.3 14.1 2.6 8.2 28.2
Any personality disorder 269 43.9 36.8f 2.2 15.6d 1.5 7.9 33.1f
No personality disorder 1,784 38.8 46.5 2.4 10.4 1.8 5.1 23.2

Number 2,053 693 830 43 185 33 43 253

At the time of interview.
bExcluding those
who are never married.
Excluding those who are never married and those who are divorced.
d X2 test, P<0.05
e 2
X test, P <0.001
fX2 test, P<0.01
Prevalence, Sociodemographics, and Functional Impairment 93

the real effect of other sociodemographic al. 2001). Those with schizoid personality
variables such as marital status is difficult to disorder are more seldom separated (Zim-
determine. Younger people are less often merman and Coryell 1989), more often never
married, and education is related to gender married or separated/divorced/widowed
and age. The best way to determine the inde- (Grant et al. 2004), and more often living
pendent effect of other sociodemographic alone (Torgersen et al. 2001). Those with
variables is to apply multivariate methods. schizotypal personality disorder have more
However, to apply such methods one needs often been separated (Zimmerman and
large samples. Thus multivariate methods Coryell 1989) and live more often alone
have been used in very few studies. In the (Torgersen et al. 2001).
study of Torgersen et al. (2001), such multi- Among the Cluster B disorders, persons
variate analyses have been carried out for with histrionic personality disorder have
living alone versus living with a partner. more often been separated or divorced (Zim-
Those with eccentric personality disor- merman and Coryell 1989). They are also
ders have more often been divorced or sepa- more often not married when interviewed
rated (Coid et al. 2006; Samuels et al. 2002); (Zimmerman and Coryell 1989), more often
they are more often divorced when inter- never married or divorced/separated/wid-
viewed, and they have seldom been married owed (Grant et al. 2004), and live more often
(Samuels et al. 2002; Table 53). Those with alone (Torgersen et al. 2001). Those with an-
dramatic personality disorders are also often tisocial personality disorder also more often
unmarried and live more often alone (Torg- have been divorced, separated (Zimmerman
ersen et al. 2001), and they are more often sep- and Coryell 1989), or never married (Grant et
arated or divorced (Coid et al. 2006). Those al. 2004; Table 53); are less often married
with fearful personality disorders are also when interviewed; and live more often alone
less often married (Samuels et al. 2002) and (Torgersen et al. 2001). Persons with border-
live more often alone (Torgersen et al. 2001). line personality disorder also have more of-
When we examine the specific personal- ten been separated if married, are more often
ity disorders, we encounter problems in com- divorced, and are not married when inter-
paring the different studies. Marital status viewed (Zimmerman and Coryell 1989).
does not seem to be as important in the Nor- They are more often never married (Zimmer-
wegian study, perhaps because many Nor- man and Coryell 1989), and live more often
wegians live in stable relationships without alone (Torgersen et al. 2001). Finally, those
being married. When we include living to- with narcissistic personality disorder also
gether with a partner from the study of Torg- more often live alone (Torgersen et al. 2001).
ersen et al. (2001) and consider this life situa- Among persons with anxious/fearful
tion as analogous to marriage, we find more personality disorders, those with avoidant
similarity between this study and the study personality disorder have more often been
by Zimmerman and Coryell (1989). It is im- separated (Zimmerman and Coryell 1989).
portant to note that the relationships in the They are more often separated/divorced/
Torgersen et al. study are based on logistic widowed when interviewed, and more often
and linear regression analysis, taking into ac- never married (Grant et al. 2004). Those with
count a number of other sociodemographic dependent personality disorder more often
variables. have been separated when interviewed (Zim-
Among the odd/eccentric personality merman and Coryell 1989), never married
disorders, those with paranoid personality (Grant et al. 2004; Table 53), or separated/
disorder are more often nev er married divorced/widowed (Grant et al. 2004). Those
(Grant et al. 2004), divorced (Grant et al. with obsessive-compulsive traits are less of-
2004; Table 53), or living alone (Torgersen et ten married (Torgersen et al. 2001).
94 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

Among the proposed personality disor- In applying logistic regression analysis

ders, persons with passive-aggressive per- and taking into account a number of other
sonality disorder have more often been di- sociodemographic variables, Torgersen et al.
vorced and are less often married when (2001) observed that paranoid and avoidant
interviewed (Zimmerman and Coryell 1989) personality disorders and traits and schiz-
and live more often alone (Torgersen et al. oid, schizotypal, antisocial, borderline, de-
2001). Those with self-defeating personality pendent, and self-defeating personality traits
disorder have more often been divorced were related to lower education. Interest-
(Zimmerman and Coryell 1989), are more of- ingly, those with obsessive-compulsive dis-
ten divorced (Table 53) and not married order or traits in fact had higher education.
when interviewed (Zimmerman and Coryell Only histrionic, narcissistic, and passive-
1989), and more often live alone (Torgersen et aggressive personality disorders or traits
al. 2001). were unrelated to education. Grant et al.
In conclusion, those with personality dis- (2004) found that lower education was re-
orders typically live alone, which is most lated to all the studied personality disorders
likely specifically for those with self-defeat- (paranoid, schizoid, antisocial, histrionic,
ing, borderline, or schizotypal personality avoidant, and dependent), with the excep-
disorder. Those with obsessive-compulsive tion of obsessive-compulsive personality
personality disorder may be an exception. disorder, which was related to higher educa-
Never being married is often observed tion (as Torgersen et al. [2001]) found.
among those with antisocial and dependent Coid et al. (2006) found that Cluster A
personality disorder. The risk of divorce/ disorders were related to unemployment
separation is high among those with para- and lower social class, Cluster B disorders
noid personality disorder. In cultures where were related to lower social class, and Clus-
it is more common to live together unmar- ter C disorders were related to being eco-
ried, a breakup in the relationship is less easy nomically inactive but not unemployed.
to record. For whatever reason, living with- Grant et al. (2004) found that lower income
out a partner is very common among those was related to all the studied personality dis-
with personality disorders. orders, except obsessive-compulsive person-
ality disorder. Lenzenweger et al. (2007)
Education and Income found that only borderline personality disor-
der was related to unemployment.
Relatively few studies have investigated the Samuels et al. (2002) also investigated the
relationship between personality disorders relationship between income and personal-
and education and income. Torgersen et al. ity disorders but did not find any association.
(2001) observed that those with any person- Jackson and Burgess (2000) did not find any
ality disorder had less education. The same relationship to unemployment. It is impor-
was observed for those with odd/eccentric tant to note that these studies applied multi-
personality disorders, and those with dra- variate methods, taking into account other
matic/emotional as well as anxious/fearful sociodemographic variables.
personality disorder traits. Samuels et al. In conclusion, personality disorders,
(2002) and Lenzenweger et al. (2007) con- with a few exceptions, are related to lower
firmed that those with dramatic/emotional socioeconomic status and economic prob-
personality disorders had less education but lems. This holds true for all of the Cluster A
not those with odd/eccentric or anxious/ disorders (paranoid, schizoid, and schizo-
fearful personality disorders. Coid et al. typal) and for at least two Cluster B disorders
(2006), however, found lower education (antisocial and borderline). It is not true for
among those with Cluster A disorders. narcissistic personality disorder, and the so-
Prevalence, Sociodemographics, and Functional Impairment 95

cioeconomic status of those with histrionic makes it easier to express the less socially ac-
personality disorder is equivocal. As to Clus- ceptable aspects of ones personality. We
ter C disorders, avoidant and dependent per- used to think that excessive social control
sonality disorders imply poorer socioeco- creates mental problems. Perhaps social con-
nomic status, whereas the opposite is true for trol hinders the development of accentuated
obsessive-compulsive personality disorder. eccentric, narcissistic, and impulsive person-
For the provisional disorders, there exists only ality styles.
one study (Torgersen et al. 2001) suggesting
lower education for those with self-defeating
and sadistic, but not passive-aggressive, per- QUALITY OF LIFE AND
sonality disorders.
Urban Location Central to the definition of personality disor-
der are the interpersonal problems, reduced
The study of Torgersen et al. (2001) showed well-being, and dysfunction that personality
that those living in the populated center of disorders imply. In the sample studied by
the city more often had a personality disor- Torgersen et al. (2001), quality of life was as-
der. The same was true for all clusters of per- sessed by interview and included the follow-
sonality disorders and all specific disorders ing aspects: subjective well-being, self-
except antisocial, sadistic, avoidant, and de- realization, relation to friends, social support,
pendent personality. Grant et al. (2004) found negative life events, relation to family of ori-
this to be true only for paranoid and avoidant gin, and neighborhood quality (Cramer et al.
personality disorders, not antisocial, histri- 2003, 2006, 2007). All aspects were integrated
onic, schizoid, dependent, or obsessive-com- in a global quality-of-life index.
pulsive personality disorders. The two stud- Personality disorders turned out to be
ies agree that paranoid personality disorder, more strongly related to quality of life than
but not antisocial and dependent personality Axis I mental disorders, somatic health, and
disorders, are related to urbanicity. They dis- any other socioeconomic, demographic, or
agree about schizoid, histrionic, avoidant, life situation variable. Among the specific
and obsessive-compulsive personality disor- personality disorders, avoidant personality
ders, while the rest of the personality disor- disorder was most strongly related to quality
ders are not included in the study. of life, controlling for all the aforementioned
Given that more people with personality variables. Next came schizotypal, then para-
disorders are found in the center than in the noid, schizoid, borderline, dependent, anti-
outskirts of a city, one may speculate about social personality disorders, and narcissistic
the reason for this. Quality of life is generally and self-defeating personality disorders to a
lower in the center of the city (Cramer et al. lesser degree. Histrionic, obsessive-compul-
2004), and there is a higher rate of symptom sive, and passive-aggressive personality dis-
disorders in the city or in the center of the city orders were unrelated to quality of life. Some
(Kringlen et al. 2001; Lewis and Booth 1992, may be surprised that borderline personality
1994; Marcelis et al. 1998; Sundquist et al. disorder was not more strongly related to re-
2004; van Os et al. 2001). One reason may be duced quality of life. The reason for this is
that the concentrated urban life creates stress that the disorder is related to a number of
leading to personality disorders. Another other variables that are related to quality of
reason may be that individuals with person- life. Hence, the variables become weaker in a
ality problems drift to the center, where they multiple regression analysis.
can lead an anonymous life. A third explana- A dysfunction index was created by com-
tion may be that less social control simply bining quality of life (reversed); the answer
96 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

to the Structured Interview for DSM-III Per- ments for any specific number of criteria to
sonality DisordersRevised question do define a personality disorder if one uses
you feel that the way you usually deal with quality of life or dysfunction as validation
people and handle situations causes you variables. There is no natural cutoff point.
problems?; the number of lifetime Axis I di- A high level of dysfunction and disability
agnoses; and any incidence of seeking treat- was also observed among those with schizo-
ment with varying degrees of seriousness, typal personality disorder, followed by bor-
from private psychologist and psychiatrist derline and avoidant personality disorders,
via outpatient and inpatient clinicsto psy- in a large-scale multicenter study (Skodol et
chiatric hospitals. The dysfunction index was al. 2002). It was also observed that those with
related to personality disorder much as the obsessive-compulsive personality disorder
global quality-of-life index was. The only dif- showed much less disability.
ferences found in comparing results derived In another study, Ullrich et al. (2007),
from the dysfunction index with those from found that obsessive-compulsive personality
the global quality-of-life index were that disorder was not related to poor function-
those persons with borderline, dependent, or ingin fact, it was quite the opposite. Also
self-defeating personality disorder appeared histrionic personality disorder was positively
more dysfunctional, and those persons with related to status and wealth, while narcissis-
antisocial personality disorder appeared less tic and paranoid personality disorders were
dysfunctional. The reason for the differences unrelated to this index as well as to success-
is mainly that those with borderline, histri- ful intimate relationships. Taken together,
onic, dependent, and self-defeating person- those with schizoid personality disorder
ality disorders are more likely to seek treat- scored poorest on these two indexes, followed
ment and those with antisocial personality by antisocial, schizotypal, avoidant, border-
disorders are less likely to seek treatment. line, and dependent personality disorders.
However, the most important result in Zimmerman and Coryell (1989) also
this study was that for both quality of life and found a high frequency of psychosexual dys-
dysfunction, there was a perfect linear dose function among persons with avoidant per-
response relationship to numbers of criteria sonality disorder. Surprisingly, this dysfunc-
fulfilled for all personality disorders together tion was infrequent among persons with
and to the number of criteria fulfilled for any borderline personality disorder, and, not sur-
specific personality disorder. Thus, if a per- prisingly, it was also infrequent among those
son has one criterion fulfilled for one or an- with antisocial personality disorder.
other personality disorder, the quality of life Grant et al. (2004) applied a short form of
is lower and dysfunction is higher than quality of life assessment, the Short Form
among those with no criteria fulfilled. Those Health Survey, Version 2 (SF-12v2; Ware et al.
with two criteria fulfilled on one or more spe- 2002) and found that those with dependent
cific disorders have more problems than personality disorder had the poorest quality
those with one, those with three criteria have of life, followed by those with avoidant,
more problems than those with two, and so paranoid, schizoid, or antisocial personality
on. In other words, when those with zero disorders. There was no reduction in quality
criteria on all disorders were grouped to- of life for those with histrionic personality
getherthat is, those with a maximum of disorder, and a reduction on only one of three
one criterion on any disorder, those with a scores for those with obsessive-compulsive
maximum of two, and so onthe relation- personality disorder.
ship to global quality of life and dysfunction Crawford et al. (2005) studied impair-
was perfectly linear (Figures 51 and 52). ment using the Global Assessment of Func-
This result means that there are no argu- tioning (GAF) scale. Those with borderline
2.45 2.36

Prevalence, Sociodemographics, and Functional Impairment

2 1.76
Quality of life




0 1 2 3 4 5 6 79

Number of criteria

Figure 51. The relationship between maximum number of criteria fulfilled on any personality disorder and quality of life.
As explained in text, the ordinate (quality of life) is a composite of subjective well-being, self-realization, social support, negative life events, and relation to family,
friends and neighbors. The mean is set to 2 and the standard deviation is 1.


Level of dysfunction

2 1.73

E s s e n t i a l s o f P e r so n a l i t y D i s o r d e r s

1.5 1.25


0 1 2 3 4 5 6 7 9

Number of criteria

Figure 52. The relationship between maximum number of criteria fulfilled on any personality disorder and dysfunction.
As explained in the text, the ordinate (dysfunction) is a composite of life quality (reversed), treatment seeking, the number of lifetime Axis I diagnoses, and the
notion that ones behavior causes problems. The mean and standard deviation are 1.
Prevalence, Sociodemographics, and Functional Impairment 99

personality disorder had the poorest func- the percentage of the population with a disor-
tioning, followed by avoidant, schizotypal, der during the last 2 weeks, 1 month, 1 year,
narcissistic, antisocial, paranoid, histrionic, 2 years, or 5 years, the percentage will be lower
dependent, and schizoid personality disor- than if we follow a population throughout the
ders. Only obsessive-compulsive personality whole life span. This obvious fact has long
disorder had no indication of dysfunction. since been established for Axis I mental disor-
In conclusion, all studies taken together ders. The same will hold true for personality
show that reduced quality of life and dys- disorders, provided the disorders are not there
function are highest among those with already in early age and stay chronic through-
avoidant personality disorder, followed out life. While some theories may suggest this
closely by those with schizotypal or border- to be the case, empirical research shows that
line personality disorders. Those with para- many individuals are free of the personality
noid, schizoid, dependent, or antisocial per- disorder after a relatively short time (Grilo et
sonality disorder follow. There are few al. 2004; Zanarini et al. 2006). At the same time,
studies showing impaired quality of life for the 25 years point prevalence is much the
histrionic, narcissistic, or obsessive-compul- same over age, as we have seen. The implica-
sive personality disorder. The same is true tion is that new cases have to debut in the pop-
for the quality of provisional personality dis- ulation replacing those that disappear, even if
orders, passive-aggressive, self-defeating, some few reappear (Ferro et al. 1998; Durbin
and sadistic personality disorders. However, and Klein 2006). The consequence is that the
it should be noted that a problem with many percentage of the population that will fulfill
of these studies is a lack of control for other the criteria for a personality disorder in one or
variables correlating with specific personal- more periods of their life is much higher than
ity disorders and quality of life, dysfunction, 10%. This may be surprising for some, but it is
and impairment. an inevitable consequence of the continuous
There is reason to question whether his- nature of personality disorders and the appli-
trionic and obsessive-compulsive personal- cation of a threshold number of criteria to de-
ity disorders, in spite of their long histories, fine personality disorders. (Some with person-
deserve a status as personality disorder. Nar- ality disorders regress toward the population
cissistic personality disorder was not in- mean and do not any longer show the pre-
cluded in ICD-10, which some would view as scribed number of criteria. Other individuals
a wise decision. The provisional disorders with too few criteria display an increase and
provided for further study do not appear jump over and above the threshold.)
to deserve an inclusion in the nosology of In the future, longitudinal studies will
personality disorders. give us valuable insight into the epidemiol-
In the future, there is reason to believe ogy of personality disorders.
that we will see more studies of quality of
life, dysfunction, impairment, and disability
among subjects with personality disorders, CONCLUSION
either in the general population or in patient
samples. Personality disorders are prevalent: more
than 1 in 10 adult individuals has a personal-
ity disorder.
POINT PREVALENCE AND The average prevalence of the specific
LIFETIME PREVALENCE personality disorders is a little above 1%,
somewhat higher for obsessive-compulsive,
Lifetime prevalence for disorders is necessarily passive-aggressive, avoidant, histrionic, and
higher than point prevalence. If we measure antisocial personality disorders, and some-
100 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

what lower for sadistic, narcissistic, and self- tion, relational problems, lack of social sup-
defeating personality disorders. Two of these port, and frequent negative life events.
low-prevalence disorders, sadistic and self- Among the personality disorders, indi-
defeating, are only provided for further viduals with avoidant, schizotypal, and bor-
study in DSM-III-R (American Psychiatric derline personality disorders tend to have
Association 1987), and none of the three ex- the most reduced quality of life, dysfunction,
ists in ICD-10. and disability, followed by those with para-
Those with dependent, borderline, obses- noid, schizoid, dependent, and antisocial
sive-compulsive, avoidant, and schizotypal personality disorders. Individuals with ob-
personality disorders are overrepresented in sessive-compulsive, histrionic, narcissistic,
patient populations, both overall and when passive-aggressive, or sadistic personality
gender is controlled for, whereas those with disorders tend to have little or no reduction
antisocial, schizoid, and paranoid personal- in quality of life, dysfunction, and disability.
ity disorders are underrepresented. There is an even reduction in quality of
The highest prevalences of personality life and an even increase in dysfunction for
disorders in the general population are ob- each criterion manifested. Thus, there is a
served among subjects with lower education continuous relationship between those with
living in populated areas, for example in the no or small personality problems, those with
city center. They often have a history of di- moderate problems, and those with severe
vorce and separation and are more often liv- problems. No natural cutoff point exists. Any
ing without a partner. definition of how many criteria are required
Men have typically a schizoid, antisocial, for a personality disorder is arbitrary. Even
or obsessive-compulsive personality disor- so, to have a definition is important for com-
der, whereas women are more inclined to- munication. However, a change in criteria
ward a dependent or histrionic personality will immediately change the prevalence esti-
disorder. Antisocial, borderline, and passive- mates in the society. Consequently, correla-
aggressive personality disorders are more of- tions between personality disorders and
ten obse rv ed among younger persons, other variables are more important than
whereas older individuals more often have a prevalence rates. These correlations appear
schizoid personality disorder. It is typical for to be independent of how strictly personality
those with personality disorder (with an ex- disorders are defined.
ception for obsessive-compulsive personal- Because of the continuous nature of per-
ity disorder) to to live alone. In many cases, sonality disorders, their tendency to disap-
some are never married (antisocial, depen- pear, and the even distribution of point prev-
dent), have a history of frequent divorces alence over age, new personality disorders
(borderline), are divorced (paranoid), or not have to arise over the life span. Consequently,
married (histrionic) when interviewed. the likelihood of having a personality disor-
Lower education is most typical for those der once in the lifetime may be surprisingly
with paranoid and avoidant personality dis- high.
orders, whereas those with obsessive-com- Epidemiological research has perhaps
pulsive personality disorder in fact have changed some stereotypic notions about per-
higher education than those without the dis- sonality disorders. They are more frequent in
order. Those with paranoid, schizotypal, and the general population than we generally be-
passive-aggressive personality disorders are lieved, especially the introverted personality
most inclined to live in the city center. disorders. Borderline personality disorder is
Personality disorders imply dysfunction not a female disorder. Living without a
and reduction in quality of life, including re- partner is a risk factor for personality disor-
duced subjective well-being and self-realiza- ders, but being unmarried is less a risk factor
Prevalence, Sociodemographics, and Functional Impairment 101

than many would have believed. Those liv- ders in depressed outpatients. J Personal Dis-
ing in a partnership without being married ord 13:6774, 1999
Coid J, Yang M, Tyrer P, et al: Prevalences and cor-
function well.
relates of personality disorder. Br J Psychiatry
Care must be taken to avoid believing 188:423431, 2006
that these correlations display one-direc- Cramer V, Torgersen S, Kringlen E: Personality
tional causal relationships. Personality disor- disorders, prevalence, sociodemographic cor-
ders may hinder obtaining higher levels of relations, quality of life, dysfunction, and the
education and may create socioeconomic dif- question of continuity. Persnlichkeitsstrun-
gen. Theorie und Therapie 7:189198, 2003
ficulties. Problematic personality traits may Cramer V, Torgersen S, Kringlen E: Quality of life
prevent a person from going into a relation- in a city: the effect of population density. Soc
ship or may lead to the breaking-up of rela- Indic Res 69:103116, 2004
tionships, rather than having relationship is- Cramer V, Torgersen S, Kringlen E: Personality
sues and problems causing problematic disorders and quality of life. A population
study. Compr Psychiatry 47:178184, 2006
personality traits.
Cramer V, Torgersen S, Kringlen E: Socio-demo-
Personality disorders are not something graphic conditions, subjective somatic health,
that a person necessarily has for life. Impul- Axis I disorders and personality disorders in the
sive and aggressive features may decrease common population: the relationship to quality
quickly with age, whereas introverted traits of life. J Personal Disord 21:552567, 2007
Crawford TN, Cohen P, JohnsonJG, et al: Self-re-
increase as people get older.
ported personality disorder in the children in
Perhaps one of the most important as- the community sample: convergent and pro-
pects of personality disorders is the reduc- spective validity in late adolescence and
tion of quality of life that is implied. How- adulthood. J Personal Disord 19:3052, 2005
ever, a low quality of life does not necessarily Durbin EC, Klein DN: Ten-year stability of person-
ality disorders among outpatients with mood
create a personality disorder; the opposite is
disorders. J Abnorm Psychol 115:7584, 2006
just as likely. Ferro T, Klein DN, Schwartz JE, et al: Thirty-
month stability of personality disorder diag-
noses in depressed outpatients. Am J Psychia-
REFERENCES try 155:653659, 1998
Fossati A, Feeney JA, Donati D, et al: Personality
Alns R, Torgersen S: DSM-III symptom disor- disorders and adult attachment dimensions in
ders (Axis I) and personality disorders (Axis a mixed psychiatric sample: a multivariate
II) in an outpatient population. Acta Psychiatr study. J Nerv Ment Dis 191:3037, 2003
Scand 78:348355, 1988 Golomb M, Fava M, Abraham M, et al: Gender
American Psychiatric Association: Diagnostic and differences in personality disorders. Am J Psy-
Statistical Manual of Mental Disorders, 3rd chiatry 152:579582, 1995
Edition, Revised. Washington, DC, American Grant BF, Hasin DS, Stinson FR, et al: Prevalences,
Psychiatric Association, 1987 correlates, and disability of personality disor-
American Psychiatric Association: Diagnostic and ders in the United States: results from the Na-
Statistical Manual of Mental Disorders, 4th tional Epidemiologic Survey on Alcohol and
Edition. Washington, DC, American Psychiat- Related Conditions. J Clin Psychiatry 65:948
ric Association, 1994 958, 2004
American Psychiatric Association: Diagnostic and Grilo CM, Sanislow CA, Gunderson JG, et al: Two-
Statistical Manual of Mental Disorders, 4th year stability and change of schizotypal, bor-
Edition, Text Revision. Washington, DC, derline, avoidant, and obsessive-compulsive
American Psychiatric Association, 2000 personality disorders. J Consult Clin Psychol
Black DW, Noyes R Jr, Pfohl B, et al: Personality dis- 72:767775, 2004
order in obsessive-compulsive volunteers, well Jackson HP, Burgess PM: Personality disorders in
comparison subjects, and their first-degree rel- the community: a report from the Australian
atives. Am J Psychiatry 150:12261232, 1993 National Survey of Mental Health and Well
Carter JD, Joyce PR, Mulder RT, et al: Gender dif- Being. Soc Psychiatry Psychiatr Epidemiol
ferences in the frequency of personality disor- 35:531538, 2000
102 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

Klein DN, Riso LP, Donaldson SK, et al: Family schizotypal, borderline, avoidant, or obses-
study of early onset dysthymia: mood and sive-compulsive personality disorder. Am J
personality disorders in relatives of outpa- Psychiatry 159:276283, 2002
tients with dysthymia and episodic major de- Srivastava S, John OP, Gosling SD, et al: Develop-
pressive and normal controls. Arch Gen Psy- ment of personality in early and middle adult-
chiatry 52:487496, 1995 hood: set like plaster or persistent change?
Kringlen E, Torgersen S, Cramer V: A Norwegian J o u rn a l o f P ers on a l S o c ia l P sy c h o lo g y
psychiatric epidemiological study. Am J Psy- 84:10411053, 2003
chiatry 158:10911098, 2001 Stangl D, Pfohl B, Zimmerman M, et al: A struc-
Lenzenweger MF, Loranger AW, Korfine L, et al: tured interview for the DSM-III personality
Detecting personality disorders in a nonclini- disorders. Arch Gen Psychiatry 42:591596,
cal population: application of a 2-stage proce- 1985
dure for case identification. Arch Gen Psychi- Sundquist K, Frank G, Sundquist J: Urbanisation
atry 54:345351, 1997 and incidence of psychosis and depression. Br
Lenzenweger MF, Lane MC, Loranger AW, et al: J Psychiatry 184:293298, 2004
DSM-IV personality disorders in the national Torgersen S, Kringlen E, Cramer V: The preva-
comorbidity survey replication. Biol Psychia- lence of personality disorders in a community
try 62:553564, 2007 sample. Arch Gen Psychiatry 58:590596, 2001
Lewis G, Booth M: Regional differences in mental Ullrich S, Farrington DP, Coid JW: Dimensions of
health in Great Britain. J Epidemiol Commu- DSM-IV personality disorders and life-suc-
nity Health 46:608611, 1992 cess. J Personal Disord 21:657663, 2007
Lewis G, Booth M: Are cities bad for your mental van Os J, Hanssen M, Bijl RV, et al: Prevalence of
health? Psychol Med 24:913915, 1994 psychotic disorder and community level of
Maier W, Lichtermann D, Klingler T, et al: Preva- psychotic symptoms: an urban-rural compari-
lences of personality disorders (DSM-III-R) in son. Arch Gen Psychiatry 58:663668, 2001
the community. J Personal Disord 6:187196, Ware JE, Kosinski M, Turner-Bowker DM, et al:
1992 How to Score Version 2 of the SF-12 Health
Marcelis M, Navarro-Mateu F, Murray R: Urban- Survey. Lincoln, RI, Quality Metric, 2002
ization and psychosis: a study of 19421978 World Health Organization: The ICD-10 Classifica-
birth cohorts in the Netherlands. Psychol Med tion of Mental and Behavioural Disorders: Clin-
28:11971203, 1998 ical Description and Diagnostic Guidelines.
Moldin SO, Rice JP, Erlenmeyer-Kimling L, et al: Geneva, World Health Organization, 1992
Latent structure of DSM-III-R Axis II psycho- Zanarini MC, Frankenburg FR, Hennen J, et al:
pathology in a normal sample. J Abnorm Psy- Prediction of the 10-years course of borderline
chol 103:259266, 1994 personality disorder. Am J Psychiatry 163:827
Samuels J, Eaton WW, Bienvenu OJ III, et al: Preva- 832, 2006
lences and correlates of personality disorders in Zimmerman M, Coryell W: DSM-III personality
a community sample. Br J Psychiatry 180:536 disorder diagnoses in a nonpatient sample:
542, 2002 demographic correlates and comorbidity.
Saulsman LM, Page AC: The five-factor model Arch Gen Psychiatry 46:682689, 1989
and personality disorders empirical literature: Zimmerman M, Coryell WH: Diagnosing person-
a meta-analytic review. Clin Psychol Rev ality disorders in the community: a compari-
23:10551085, 2004 son of self-report and interview measures.
Skodol AE, Gunderson JG, McGlashan TH, et al: Arch Gen Psychiatry 47:527531, 1990
Function al impairment in patients with
Emil F. Coccaro, M.D.
Larry J. Siever, M.D.

The study of personality disorders involves degree to which personality (or temperamen-
the study of both disordered character and tal) traits are under genetic influence. This
disordered temperament. Character relates to work largely involves studies of families and
how we see and operate in our world and is twins and is designed to document familial, if
based on how we develop and what we are not genetic, components to behavior. Previ-
taught about how to go through life. Temper- ous work defining the genetic underpinnings
ament, in contrast, relates to our innate ten- of temperament has been critical to our cur-
dency to behave and to react to any of a vari- rent understanding that temperament is in-
ety of challenges presented by other people herently biological in nature. Neuropsycho-
and our environment. Although both aspects pharmacological study informs us about the
of personality may be studied empirically, nature of brain chemistry and how the regula-
the study of temperament is uniquely suited tion of any of a variety of brain neurotrans-
to biological study because temperament has mitters influences temperament. Work in this
known genetic and neurobiological corre- area has led to the understanding that brain
lates, both of which are linked to critical pro- serotonin, for example, is critical in modulat-
cesses involving cognition, emotion, and be- ing impulsive aggressive behavior in individ-
havior. uals with personality disorder. Consequently,
The neurobiology of temperament, as it work in neuropsychopharmacology leads to
appears in personality disorders, can be work in molecular genetics whereby the pres-
studied in a variety of ways, including those ence of a specific copy of a specific gene (e.g.,
that involve behavioral genetics, neuropsy- for a component of the brain serotonin sys-
chopharmacology, molecular genetics, psy- tem) influences a temperamental trait. For ex-
chophysiology, and neuroimaging. Behav- ample, individuals carrying a specific gene
ioral-genetic study informs us about the for the serotonin transporter may be more

104 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

anxious than other individuals who do not lustrating some of the points made about the
carry this gene. Finally, work in psychophys- psychobiology of prototypical patients with
iology and neuroimaging brings investiga- selected personality disorders.
tive work up to a level that integrates genes,
neuropsychopharmacology, and networks of
neural transmission. In this methodology,
both brain structure and brain function are
examined regarding their contribution to the DISORDERS
expression of various temperamental traits.
In some groups of patients with personality The Cluster A personality disorders include
disorder, neuroimaging has revealed differ- schizotypal, paranoid, and schizoid person-
ences in the size and function of specific ality disorder. The criteria of these disorders
structures. capture shared characteristics of social isola-
Ultimately, the study of the neurobiology tion, detachment, suspiciousness, and in the
of personality disorders is conducted to lead case of schizotypal personality disorder, psy-
to a more comprehensive understanding of chotic-like cognitive/perceptional distortion.
the biological substrates of personality disor- Schizotypal personality was formulated in
der so that better treatments may be discov- part on the clinical profile observed in rela-
ered and existing treatments may be im- tives of schizophrenic probands, whereas the
proved. Uncovering the biological substrate other two were defined more in a clinical tra-
for a specific temperamental trait naturally dition. A high degree of overlap exists be-
leads to treatment strategies aimed at this spe- tween schizotypal and paranoid personality
cific substrate. The best example of this ap- disorder, whereas schizoid personality disor-
proach is the use of serotonin uptake inhibi- der is not frequently diagnosed in the clinical
tors in the treatment of impulsive aggression setting and may represent a milder version of
in individuals with personality disorders. Cu- the Cluster A personality disorders. These
riously, work in this area revealed the likely disorders can be perceived as consisting of a
presence of two treatment response groups: dimension of social deficits (no friends, de-
one responsive to serotonin uptake inhibitors, tached affect) and cognitive impairment, and
the other responsive to mood stabilizers. in the case of schizotypal personality disor-
In this chapter, we discuss the various as- der, a psychotic-like dimension. Because of
pects of the neurobiology of personality dis- its relationship to schizophrenia and its more
order on a cluster-by-cluster basis. We have common prevalence in clinical populations,
chosen this organization because the proto- most of the neurobiological research on this
typical personality disorders of interest tend cluster has focused on schizotypal personal-
to break out into one of the three personality ity disorder and is summarized here in rela-
disorder clusters. Despite this type of organi- tion to these dimensions.
zation, we should note that research has
clearly shown the relevance of a dimensional Behavioral Genetics
approach to the study of personality. Each
section begins with a brief summary of the Schizotypal personality disorder is found
phenomenology characteristic of each per- more frequently in the relatives of schizo-
sonality disorder cluster and follows with a phrenic probands than in the relatives of
summary of data relevant to behavioral ge- control subjects, and this association is
netics, neuropsychopharmacology (and mo- grounded in genetics rather than shared fa-
lecular genetics where relevant), and neu- milial environment as suggested by adoptive
ropsychology and neuroimaging. Each of the and twin studies (Siever 1991). The genetics
first two sections ends with a brief vignette il- of paranoid personality disorder are less well
Neurobiology 105

understood, but it has a high overlap with schizotypal personality disorder. Interest-
schizotypal personality disorder, and its ingly, however, when the negative symp-
presence may be greater in families of pa- toms were entered as a covariant, the positive
tients with schizophrenia or delusional disor- relationship with psychotic-like symptoms in
der (Webb and Levinson 1993). Schizoid per- plasma HVA emerged (Amin et al. 1997). Re-
sonality disorder has received little or no duced plasma HVA concentrations have
genetic study but is more common in the rel- been associated with impairment in tests of
atives of patients with schizophrenia (Kalus frontally mediated executive function such as
et al. 1993). the Wisconsin Card Sort Test (Siever et al.
1991). Thus, these results suggest that
Neuropsychopharmacology: dopaminergic activity may be relatively in-
Dopamine System creased or decreased depending on the pre-
dominance of psychotic-like versus deficit-
The dopamine system has been extensively like symptoms, respectively. This distinction
studied in patients with schizophrenia and is consistent with formulations that increased
particularly associated with the psychotic dopaminergic activity, particularly in stria-
symptoms of this disorder, consistent with tum, is associated with psychotic-like symp-
the antipsychotic effects of the neuroleptics, toms and that decreased dopaminergic ac-
which act as dopamine antagonists. Accord- tivity, particularly in prefrontal regions, is
ingly, given the phenomenological and ge- especially associated with deficit-like symp-
netic relationships between schizophrenia toms (Siever and Davis 2004).
and schizotypal personality disorder, the
dopaminergic system has been the primary Acute Pharmacological Interventions
neurotransmitter system studied in schizo-
typal personality disorder. Amphetamine, which stimulates the release
of the monoamines, particularly dopamine
Neurochemistry and norepinephrine, has been shown to im-
prove the cognitive performance of schizo-
Plasma homovanillic acid (HVA), a major typal personality disorder subjects on tests of
metabolite of dopamine, has been found to be executive function, working memory, and to
elevated in clinically selected patients with a lesser extent, sustained attention and ver-
schizotypal personality disorder, and this el- bal learning (Kirrane et al. 2000; Siegel et al.
evation is significantly correlated with psy- 1996). These improvements are more consis-
chotic-like criteria for this disorder, such that tent than those observed in schizophrenic
statistical correction for the presence of psy- subjects given amphetamine and are not ac-
chotic-like symptoms abolishes the differ- companied by the behavioral worsening
ence between groups (Siever et al. 1991). An that is, increased psychotic symptoms
identical configuration of results is found found after amphetamine administration in
with respect to cerebrospinal fluid (CSF) schizophrenic patients. Indeed, the deficit-
HVA (Siever et al. 1993). On the other hand, like symptoms of schizotypal personality
among relatives of patients with schizophre- disorder tend to improve following amphet-
nia, who are generally characterized more by amine administration (Laruelle et al. 2002;
the social and cognitive deficit-like symp- Siegel et al. 1996). These results suggest that
toms of schizotypal personality disorder, agents that enhance catecholamines, includ-
plasma HVA is lower in subjects with schizo- ing dopamine, may have beneficial effects on
typal personality disorder than control sub- cognition, presumably through stimulation
jects (Amin et al. 1999). In these studies, of D1 receptors in prefrontal cortex.
plasma HVA was negatively correlated with Similarly, the administration of a glu-
the negative or deficit-like symptoms of copyruvic stressor, 2-deoxyglucose, which
106 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

activates stress-sensitive subcortical systems phism, the allele for the COMT gene codes
such as the dopamine system and the hypo- for the amino acid valine (Val), as opposed to
thalamic-pituitary-adrenal (HPA) axis, re- methionine (Met), in the COMT enzyme. The
sults in greater stress-related (i.e., plasma substitution of Val for Met leads to a COMT
cortisol and HVA) responses in patients with enzyme that has far more activity than a
schizophrenia than in control subjects. In COMT enzyme coded by the MET allele.
contrast, patients with schizotypal personal- Thus, individuals with VAL alleles should
ity disorder show normal (plasma HVA) or have increased activity of COMT compared
even reduced (cortisol) activation compared with those with the MET allele. Because in-
with control subjects, suggesting that pa- creased COMT activity is associated with in-
tients with schizotypal personality disorder creased destruction of catecholamines, indi-
have better-buffered subcortical stress-re- viduals with VAL alleles should have less
sponsive systems than patients with schizo- central dopamine activity than those with
phrenia. Consequently, it is possible that this MET alleles. Consistent with this idea, cogni-
buffer provides a protective factor against tive impairment, particularly evident in
psychosis in patients with schizotypal per- dopamine-dependent working memory, has
sonality disorder (Siever and Davis 2004). been associated with the presence of the VAL
Longer-term pharmacological interven- allele in patients with schizophrenia (Wein-
tions have been evaluated in individuals berger et al. 2001) as well as their healthy sib-
with schizotypal personality disorder to de- lings and control subjects (Goldberg et al.
termine their effects on cognitive function. 2003). Studies in patients with schizotypal
Studies of guanfacine, an 2-adrenergic ago- personality disorder also suggest an associa-
nist, and pergolide, a D1/D2 agonist, suggest tion between cognitive impairment and the
improvement in cognitive function, particu- VAL allele, consistent with the role of re-
larly working memory, with these catechola- duced dopaminergic activity hypothesized
minergic interventions, consistent with the to contribute to the cognitive dysfunction in
facilitatory effects of the catecholamines on the schizophrenia spectrum disorders such
cognitive function and prefrontal cortex (Mc- as schizotypal personality disorder (Minzen-
Clure et al. 2007 ) . Cognitive function may berg et al. 2006).
also improve with risperidone (Koenigsberg
et al. 2003), possibly due to the effects that 5- Cognitive Function and
hydroxytriptamine (serotonin) type 2 recep- Psychophysiology
tor (5-HT 2 ) blockade has on facilitating
dopaminergic activity in frontal lobe. Anti- Although cognitive dysfunction may exist in
psychotic effects have been documented in a subtle forms in a variety of personality disor-
number of clinical trials of atypical and typi- ders, the most consistent and robust changes
cal neuroleptics in individuals with schizo- are found in people with Cluster A personal-
typal personality disorder (Hymowitz et al. ity disorders, more specifically schizotypal
1986; Schulz et al. 2003). personality disorder. Patients with schizo-
typal personality disorder show attenuated
DNA Polymorphisms patterns of cognitive impairment similar to
those of patients with schizophrenia but
Catechol-O-methyltransferase (COMT) somewhat more specific. For example, over-
plays a critical role in inactivation of dopa- all intelligence may not be impaired (Mitro-
mine in the frontal lobe, where the dopamine poulou et al. 2002; Trestman et al. 1995),
transporter is not the primary mode of inac- whereas specific disturbances in sustained
tivation of dopamine. Recently, a single nu- attention, in verbal learning, and particularly
cleotide polymorphism in the COMT gene in working memory have been reported in
has been discovered. With this polymor- patients with schizotypal personality disor-
Neurobiology 107

der compared with patients with other non- thermore, they show less capacity for inhibi-
schizophrenia-related personality disorders, tion on a prepulse inhibition paradigm and
the latter of whom are generally not im- P50-evoked potential paradigm (Cadenhead
paired in these indices, and with normal con- 2002). The latter finding is of particular inter-
trol subjects (Mitropoulou et al. 2002, 2005). est because it has been linked to a specific al-
Alth oug h patie nts with schizophrenia lele of the nicotinic receptor in families of
showed deviations from normal control sub- patients with schizophrenia. Backward mask-
jects on the order of two standard deviations, ing, reflecting early visual processing, has
patients with schizotypal personality disor- also been reported to be abnormal in patients
der have more on the order of one standard with schizotypal personality disorder and
deviation below the mean or less (Mitropou- schizophrenia (Siever and Davis 2004; see
lou et al. 2002, 2005). The deficits in working Braff and Freedman [2002] for an overview of
memory and attention may contribute to the these psychophysiological abnormalities).
impaired rapport and misreading of verbal
and facial cues in patients with schizotypal
personality disorder, who often clinically
complain that they have a hard time focusing Structural Imaging
on others, which detracts from their ability to
engage. Indeed, performances on working Patients with schizotypal personality disor-
memory tasks have been reported to be cor- ders show ventricular enlargement and re-
related with interpersonal impairment and duced volumes of several brain regions, as
account for the other cognitive impairments do patients with schizophrenia. In studies of
in schizotypal personality disorder (Mitro- patients with schizotypal personality disor-
poulou et al. 2005; Siever et al. 2002). der, ventricular volume is increased, al-
A variety of psychophysiological en- though studies of relatives of patients with
dophenotypes that may reflect genetic sub- schizotypal personality disorder are mixed
strates to the schizophrenia spectrum disor- (Shihabuddin et al. 1996; Siever 1995). Tem-
ders have been found to be abnormal in poral volume reductions in patients with
patients with schizotypal personality disor- schizotypal personality disorder appear to
der as well as in patients with chronic schizo- be comparable with th ose observ ed in
phrenia. Many of these psychophysiological schizophrenic patients and occur in both su-
abnormalities have also been found in rela- perior temporal gyrus and other temporal re-
tives of patients with schizophrenia, who gions. However, some data suggest that
may have mild schizophrenia-spectrum frontal volumes are relatively preserved, es-
symptoms or may even appear to be clini- pecially frontal pole (Brodmann area 10
cally healthy, raising the possibility that [BA10]), suggesting that greater frontal ca-
these abnormalities reflect an underlying ge- pacity may serve as a buffer against the se-
netic susceptibility to the schizophrenia vere cognitive and social deterioration we
spectrum that is variably expressed. Al- see in schizophrenia (Hazlett et al. 2008).
though a detailed review of psychophysio- Whereas striatal volumes of patients with
logical abnormalities is beyond the scope of schizophrenia are enlarged secondary (in
this chapter, abnormalities in eye movement, large part) to neuroleptic medications, the
visual processing, and inhibition of startle re- striatal volumes (including putamen [Shi-
sponse are among the most consistently rep- habuddin et al. 2001] and caudate [Levitt et
licated. Thus, individuals with schizotypal al. 2002]) of patients with schizotypal per-
personality d isorder showed impaired sonality disorder are reduced in comparison
smooth-pursuit eye movement, antisaccade with normal control subjects and unmedi-
generation, and velocity discrimination. Fur- cated patients with schizophrenia. Reduced
108 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

striatal volumes are consistent with the pos- striatum, which is normally inhibited by
sibility of reduced dopaminergic activity, dopamine, in unmedicated schizotypal pa-
which may be protective against the emer- tients compared with control subjects and
gence of psychosis. unmedicated schizophrenic patients as well
as the reduced plasma HVA responses to 2-
Functional Imaging deoxyglucose and striatal volumes noted
earlier in these studies, suggesting dopamin-
Both positron emission tomography (PET) ergic activity that is better buffered than that
and single photon emission computed to- of schizophrenic patients.
mography (SPECT) functional imaging stud-
ies suggest that patients with schizotypal per- Case Example
sonality disorder do not activate regions such
as dorsolateral prefrontal cortex in response Mr. B is a 56-year-old, common-law
to an executive function or learning task to the married male, emplo yed in his ex-
same degree as control subjects, but do so to a tended familys business, whose cur-
rent complaint is that people at work
greater degree than do patients with schizo-
are accusing me of saying things that I
phrenia. However, patients with schizotypal am not saying. Mr. B has been seen by
personality disorder are able to activate other the psychiatry service for more than 20
compensatory regions, including the anterior years, after he was admitted to medi-
pole of frontal cortex (BA 10), which is be- cine for complaints of back pain. He
lieved to be a high-level executive region was transferred to psychiatry because
(Buchsbaum et al. 2002). A recent functional he couldnt stand up. He was first
psychiatrically hospitalized when he
magnetic resonance imaging (fMRI) study
was in the Navy for an episode of go-
(Koenigsberg et al. 2005) using a visuospatial ing crazy after a dispute with his cap-
working-memory task also showed increased tain. He had symptoms of deperson-
activation in patients with schizotypal per- alization, irritability, and difficulty
sonality disorder in BA 10, but lesser activa- getting along with his peers. Six years
tion in dorsolateral prefrontal cortex than in after his tour in the Navy, he saw a ther-
normal control subjects. Thus, patients with apist but would have vivid dreams that
were disturbing to both the therapist
schizotypal personality disorder may have
and himself, at which point Mr. B states
compensatory mechanisms available to them
his therapy ended. He has had paranoid
that patients with schizophrenia do not have ideation, thinking that people at work
in the face of diminished capacity to use dor- are against him, although this suspi-
solateral prefrontal cortex. The compensatory ciousness and ideation are responsive
mechanisms may involve using higher execu- to reality testing, as are his ideas of ref-
tive regions than are required for normal in- erence. He has prolonged periods of an-
dividuals. hedonia and demoralization but, other
than insomnia at times, does not have
A n I B Z M SP E C T st u d y m e as ur in g
extensive vegetative symptoms of de-
dopamine released by displacement of [11C] pression. He has experienced episodes
iodine-methoxybenzamide (IBZM) demon- of depersonalization described as look-
strated that subjects with schizotypal per- ing down at himself. He complains of
sonality disorder released significantly more low self-esteem but denies worthless-
dopamine in response to amphetamine ad- ness, hopelessness, or helplessness. Mr.
ministration than did normal control sub- B notes that he was always a loner and
had no close friends since the fifth
jects but less than did acute schizophrenic
grade. He went to college just before he
patients (Siever et al. 2002). These results are went to Vietnam. He smokes one pack
consistent with functional imaging studies of cigarettes per day, does not use recre-
suggesting increased activation of ventral ational drugs, and drinks up to three
Neurobiology 109

drinks per night, although he goes for threshold of overt psychosis. For these rea-
periods without drinking significantly. sons, they present more in the context of their
Mr. B underwent a research evalua- disturbed interpersonal style and coping
tion in the Mood and Personality Dis-
mechanisms rather than in the context of
orders Program. Research diagnostic
evaluation revealed the presence of a overt psychosis as in schizophrenia. How-
schizotypal personality disorder with ever, this disorder provides an example of a
traits of paranoid and narcissistic per- spectrum that in its more extreme forms
sonality disorder; he was also found to manifests as an Axis I disorder (schizophre-
meet DSM-IV-TR (American Psychiatric
nia) but in milder forms as an Axis II disor-
Association 2000) criteria for alcohol
abuse (past). Neurobiological evalua- der. There are few biological data regarding
tion uncovered a number of abnormali- paranoid personality disorder when it is not
ties. First, he displayed modestly im- comorbid with schizotypal personality dis-
paired eye-movement accuracy (3.38 on order.
a 1=best to 5=worst scale) and mild cog-
nitive impairment. His dopaminergic in-
dices were high, with a plasma HVA
level of 14.5 ng/mL (mean for normal CLUSTER B PERSONALITY
subjects is 7.4 1.8 ng/mL) and a CSF
HVA level of 38.0 ng/mL (mean for nor- DISORDERS
mal control subjects = 24.1 6 ng/mL).
In addition, Mr. B showed hypofrontal- The Cluster B personality disorders include
ity on a PET scan during a verbal mem- antisocial, borderline, histrionic, and narcis-
ory task. Finally, he showed modest im- sistic personality disorders. Individuals with
provement following administration of
these disorders present with varied degrees
amphetamine. Since evaluation, Mr. B
has been treated with low-dose neuro- of impulsivity, aggression, and emotional
leptic medication that helps him control dysregulation. As in other clusters, there is a
multiple symptoms, including an olfac- high degree of overlap among the disorders
tory hallucination-like experience of the in Cluster B, particularly between antisocial
smell of cordite, a feeling that others personality disorder (ASPD) and borderline
are staring at him, a feeling of being de-
tached or separated by a bubble from personality disorder (BPD). ASPD and BPD
other people, a lack of any close friends are the best studied of the cluster, due to
other than a common-law wife, feelings clear and reliable criteria for the former and
in the past that his wife might be fol- the high prevalence of the latter in clinical
lowing around, and a feeling in the past populations.
that he has seen future events.

Behavioral Genetics
Twin studies suggest that the genetic influ-
These studies suggest that patients with ence underlying personality disorders is at
schizotypal personality disorder have at least as high as that of personality traits that
least a profile of cognitive impairment and underlie the various personality disorders.
structural brain abnormalities, particularly In a relatively small twin study (Torgersen et
in temporal cortex, similar to that found in al. 2000) that may tend to overestimate the
patients with schizophrenia, but a combina- underlying genetic influence of any of a vari-
tion of better prefrontal reserves and more ety of personality disorders, the heritability
subdued dopaminergic activity subcortically for Cluster B personality disorders was 0.60.
protects them from the emergence of psycho- The heritabilities of the specific Cluster B dis-
sis. Their more subtle cognitive impairments orders in this study were 0.79 for narcissistic
are reflected in their eccentricity and inter- personality disorder, 0.69 for BPD, and 0.67
personal disengagement but do not reach the for histrionic personality disorder. The best-
110 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

fitting models did not include shared famil- disorder diagnoses (Brown et al. 1979). This
ial environment effects, although such effects finding was extended (Brown et al. 1982) to
may influence the development of BPD. include a trivariate relationship between his-
Adoption studies of ASPD confirm a strong tory of aggression, suicide attempts, and
genetic, although a less strong environmen- reduced CSF 5-HIAA, whereby history of
tal, influence for this disorder (Cadoret et al. aggression and suicide attempts were corre-
1985). Although adoption studies of other lated directly with each other and inversely
Cluster B personality disorders have not with CSF 5-HIAA. Later work with violent
been conducted, the results of family history offenders (Linnoila et al. 1983) found re-
studies suggest a complex pattern of familial duced CSF 5-HIAA in impulsive, but not
aggregation in which traits related to impul- nonimpulsive, violent offenders with a vari-
sive aggressiveness and mood dysregulation, ety of DSM-II personality disorder diag-
rather than BPD itself, are transmitted in noses, suggesting that impulsive aggression
families (Silverman et al. 1991). was the form most associated with reduced
CSF 5-HIAA concentration. Although these
Neuropsychopharmacology findings have been replicated, an inverse re-
lationship between CSF 5-HIAA and aggres-
The 5-HT system has been extensively stud- sion has not been reported in samples of in-
ied in individuals with personality disorder dividuals with personality disorder without
in general and in particular as an inverse cor- a prominent history of criminal activity (Coc-
relate of impulsive aggressive behavior. caro et al. 1997a, 1997b; Gardner et al. 1990;
Other neurotransmitters and/or modulators Simeon et al. 1992). It is likely that CSF 5-
have also been studied in this regard, but to a HIAA, being a relatively insensitive index of
much lesser degree. 5-HT activity, is most reduced in the most se-
verely aggressive individuals and that it is
Serotonin difficult to detect this relationship in less se-
verely aggressive individuals.
There is a clear and consistent role for 5-HT
in the regulation of aggression and/or im- Acute pharmacological interventions. T h e r e
pulsivity, particularly in individuals with are a variety of 5-HT acute pharmacological
personality disorder. Most data suggest an challenge studies that have been performed
inverse relationship between any of a variety in individuals with personality disorder in
of measures of 5-HT levels and levels of ag- the context of the study of aggression. Typi-
gression or impulsivity. Although some cally, hormonal (e.g., prolactin) responses to
studies suggest a primary relationship with the 5-HT selective agents are reported to cor-
impulsivity, most studies report a 5-HT rela- relate inversely with various measures of ag-
tionship more consistent with the construct gression and impulsivity (Coccaro et al. 1989,
of impulsive aggression. 1997a, 1997b; Dolan et al. 2001; Moss et al.
1990; OKeane et al. 1992; Paris et al. 2004;
Neurochemical studies. Inverse relationships Siever and Trestman 1993). Pharmacological
between human aggression and measures of challenge studies using putatively receptor-
central 5-HT function have been reported selective 5-HT agents also seem to support
since 1979, when Brown and colleagues re- the hypothesis of an inverse relationship be-
ported an inverse relationship between CSF tween 5-HT and measures of aggression and
levels of the main central 5-HT metabolite, suggest a role for at least the 5-HT1A receptor
5-hydroxyindoleacetic acid (5-HIAA), and in particular (Cleare and Bond 2000; Coccaro
life history of actual aggressive behavior in et al. 1990, 1995; Hansenne et al. 2002). A
males with a variety of DSM-II (American more complex picture in regard to central 5-
Psychiatric Association 1968) personality HT1A receptors has been suggested by the
Neurobiology 111

observation of reduced 5-HT 1A receptor on the genotyping gel) was found to have
mediated responses in females with BPD some association with clinically relevant
with a history of sustained child abuse variables. For example, impulsive violent of-
(Rinne et al. 2000). Because childhood abuse fenders (nearly all with a personality disor-
has been linked to impulsive aggression in der) with at least one copy of the L TPH allele
later adolescence and adulthood (Crick and have been reported to have significantly
Dodge 1996), it remains to be determined lower CSF 5-HIAA compared with impul-
whether the relationships between 5-HT and sive violent offenders with the UU genotype
aggression are linked to this environmental/ (U referred to the upper band on the geno-
developmental variable. Although behav- typing gel) in at least one study (Nielson et
ioral responses to 5-HT stimulation in indi- al. 1994). This finding did not generalize to
viduals with personality disorder have not nonimpulsive violent offenders (many of
received much attention, at least one study whom also had a personality disorder) or to
reported a significant reduction in anger in normal control subjects and was not repli-
12 patients with BPD after administration of cated in a later study by the same authors
the mixed 5-HT agonist m-chlorophenylpip- (Nielson et al. 1998).
erazine (m-CPP) but not placebo (Hollander The presence of the L allele was associ-
et al. 1994); a reduction in fear was also ob- ated with an increased risk of suicidal behav-
served in the males with BPD. ior in all violent offenders in this and in a
later study by these authors (Nielson et al.
Platelet receptor markers. Despite consider-
1994, 1998). New et al. (1998) have also re-
able platelet receptor work in other psychiat-
ported that the self-reported tendency to-
ric populations, relatively little research in
ward aggression varies as a function of TPH
this area has been published on subjects with
genotype whereby subjects with the LL gen-
personality disorder. Inverse correlations be-
otype had higher aggression scores than
tween the number of platelet 3H-imipramine
those with the UU genotype. Curiously,
(5-HT transporter) binding sites and self-
however, the reverse finding was reported
mutilation and impulsivity have been re-
by Manuck et al. (1999) in a sample of healthy
ported in individuals with personality disor-
volunteers from the community: higher ag-
der but not in patients without a history of
gression scores were associated with the
self-mutilation (Simeon et al. 1992). Similarly,
presence of the U allele.
an inverse correlation between the number of
These disparate findings may be due to
platelet 3 H-paroxetine (5-HT transporter)
critical differences in the subject samples. As
binding sites (Coccaro et al. 1996), the quan-
such, the relationship between the TPH allele
tity of platelet serotonin (Goveas et al. 2004),
and 5-HT function may be dependent on the
and life history of aggression has been re-
TPH alleles relationship with some other
ported in persons with personality disorder.
gene depending on the subject sample. Lap-
DNA polymorphism studies. W o r k i n t h i s palainen et al. (1998) reported an association
area began with an examination of DNA between antisocial alcoholism (i.e., alco-
polymorphisms in the gene for tryptophan holism with ASPD or intermittent explosive
hydroxylase (TPH). TPH is the rate-limiting disorder) and the C allele for the 5-HT 1D
step for the synthesis of serotonin, and it was beta-receptor polymorphism. Because the
thought that polymorphisms in TPH would 5-HT1D beta receptor is a critical receptor in-
lead to TPH enzymes of different activities. volved in the regulation of 5-HT release on
Although this TPH polymorphism was not neuronal impulse, this finding could be
found to have a clear functional consequence highly relevant to the understanding of
regarding serotonin synthesis, the presence ASPD comorbid with alcoholism. Alleles of
of the L allele (L referred to the lower band the 5HTT transporter, 5HT 2A, and TPH2
112 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

have been found to be associated with BPD concentrations, it is possible that findings
or its traits (Ni et al. 2006a, 2006b, 2007; with CSF HVA may be related to similar find-
Siever et al. 2006). ings with CSF 5-HIAA concentration. If so, a
specific assessment of CSF HVA may not be
Catecholamines made unless the effect of CSF 5-HIAA con-
centration is accounted for, a statistical ad-
Compared with serotonin, far fewer data justment that has not been made in pub-
have been published regarding the role of lished studies to date.
other neurotransmitters and behavioral di-
Acute pharmacological interventions. E a r l y
mensions of relevance to the Cluster B per-
studies of the acute administration of am-
sonality disorders.
phetamine in patients with BPD demon-
Neurochemical studies. A positive correlation strated a greater behavioral sensitivity to am-
between CSF 3-methoxy-4-hydroxy-phenyl- phetamine challenge among the patients with
clycol (MHPG, the major metabolite of nore- personality disorder than among control sub-
pinephrine) concentrations and life history of jects (Schulz et al. 1985). Replication studies
aggression has been reported in males with found that global worsening in psychopathol-
personality disorder, although further analy- ogy after amphetamine was typical of pa-
sis revealed that CSF 5-HIAA concentration tients with both borderline and schizotypal
accounted for most (80%) of the variance in personality disorder, whereas global im-
aggression scores. Similarly, one study re- provement was typical of borderline subjects
ported a small positive correlation between without comorbid schizotypal personality
plasma norepinephrine and self-reported im- disorder (Schulz et al. 1988). This finding sug-
pulsivity in males with personality disorder gests important biological differences among
(Siever and Trestman 1993). In contrast, at patients with BPD as a function of comorbid
least one study (Virkkunen et al. 1987) re- schizotypy (perhaps because of preexisting
ported a significant reduction in CSF MHPG dopaminergic hyperactivity in mesolimbic
concentration in males who have committed dopamine circuits). In other studies of am-
violent offenses. Finally, Coccaro et al. (2003) phetamine challenge relevant to Cluster B
reported an inverse relationship between personality disorder, a direct relationship
plasma-free MHPG and life history of aggres- with affective lability has been noted in
sion in males with personality disorder. healthy volunteers, suggesting that increases
Compared with patients with nonborderline in norepinephrine and/or dopamine may
personality disorders, patients with BPD had play a role in the moment-to-moment dysreg-
lower plasma free MHPG compared with the ulation of affect seen in patients with BPD
nonborderline control subjects; a finding that (Kavoussi et al. 1993).
disappeared after differences in aggression Limited data are available regarding the
scores were accounted for. study of norepinephrine receptormediated
Evidence for the role of dopamine in ag- responses related to the features of Cluster B
gression in individuals with personality dis- personality disorder. One study reported a
order is limited and contradictory. Although positive correlation between the growth hor-
some studies demonstrate no relationship mone response to the 2 norepinephrine ago-
between CSF HVA concentration and aggres- nist clonidine and self-reported irritability
sion (Brown et al. 1979; Virkkunen et al. (a correlate of aggression) in a small sample of
1987), other studies demonstrate an inverse males with personality disorder and healthy
relationship between these variables (Lin- volunteers (Coccaro et al. 1991). A more re-
noila et al. 1983; Virkkunen et al. 1989). Given cent study of females with BPD, however, re-
the consistent observation of a strong corre- ported no difference in growth hormone re-
lation between CSF 5-HIAA and CSF HVA sponses to clonidine (Paris et al. 2004).
Neurobiology 113

DNA polymorphism studies. The presence of Other neurotransmitters or neuromodu-

the low-functioning monoamine oxidase A lators that may play a role in Cluster B
(MAO-A) allele in young men combined related features include vasopressin, which
with a history of childhood maltreatment has may have a direct relationship with aggres-
recently been shown to be associated with an sion (Coccaro et al. 1998); substances related
increased risk of aggressive and criminal of- to limbic HPA axis functioning (corticotropin
fending (e.g., antisocial) behavior (Caspi et releasing factor, adrenocorticotropic hor-
al. 2002). This specific MAO-A allele is asso- mone, cortisol), which may have varied rela-
ciated with reduced catabolism of catechola- tionships regarding aggressive behavior de-
mines (and serotonin) and accordingly with pendent on social context and stress (Rinne
higher levels of these neurotransmitters that et al. 2002); testosterone, which is variably
may be associated with aggressive behavior. correlated with aggression, particularly in vi-
These data suggest that although the pres- olent offenders with ASPD (Virkkunen et al.
ence of this allele may be important in in- 1994); and cholesterol and fatty acids, which
creasing the risk of antisocial behavior, the may play a role in both aggression (both: At-
co-occurrence of childhood maltreatment in maca et al. 2002; New et al. 1999) and mood
vulnerable individuals is also needed to regulation (fatty acids: Zanarini and Fran-
meaningfully increase the risk of antisocial kenburg 2003).
Acetylcholine and Other
Neurotransmitters/Neuromodulators Structural Imaging
Reduced prefrontal gray matter (e.g., by 11%)
Studies of acetylcholine function in personal- has been associated with autonomic deficits
ity disorder have been limited to two studies. in individuals with ASPD characterized by
In the first (Steinberg et al. 1997), patients aggressive behaviors (Raine et al. 2000). Con-
with BPD reported greater self-rated depres- versely, increases in corpus callosum white
sion scores in response to the cholinomimetic matter volume and length have been de-
agent physostigmine than did patients with scribed in similar subjects (Raine et al. 2003),
nonborderline personality disorders or where larger callosal volumes were also asso-
healthy volunteer control subjects. Peak phy- ciated with affective/interpersonal deficit,
sostigmine-induced depression scores corre- low autonomic stress reactivity, and spatial
lated positively with the number of affective ability. Given the complex role these struc-
instability, but not with the number of im- tures play in mediating cognitive and affec-
pulsive aggression, borderline personality tive processes, these findings may represent
traits. This finding suggests that the trait of anatomical correlates of the complex behav-
affective lability in patients with BPD may be iors seen in ASPD. A confounding role for al-
mediated in part by a heightened sensitivity coholism in these matters must always be ad-
to acetylcholine. In the second study (Paris et dressed, however, because it also has been
al. 2004), however, no differences in hor- shown that volume changes may be corre-
monal responses to a different cholinomi- lated with duration of alcoholism (Laakso et
metic agent, pyridostigmine, were seen be- al. 2002).
tween females with BPD and control subjects. Similar structural imaging studies of fe-
These divergent findings suggest the possi- males with BPD report reductions in the vol-
bility that the cholinergic receptors mediating ume of subcortical structures such as the
behavioral and hormonal responses to cho- amygdala (Rusch et al. 2003; Schmahl et al.
linergic agents in these subjects may be very 2003; Tebartz van Elst et al. 2003; but see New
different by virtue of brain location. et al. 2007) and hippocampus (Schmahl et al.
114 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

2003; Tebartz van Elst et al. 2003). Studies tal and prefrontal regions as well as in the
also report reductions in the volumes of both hippocampus and cuneus (Juengling et al.
orbitofrontal and anterior cingulate (Hazlett 2003), supporting previous structural studies
et al. 2005; Tebartz van Elst et al. 2003). Given that demonstrated reductions in the volumes
the role these structures are thought to play of these brain areas. Although most of these
in emotional information processing, it is PET studies were performed in the resting
tempting to speculate that these structures condition, one PET study in females with
represent anatomical correlates of the emo- BPD showed that the replay of abandonment
tional dysregulation (including impulsive scripts prior to PET scan was associated with
aggression) seen in patients with BPD. greater increases in activity in dorsolateral
prefrontal cortex (bilaterally) and in cuneus,
Functional Imaging (PET and SPECT) but with reductions in activity in the right
anterior cingulate (Schmahl et al. 2003).
Whereas structural imaging yields only a Given that several of these structures have
static picture of the brain, SPECT or PET been shown to be smaller in these subjects
scanning can yield functional information re- compared with control subjects, the in-
lated to cerebral blood flow or cerebral glu- creased activity in these regions after the
cose metabolism, respectively. For example, abandonment task is quite notable.
SPECT studies have demonstrated reduced PET studies may also be performed after
perfusion in prefrontal cortex as well as focal the administration of neurotransmitter-
abnormalities in left temporal lobe and in- specific agents so that the activity of brain re-
creased activity in anteromedial frontal cor- gions in response to activation of specific re-
tex in limbic system in aggressive individu- ceptors by these agents can be assessed. To
als with ASPD and alcoholism (Amen et al. date, at least four studies of patients with per-
1996). A more recent study using SPECT re- sonality disorder have been performed in this
ported significant correlations between re- way. Two utilized the indirect 5-HT agonist
duced cerebral blood flow in frontal and fenfluramine, one utilized the more direct
temporal brain regions and the disturbed postsynaptic 5-HT agonist m-CPP, and one
interpersonal attitude factor from the Psy- examined the trapping of a 11 C analogue of
chopathy ChecklistRevised (Soderstrom et tryptophan. In the first fenfluramine study,
al. 2002). In homicide offenders (many of patients with prominent histories of impul-
whom presumably had ASPD), a bilateral sive aggression and BPD demonstrated
diminution of glucose metabolism has been blunted responses of glucose metabolism in
reported in both medial frontal cortex and at orbital frontal, ventral medial frontal, and
a trend level in orbital frontal cortex (Raine et cingulate cortex compared with normal sub-
al. 1994). In a study of patients with a variety jects (Siever et al. 1999). A similar result was
of personality disorders, an inverse relation- reported in the second fenfluramine study, in
ship was found between life history of ag- which patients with BPD displayed reduced
gressive impulsive behavior and regional glucose metabolism (relative to placebo)
glucose metabolism in orbital frontal cortex compared with control subjects in right me-
and right temporal lobe (Goyer et al. 1994). dial and orbital frontal cortex, left middle and
Patients meeting criteria for BPD had de- superior temporal gyri, left parietal lobe, and
creased metabolism in frontal regions corre- left caudate (Soloff et al. 2000). In the PET
sponding to BA 46 and BA 6 and increased study involving m-CPP, patients with promi-
metabolism in superior and inferior frontal nent histories of impulsive aggression and
gyrus (BA 9 and BA 45; Goyer et al. 1994). personality disorder were found to have re-
More-recent PET studies in females with duced activation of the anterior cingulate and
BPD reported hypometabolism in both fron- increased activation of the posterior cingulate
Neurobiology 115

compared with control subjects, as well as a ing emotionally aversive images (e.g., crying
relative amygdala-prefrontal disconnec- children) than did control subjects (Herpertz
tion (New et al. 2002, 2007). Given the role of et al. 2001). Another fMRI study in females
the anterior cingulate in emotional informa- with BPD reported a generally similar find-
tion processing, it is noteworthy that this area ing (left amygdala as opposed to bilateral ac-
is underactivated by 5-HT stimulation. In the tivation) using emotional faces (Donegan et
PET study examining the unilateral trapping al. 2003). Other recent studies from the two
of a 11C analogue of tryptophan, evidence for separate laboratories of the authors suggest
a reduction in 5-HT synthesis was present in a) increased activation in amygdala, and re-
the corticostriatal (e.g., medial frontal, ante- duced activation in orbitomedial prefrontal
rior cingulate, superior temporal gyri, and cortex, to anger faces in impulsively aggres-
corpus striatum) brain areas of subjects with sive subjects with personality disorders
BPD (Leyton et al. 2001). Reduction in 5-HT (none of whom was psychopathic; Coccaro et
synthesis in these regions was reported to al. 2007) and b) increased superior temporal
correlate with a laboratory measure of behav- cortex, rather than middle temporal cortex,
ioral disinhibition. Reduced transporter BOLD responses, suggesting increased re-
binding has also been reported in patients flexive processing and increased amygdala,
with impulsive aggressive personality disor- fusiform, and occipital visual area activation
der (Frankle et al. 2005), and increased 5 to emotional provocation (Koenigsberg et al.
HT2A receptor binding has been reported in 2007). Given the clear differences in known
currently aggressive subjects (Siever et al. emotional information processing between
2006). psychopathic antisocial subjects on the one
hand and borderline subjects on the other,
Functional Imaging (fMRI) these data suggest the brain sites of these dif-
Unlike PET or SPECT, fMRI does not require ferences.
the injection of a radiolabeled agent. Instead,
fMRI assesses changes in cerebral blood flow Case Example
using changes in the blood oxygenation
Mr. C is a 29-year-old, married male
leveldependent (BOLD) signal in the mag- computer technician referred for treat-
netic resonance imaging scanner. This offers ment of his impulsive aggressive out-
a much greater spatial and temporal resolu- bursts in the context of a threatened
tion compared with either PET or SPECT and separation from his wife of 4 years. Mr.
allows a finer assessment of the activation C reports impulsive aggressive out-
and deactivation of discrete regions of the bursts since his mid-teens. These out-
bursts typically involve screaming,
brain in response to specific stimuli. To date,
shouting, and throwing things around;
at least three studies using fMRI in patients he has only occasionally physically hit
with personality disorder have been pub- anyone. However, these aggressive out-
lished. In one study using fMRI, males with bursts occur several times a month and
(psychopathic) ASPD activated preselected usually several times a week, particu-
frontal and temporal regions of interest less larly when Mr. C is held up in traffic.
than did control subjects during trials of neg- Most recently, he has been having seri-
ous marital difficulty, and his wife is
atively charged emotional words (Kiehl et al.
now threatening to leave him if he does
2001), suggesting an important deficit in not get help for his anger problem. He
emotional information processing. In a simi- reports that his relationship with his
lar fMRI study in females with BPD, the wife is often stormy, with frequent
study group demonstrated greater activation fighting that sometimes goes on for
of the amygdala bilaterally (as well as activa- hours. Sometimes in the aftermath of
tion of selected frontal regions) while view- these fights Mr. C runs off and gets ex-
116 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

ceedingly drunk and drives recklessly norepinephrine, vasopressin for impulsivity

around to wn while high. At o ther and aggression, possibly acetylcholine for
times, he reports, he beats his head so mood reactivity) and in brain structures re-
hard against a wall that his forehead
lated to behavioral inhibition and emotional
bleeds (once he needed stitches). Still, at
other times he frantically pleads with
information processing (e.g., orbitofrontal
his wife not to leave him; once he took cortex, amygdala). Although patients with
an overdose of aspirin, in front of his BPD are often the most extreme in these fea-
wife, to get her to stay with him. Mr. C tures and in related biological dysfunction,
reports a history of alcohol abuse in his specific biological dysfunction related to spe-
late teens and early twenties and a his- cific traits (e.g., serotonin dysfunction with
tory of gambling to excess up until impulsive aggression) can be seen in patients
1 year prior to evaluation.
with other, nonborderline personality disor-
Mr. C underwent a research evalua-
tion in the Mood and Personality Disor- ders. As such, it is doubtful that any assess-
ders Program. Diagnostic evaluation re- ment of specific neurobiological function will
vealed the presence of BPD with traits of be specific to patients with BPD.
histrionic, narcissistic, and obsessive-
compulsive personality disorder. He
was also found to meet DSM-IV-TR cri-
teria for two episodes of major depres-
sion in the past and for alcohol abuse
The Cluster C personality disorders include
(past) and pathological gambling (past).
avoidant, dependent, and obsessive-compul-
He underwent a variety of research-
related studies including d-fenfluramine sive personality disorders. Individuals with
(d-FEN) challenge and was found to these disorders present with varied degrees
have a blunted, but not absent, prolactin of anxiety sometimes expressed as rigidity,
response to d-FEN (2.3 ng/mL com- particularly in the case of obsessive-compul-
pared with 6.3 3.4 ng/mL for healthy sive personality disorder. Of the three disor-
male control subjects); his CSF 5-HIAA ders, avoidant personality disorder is most
level was not abnormal (23.9 ng/mL
like generalized social phobia in Axis I, and a
compared with 20.0 4.9 ng/mL for
healthy male control subjects). The mod- great degree of comorbidity occurs between
est magnitude of his prolactin response the two diagnoses (Dahl 1996). As in other
to d-FEN suggests a limited degree of personality disorder clusters, there is overlap
central serotonin system dysfunction. among the disorders in this cluster and with
Mr. C entered a treatment trial of those in other personality disorder clusters,
fluoxetine and experienced a reduction
particularly Cluster B. To date, there has been
in overt aggressive behavior over a pe-
riod of several weeks. Over this time his much less empirical neurobiological research
relationship with his wife somewhat with patients in Cluster C.
improved, and he is now in dialectical
behavioral therapy to work on other as- Behavioral Genetics
pects of his interpersonal difficulties
with others in his life. As with the Cluster B personality disorders,
twin studies suggest substantial genetic in-
Summary fluence for each of the Cluster C personality
disorders (Torgersen et al. 2000). Heritability
The studies discussed in this section suggest for Cluster C personality disorders as a
that patients with Cluster B personality disor- group was estimated at 0.62; heritabilities for
der have dysfunction in a variety of neurobi- each disorder in the study were 0.78 for ob-
ological areas that may underlie their clinical sessive-compulsive, 0.57 for dependent, and
presentation. Dysfunction can occur in multi- 0.28 for avoidant personality disorder. The
ple monoaminergic systems (e.g., serotonin, best-fitting models did not include shared fa-
Neurobiology 117

milial environment effects, although a model creased cortisol responses to serotonergic

consisting only of shared familial and unique agents (Tancer et al. 1999), and social anxiety
environmental effects could not be defini- disorders respond to selective serotonin re-
tively ruled out for dependent personality uptake inhibitors that re-regulate serotoner-
disorders. Family studies suggest a familial gic activity (Schneier et al. 2003). Shyness (re-
association between social anxiety disorder lated to avoidant traits) has been associated
and avoidant personality disorder (Schneier with the serotonin transporter reporter re-
et al. 2002). Avoidant, dependent, and anx- gion L allele but not to COMT, MAO-A, or
ious cluster personality disorders show sig- DRD4 alleles. Growth hormone regulation
nificant familiarity (Reich 1989), and both has also been associated with social anxiety
avoidant and independent personality traits (Schneier et al. 2002).
are found in relatives of patients with panic
disorder (Reich 1991). Neuropsychological and
Psychophysiological Correlates
Increased amygdala activation in fMRI has
There has been little biological study of the been shown in social phobia in one study
Cluster C personality disorders. However, (Schneier et al. 1999) as well as in recognition
low dopamine metabolites in CSF have been bias for recall of disapproving faces in an-
identified in patients with social anxiety dis- other (Foa et al. 2000). However, skin con-
order (Johnson et al. 1994), which overlaps to ductance and heart rate change and startle
a great extent with avoidant personality dis- response during viewing of slides with emo-
order, whereas nonselective monoamine ox- tionally charged themes did not distinguish
idase inhibitors (which increase dopamine patients with avoidant personality disorder
transmission) or dopaminergic antidepres- from control subjects (Herpertz et al. 2000).
sants improve social anxiety (Schneier et al. Psychophysiological studies have not been
2002). Imaging studies are also consistent extensively undertaken in the other Cluster
with this finding, with low dopamine trans- C personality disorders.
porter binding demonstrated in generalized
social anxiety disorder (Tiihonen et al. 1997)
and lower D2 receptor binding in a SPECT
study of generalized social anxiety disorder Genetic and neurobiological research has
(Schneier et al. 2000). In addition, three PET been limited in patients with Cluster C per-
studies support a relationship of reduced D2 sonality disorders, but reductions in dopa-
binding associated with detachment, which minergic activity and increases in serotoner-
correlates with social avoidance consistent gic activity are hinted at in the data available.
with that observed both in patients with
Cluster C personality disorders and in pa-
tients with schizoid personality disorder
(Schneier et al. 2000). Genetic studies of these
types of behaviors have been found in asso- Research in the psychobiology of personality
ciation with the dopamine transporter gene disorder has advanced much since the 1980s.
DAT1 (Blum et al. 1997). These studies cumu- Although there is clear evidence of a number
latively suggest low dopaminergic activity in of biogenetic correlates of personality disor-
social anxiety disorder and likely in avoidant der traits, future efforts need to be directed
personality disorder as well. along a variety of lines to increase our under-
In the serotonergic system, on the other standing of how alterations in brain function
hand, patients with social anxiety have in- lead to the development and manifestation
118 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

of these traits. Such lines of investigation brospinal fluid amine metabolites. Psychiatry
may be aimed at 1) how genetic and environ- Res 1:131139, 1979
Brown GL, Ebert MH, Goyer PF, et al: Aggression,
mental influences interact with neurotrans-
suicide, and serotonin: relationships to CSF
mitter function to lead to specific traits; amine metabolites. Am J Psychiatry 139:741
2) how neurotransmitter function interacts 746, 1982
with the regulation of cognitive and emo- Buchsbaum MS, Nenadic I, Hazlett EA, et al: Dif-
tional function across distributed neural net- ferential metabolic rates in prefrontal and tem-
works to lead to specific traits; and 3) how poral Brodmann areas in schizophrenia and
schizotypal personality disorder. Schizophr
understanding brain function at these levels Res 54:141150, 2002
can enable us to devise more effective ways Cadenhead KS: Vulnerability markers in the
to treat personality disorder traits both phar- schizophrenia spectrum: implications for phe-
macologically and psychotherapeutically. nomenology, genetics, and the identification
of the schizophrenia prodrome. Psychiatr Clin
North Am 25:837853, 2002
Cadoret RJ, OGorman TW, Troughton E, et al: Al-
coholism and antisocial personality: interrela-
REFERENCES tionships, genetic and environmental factors.
Arch Gen Psychiatry 42:161167, 1985
Amen DG, Stubblefield M, Carmicheal B: Brain Caspi A, McClay J, Moffitt TE, et al: Role of geno-
SPECT findings and aggressiveness. Ann Clin type in the cycle of violence in maltreated chil-
Psychiatry 3:129137, 1996 dren. Science 297:851854, 2002
American Psychiatric Association: Diagnostic and Cleare AJ, Bond AJ: Ipsapirone challenge in ag-
Statistical Manual of Mental Disorders, 2nd gressive men shows an inverse correlation be-
Edition. Washington, DC, American Psychiat- tween 5-HT1A receptor function and aggres-
ric Association, 1968 sion. Psychopharmacology 148:344349, 2000
American Psychiatric Association: Diagnostic and Coccaro EF, Siever LJ, Klar HM, et al: Serotonergic
Statistical Manual of Mental Disorders, 4th studies in patients with affective and person-
Edition, Text Revision. Washington, DC, ality disorder: correlates with suicidal and im-
American Psychiatric Association, 2000 pulsive aggressive behavior. Arch Gen Psychi-
Amin F, Coccaro EF, Mitropoulou V, et al: Plasma atry 46:587599, 1989
HVA in schizotypal personality disorder, in Coccaro EF, Gabriel S, Siever LJ: Buspirone chal-
Plasma Homovanillic Acid Studies in Schizo- lenge: preliminary evidence for a role for 5-
phrenia: Implications for Presynaptic Dopa- HT-1A receptors in impulsive aggressive be-
mine Dysfunction. Edited by Friedhoff AJ, havior in humans. Psychopharmacol Bull
Amin F. Washington, DC, American Psychiat- 26:393405, 1990
ric Press, 1997, pp 133149 Coccaro EF, Lawrence T, Trestman R, et al: Growth
Amin F, Silverman JM, Siever LJ, et al: Genetic an- hormone responses to intravenous clonidine
tecedents of dopamine dysfunction in schizo- challenge correlates with behavioral irritabil-
phrenia. Biol Psychiatry 45:11431150, 1999 ity in psychiatric patients and in healthy vol-
Atmaca M, Kuloglu M, Tezcan E, et al: Serum cho- unteers. Psychiatry Res 39:129139, 1991
lesterol and leptin levels in patients with bor- Coccaro EF, Kavoussi RJ, Hauger RL: Physiologic
derline personality disorder. Neuropsychobi- responses to d-fenfluramine and ipsapirone
ology 45:167171, 2002 challenge correlate with indices of aggression
Blum K, Braverman ER, Wu S: Association of in males with personality disorder. Int Clin
polymorphisms of dopamine D2 receptor Psychopharmacol 10:177180, 1995
(DRD2), and dopamine transporter (DAT1) Coccaro EF, Kavoussi RJ, Sheline YI, et al: Impul-
genes with schizoid/avoidant behaviors sive aggression in personality disorder: corre-
(SAB). Mol Psychiatry 2:239246, 1997 lates with 3H-paroxetine binding in the plate-
Braff DL, Freedman R: Endophenotypes in studies let. Arch Gen Psychiatry 53:531536, 1996
of the genetics of schizophrenia, in Neuropsy- Coccaro EF, Kavoussi RJ, Cooper TB, et al: Central
chopharmacology: The Fifth Generation of serotonin and aggression: inverse relationship
Progress. Edited by Davis KL, Charney D, with prolactin response to d-fenfluramine, but
Coyle JT, et al. Philadelphia, PA, Lippincott not with CSF 5-HIAA concentration in human
Williams & Wilkins, 2002, pp 703716 subjects. Am J Psychiatry 154:14301435, 1997a
Brown GL, Goodwin FK, Ballenger JC, et al: Ag- Coccaro EF, Kavoussi RJ, Trestman RL, et al: Seroto-
gression in humans correlates with cere- nin function in personality and mood disorder:
Neurobiology 119

intercorrelations among central indices and ag- Hansenne M, Pitchot W, Pinto E, et al: 5-HT1A
gressiveness. Psychiatry Res 73:114, 1997b dysfunction in borderline personality disor-
Coccaro EF, Kavoussi RK, Hauger RL, et al: Cere- der. Psychol Med 32:935941, 2002
brospinal fluid vasopressin: correlates with Hazlett EA, New AS, Newmark R, et al: Reduced
aggression and serotonin function in personal- anterior and posterior cingulated gray matter
ity disordered subjects. Arch Gen Psychiatry in borderline personality disorder. Biol Psy-
55:708714, 1998 chiatry 58:614623, 2005
Coccaro EF, Lee R, McCloskey M: Norepinephrine Hazlett EA, Buchsbaum MS, Haznedar MM, et al:
function in personality disorder: plasma free Cortical gray and white matter volume in un-
MHPG correlates inversely with life history of medicated schizotypal and schizophrenia pa-
aggression. CNS Spectr 8:731736, 2003 tients. Schizophr Res 101:111123, 2008
Coccaro EF, McCloskey MS, Fitzgerald DA, et al: Herpertz SC, Schwenger UB, Kunert HJ, et al:
Amygdala and orbitofrontal reactivity to so- Emotional responses in patients with border-
cial threat in individuals with impulsive ag- line as compared with avoidant personality
gression. Biol Psychiatry 62:168178, 2007 disorder. J Personal Disord 14: 339351, 2000
Crick NR, Dodge KA: Social information-process- Herpertz SC, Dietrich TM, Wenning B, et al: Evi-
ing mechanisms in reactive and proactive ag- dence of abnormal amygdala functioning in
gression. Child Dev 67:9931002, 1996 borderline personality disorder: a functional
Dahl AA: The relationship between social phobia MRI study. Biol Psychiatry 50: 292298, 2001
and avoidant personality disorder: workshop Hollander E, Stein D, DeCaria CM, et al: Seroton-
report 3. Int Clin Psychopharmacol 11(suppl ergic sensitivity in borderline personality dis-
3):109112, 1996 order: preliminary findings. Am J Psychiatry
Dolan M, Anderson IM, Deakin JF: Relationship 151:277280, 1994
between 5-HT function and impulsivity and Hymowitz P, Frances A, Jacobsberg LB, et al: Neu-
aggression in male offenders with personality roleptic treatment of schizotypal personality
disorders. Br J Psychiatry 178:352359, 2001 disorders. Compr Psychiatry 27:267271, 1986
Donegan NH, Sanislow CA, Blumberg HP, et al: Johnson MR, Lydiard RB, Zealberg JJ: Plasma and
Amygdala hyperreactivity in borderline per- CFS levels in panic patients with comorbid so-
sonality disorder: implications for emotional cial phobia. Biol Psychiatry 36:426427, 1994
dysregulation. Biol Psychiatry 54:12841293, Juengling FD, Schmahl C, Hesslinger B, et al:
2003 Positron emission tomography in female pa-
Foa EB, Gilboa-Schechtman E, Amir N: Memory tients with borderline personality disorder.
bias in generalized social phobia: remember- J Psychiatr Res 37:109115, 2003
ing negative emotional expressions. J Anxiety Kalus O, Bernstein DP, Siever LJ: Schizoid person-
Disord 14:501519, 2000 ality disorder: a review of its current status.
Frankle WG, Lombardo I, New AS, et al: Brain se- J Personal Disord 7: 4352, 1993
rotonin transporters distribution in subjects Kavoussi RJ, Coccaro EF: The amphetamine chal-
with impulsive aggressvity: a positron emis- lenge test correlates with affective lability in
sion study with [11C] McN5652. Am J Psychi- healthy volunteers. Psychiatry Res 48:219228,
atry 162:915923, 2005 1993
Gardner DL, Lucas PB, Cowdry RW: CSF metabo- Kiehl K, Smith AM, Hare RD, et al: Limbic abnor-
lites in borderline personality disorder com- malities in affective processing by criminal
pared with normal controls. Biol Psychiatry psychopaths as revealed by functional mag-
28:247-54, 1990 netic resonance imaging. Biol Psychiatry
Goldberg TE, Egan MF, Gscheidle T, et al: Execu- 50:677684, 2001
tive subprocesses in working memory: rela- Kirrane RM, Mitropoulou V, Nunn M, et al: Effects
tionship to catechol-O-methyltransferase of amphetamine on visuospatial working
Val158Met genotype and schizophrenia. Arch memory performance in schizophrenia spec-
Gen Psychiatry 60:889896, 2003 trum personality disorder. Neuropsychophar-
Goveas JS, Csernansky JG, Coccaro EF: Platelet se- macology 22:1418, 2000
rotonin content correlates inversely with life Koenigsberg HW, Reynolds D, Goodman M, et al:
history of aggression in personality-disor- Risperidone in the treatment of schizotypal
dered subjects. Psychiatry Res 126:2332, 2004 personality disorder. J Clin Psychiatry 64:628
Goyer PF, Andreason PJ, Semple WE, et al: 634, 2003
Positron-emission tomography and personal- Koenigsberg HW, Buchsbaum MS, Buchsbaum
ity disorders. Neuropsychopharmacology BR, et al: Fuctional MRI of visuospatial work-
10:2128, 1994 ing memory in schizotypal personality disor-
120 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

der: a region-of-interest analysis. Psychol Med Moss HB, Yao JK, Panzak GL: Serotonergic re-
35:10191030, 2005 sponsivity and behavioral dimensions in anti-
Koenigsberg HW, Prohovnik I, Lee H, et al: Neu- social personality disorder with substance
ral correlates of the processing of negative and abuse. Biol Psychiatry 28:325338, 1990
positive social scenes in borderline personality New AS, Gelernter J, Yovell Y, et al: Tryptophan
disorder. Biol Psychiatry 61:8S, 1045, 2007 hydroxylase genotype is associated with im-
Laakso MP, Gunning-Dixon F, Vaurio O, et al: Pre- pulsive aggression measures. Am J Med Genet
frontal volumes in habitually violent subjects 81:1317, 1998
with antisocial personality disorder and type 2 New AS, Sevin EM, Mitropoulou V, et al: Serum
alcoholism. Psychiatry Res 114:95102, 2002 cholesterol and impulsivity in personality dis-
Lappalainen J, Long JC, Eggert M, et al: Linkage of orders. Psychiatry Res 85:145150, 1999
antisocial alcoholism to the serotonin 5-HT1B New AS, Hazlett EA, Buchsbaum MS, et al: Blunted
receptor gene in two populations. Arch Gen prefrontal cortical 1 8 fluorodeoxyglucose
Psychiatry 55:989994, 1998 positron emission tomography response to meta-
Laruelle M, Kegeles L, Zea-Pance Y, et al: Amphet- chlorophenylpiperazine in impulsive aggres-
amine-induced dopamine release in patients sion. Arch Gen Psychiatry 59:621629, 2002
with schizotypal personality disorders studies New AS, Hazlett EA, Buchsbaum MS, et al:
by SPECT and [123] IBZM. Neuroimage 16:S61, Amygdala-prefrontal disconnection in bor-
2002 derline personality disorder. Neuropsycho-
Levitt JJ, McCarley RW, Dickey CC, et al: MRI pharmacology 32:16291640, 2007.
study of caudate nucleus volume and its cog- Ni X, Bismil R, Chan K, et al: Serotonin 2A receptor
nitive correlates in neuroleptic-naive patients gene is associated with personality traits, but
with schizotypal personality disorder. Am J not to disorder, in patients with borderline per-
Psychiatry 159:11901197, 2002 sonality disorder. Neurosci Lett 408:214219,
Leyton M, Okazawa H, Diksic M, et al: Brain re- 2006a
gional alpha-[11C]methyl-L-tryptophan trap- Ni X, Chan K, Bulgin N, et al: Association between
ping in impulsive subjects with borderline per- serotonin transporter gene and borderline per-
sonality disorder. Am J Psychiatry 158:775782, sonality disorder. J Psychiatr Res 40:448453,
2001 2006b
Linnoila M, Virkkunen M, Scheinin M, et al: Low Ni X, Sicard T, Bulgin N, et al: Monoamine oxidase
cerebrospinal fluid 5-hydroxyindolacetic acid a gene is associated with borderline personal-
concentration differentiates impulsive from ity disorder. Psychiatr Genet 17:153157, 2007
nonimpulsive violen t behavior. Life Sci Nielsen DA, Goldman D, Virkkunen M, et al: Sui-
33:26092614, 1983 cidality and 5-hydroxyindoleacetic acid con-
Manuck SB, Flory JD, Ferrell RE, et al: Aggression centration associated with a tryptophan hy-
and anger-related traits associated with a droxylase polymorphism. Arch Gen Psychiatry
polymorphism of the tryptophan hydroxylase 51:3438, 1994
gene. Biol Psychiatry 45:603614, 1999 Nielsen DA, Virkkunen M, Lappalainen J, et al: A
McClure MM, Barch DM, Romero MJ, et al: The ef- tryptophan hydroxylase gene marker for sui-
fects of guanfacine on context processing ab- cidality and alcoholism. Arch Gen Psychiatry
normalities in schizotypal personality disor- 55:593602, 1998
der. Biol Psychiatry 61:11571160, 2007 OKeane V, Moloney E, ONeill H, et al: Blunted
Minzenberg MJ, Xu K, Mitropoulou V, et al: Cate- prolactin responses to d-fenfluramine in soci-
chol-O-methyltransferase Val 158 Met geno- opathy: Evidence for subsensitivity of central
type variation is associated with prefrontal de- serotonergic function. Br J Psychiatry 160:643
pendent task performance in schizotypal 646, 1992
personality disorder patients and comparison Paris J, Zweig-Frank H, Kin NM, et al: Neurobio-
groups. Psychiatr Genet 16:117124, 2006 logical correlates of diagnosis and underlying
Mitropoulou V, Harvey PD, Maldari LA, et al: traits in patients with borderline personality
Neuropsychological performance in schizo- disorder compared with normal controls. Psy-
typal personality disorder: evidence regarding chiatry Res 121:239252, 2004
diagnostic specificity. Biol Psychiatry 52:1175 Raine A, Buchsbaum M, Stanley J, et al: Selective
1182, 2002 reductions in prefrontal glucose metabolism
Mitropoulou V, Harvey PD, Zegarelli G, et al: in murderers. Biol Psychiatry 36:365373, 1994
Neuropsychological performance in schizo- Raine A, Lencz T, Bihrle S, et al: Reduced prefron-
typal personality disorder: importance of tal gray matter volume and reduced auto-
working memory. Am J Psychiatry 162:1896 nomic activity in antisocial personality disor-
1903, 2005 der. Arch Gen Psychiatry 57:119127, 2000
Neurobiology 121

Raine A, Lencz T, Taylor K, et al: Corpus callosum Shihabuddin L, Silverman JM, Buchsbaum MS, et
abnormalities in psychopathic antisocial indi- al: Ventricular enlargement associated with
viduals. Arch Gen Psychiatry 60:11341142, 2003 linkage marker for schizophrenia-related dis-
Reich JH: Familiality of DSM-III dramatic and orders in one pedigree. Mol Psychiatry 1:215
anxious personality clusters. J Nerv Ment Dis 222, 1996
177:96100, 1989 Shihabuddin L, Buchsbaum MS, Hazlett EA, et al:
Reich JH: Avoidant and dependent personality Striatal size and relative glucose metabolic
traits in relatives of patients with panic disor- rate in schizotypal personality disorder and
der, patients with dependent personality dis- schizophrenia. Arch Gen Psychiatry 58:877
order, and normal controls. Psychiatry Res 884, 2001
39:8998, 1991 Siegel BV, Trestman RL, OFlaithbheartaigh SO, et
Rinne T, Westenberg HG, den Boer JA, et al: Sero- al: D-amphetamine challenge effects on Wis-
tonergic blunting to meta-chlorophenylpipera- consin Card Sort Test performance in schizo-
zine (m-CPP) highly correlates with sustained typal personality disorder. Schizophr Res
childhood abuse in impulsive and autoaggres- 20:2932, 1996
sive female borderline patients. Biol Psychia- Siever LJ: The biology of the boundaries of schizo-
try 47:548556, 2000 phrenia, in Advances in Neuropsychiatry and
Rinne T, de Kloet ER, Wouters L, et al: Hyperre- Psychopharmacology, Vol 1: Schizophrenia
sponsiveness of hypothalamic-pituitary-adre- Research. Edited by Tamminga CA, Schulz SC.
nal axis to combined dexamethasone/corti- New York, Raven, 1991, pp 181191
cotropin-releasing hormone challenge in Siever LJ: Brain structure/function and the
female borderline personality disorder sub- dopamine system in schizotypal personality
jects with a history of sustained childhood disorder, in Schizotypal Personality. Edited by
abuse. Biol Psychiatry 52:11021112, 2002 Raine A, Lencz T, Mednick F. New York, Cam-
Rusch N, van Elst LT, Ludaescher P, et al: A voxel- bridge University Press, 1995, pp 272286
based morphometric MRI study in female pa- Siever LJ, Davis KL: The pathophysiology of the
tients with borderline personality disorder. schizophrenic disorders: perspective from the
Neuroimage 20:385392, 2003 spectrum. Am J Psychiatry 161:398413, 2004
Schmahl CG, Vermetten E, Elzinga BM, et al: Mag- Siever LJ, Trestman RL: The serotonin system and
netic resonance imaging of hippocampal and aggressive personality disorder. Int Clin Psy-
amygdala volume in women with childhood chopharmacol 8 (suppl 2):3339, 1993
abuse and borderline personality disorder. Siever LJ, Amin F, Coccaro EF, et al: Plasma homo-
Psychiatry Res 122:193198, 2003 vanillic acid in schizotypal personality disor-
Schneier F, Weiss U, Kessler C: Subcortical corre- der patients and controls. Am J Psychiatry
lates of differential classical conditioning of 148:12461248, 1991
aversive emotional reactions in social phobia. Siever LJ, Amin F, Coccaro EF, et al: CSF homo-
Biol Psychiatry 45:863871, 1999 vanillic acid in schizotypal personality disor-
Schneier FR, Liebowitz MR, Abi-Dargham A: Low der. Am J Psychiatry 150:149151, 1993
dopamine D2 receptor binding potential in so- Siever LJ, Buchsbaum M, New A, et al: D,L-fenflu-
cial phobia. Am J Psychiatry 157:457459, 2000 ramine response in impulsive personality dis-
Schneier FR, Blanco C, Smita XA, et al: The social order assessed with [18]-fluorodeoxyglucose
anxiety spectrum. Psychiatr Clin North Am positron emission tomography. Neuropsy-
25:757774, 2002 chopharmacology 20:413423, 1999
Schneier FR, Blanco C, Campeas R, et al: Citalo- Siever LJ, Koenigsberg HW, Harvey P, et al: Cogni-
pram treatment of social anxiety disorder with tive and brain function in schizotypal person-
comorbid major depression. Depress Anxiety ality disorder. Schizophr Res 54:157167, 2002
17:191196, 2003 Siever LJ, Frankle WG, Ducci F, et al: The seroto-
Schulz SC, Schulz PM, Dommisse C, et al: Am- nin transporter binding and genotypes in im-
phetamine response in borderline patients. pulsive personality disorders. Abstract, The
Psychiatry Res 15:97108, 1985 Society of Biological Psychiatry 61st Annual
Schulz SC, Cornelius J, Schulz PM, et al: The am- Scientific Convention & Program, 2006
phetamine challenge test in patients with bor- Silverman JM, Pinkham L, Horvath TB, et al: Af-
derline personality disorder. Am J Psychiatry fective and impulsive personality disorder
145:809814, 1988 traits in the relatives of patients with border-
Schulz SC, Thomson R, Brecher M: The efficacy of line personality disorder. Am J Psychiatry
quetiapine vs. haloperidol and placebo: a 148:13781385, 1991
meta-analytic study of efficacy. Schizophr Res Simeon D, Stanley B, Frances A, et al: Self-mutila-
62:112, 2003 tion in personality disorders: psychological
122 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

and biological correlates. Am J Psychiatry Trestman RL, Keefe RSE, Harvey PD, et al: Cogni-
149:221226, 1992 tive function and biological correlates of cog-
Soderstrom H, Hultin L, Tullberg M, et al: Re- nitive performance in schizotypal personality
duced frontotemporal perfusion in psycho- disorder. Psychiatry Res 59:127136, 1995
pathic personality. Psychiatry Res 114:8194, Virkkunen M, Nuutila A, Goodwin FK, et al: Cere-
2002 brospinal fluid monoamine metabolite levels
Soloff PH, Meltzer CC, Greer PJ, et al: A fenflu- in male arsonists. Arch Gen Psychiatry 44:241
ramine-activated FDG-PET study of border- 247, 1987
line personality disorder. Biol Psychiatry Virkkunen M, DeJong J, Bartko J, et al: Relation-
47:540547, 2000 ship of psychobiological variables to recidi-
Steinberg BJ, Trestman R, Mitropoulou V, et al: De- vism in violent offenders and impulsive fire
pressive response to physostigmine challenge setters. Arch Gen Psychiatry 46: 600603, 1989
in borderline personality disorder patients. Virkkunen M, Rawlings R, Tokola R: CSF bio-
Neuropsychopharmacology 17:264273, 1997 chemistries, glucose metabolism, and diurnal
Tancer ME, Mailman RB, Stein MB, et al: Subcorti- activity rhythms in alcoholic, violent offend-
cal correlates of differential classical condi- ers, fire setters, and healthy volunteers. Arch
tioning of aversive emotional reactions in so- Gen Psychiatry 51:2027, 1994
cial phobia. Biol Psychiatry 45:863871, 1999 Webb CT, Levinson DF: Schizotypal and paranoid
Tebartz van Elst L, Hesslinger B, Thiel T, et al: Fr- personality disorder in the relatives of patients
ontolimbic brain abnormalities in patients with schizophrenia and affective disorders: a
with borderline personality disorder: a volu- review. Schizophr Res 11:8192, 1993
metric magnetic resonance imaging study. Weinberger DR, Egan MF, Bertolino A, et al: Pre-
Biol Psychiatry 54:163171, 2003 frontal neurons and the genetics of schizo-
Tiihonen J, Kuikka J, Bergstrom K: DA reuptake phrenia. Biol Psychiatry 50:825844, 2001
site densities in patients with social phobia. Zanarini MC, Frankenburg FR: Omega-3 fatty
Am J Psychiatry 154:239242, 1997 acid treatment of women with borderline per-
Torgersen S, Lygren S, Oien PA, et al: A twin study sonality disorder: a double-blind, placebo-
of personality disorders. Compr Psychiatry controlled pilot study. Am J Psychiat ry
41:416425, 2000 160:167169, 2003
Developmental Issues

Patricia Cohen, Ph.D.

Thomas Crawford, Ph.D.

A side from well-documented developmen- nally, we address problems in assessing

tal links between early conduct disorder and these disorders in young people based on the
antisocial personality disorder (ASPD), there limited number of measurement instru-
are large gaps in our knowledge about child- ments currently available.
hood antecedents of other DSM-IV-TR per-
sonality disorders (American Psychiatric As-
sociation 2000; Widiger and Sankis 2000). In ETIOLOGICAL AND
this chapter we discuss how developmental
processes and selected risk factors lead to the
emergence and persistence of personality In an early paper on the borderline-child-to-
disorders in young people. We highlight be, Pine (1986) identified three key factors in
changes in how children and adolescents his developmental model of borderline per-
construct mental representations of them- sonality disorder (BPD) in young people.
selves and other people and then consider First, he hypothesized how early abuse or
how distortions in this developmental pro- trauma overwhelms the child, especially
cess manifest in personality disorder. We when the trauma is ongoing or experienced
outline what we have learned about early from a variety of sources. Second, childhood
trajectories of personality disorders and dis- trauma may interfere with the development
cuss how clinicians and researchers can eval- of how trust, personal attachments, anxiety,
uate the normative and clinical significance aggression, and self-esteem are experienced
of symptoms in children and adolescents. Fi- and expressed. Third, young people may fas-

124 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

ten onto immature defenses almost as though Early Working Models of

these defenses were survival techniques for Self and Other
desperate situations. More recent empirical
research points to other factors that contrib- Attachment theory (Bowlby 1969, 1973)
ute to the emergence and persistence of per- focuses on developmental experiences re-
sonality disorders over time, including ge- flecting secure and insecure relationships
netic effects (Coolidge et al. 2001; Jang 2005) between infants and caregivers and empha-
and co-occurring Axis I disorders (Crawford sizes how young people come to perceive
et al. 2001b, 2008; Kasen et al. 2007). themselves and others (Lyddon and Alford
When individual risk factors occur in iso- 2007). A secure attachment typically occurs
lation, they often may be offset by normative when the caregiver has been available and
maturational factors in social or cognitive sensitive to the needs of the infant or toddler,
domains. Conduct disorder in childhood, for especially in times of distress. Young chil-
instance, does not usually lead to ASPD in dren can better manage negative emotions,
adulthood. When risk factors occur in combi- such as anger or fear, within a secure rela-
nation, however, they may overwhelm the tionship because these feelings have been as-
young persons ability to cope, thus leading sociated with soothing and effective re-
immature defenses to become inflexible and sponses by the caregiver (Sroufe 1996). By
maladaptive over time. On the other hand, providing this external form of affect regula-
even a child who has experienced a signifi- tion, caregivers prevent infants from being
cant trauma may be protected from lasting overwhelmed and help them gradually de-
damage to personality functioning if he or velop the ability to regulate their own affect
she is securely attached to parents who can (Mikulincer et al. 2003). When caregivers are
buffer the impact of the trauma. If trauma- inconsistent or rejecting, infants and toddlers
tized children are anxiously attached to par- instead may underregulate their own affect
ents instead, thus reducing the protective ef- or restrict it excessively.
fects, they may be at greater risk for lasting As hypothesized in attachment theory,
personality dysfunction (Alexander 1992; very basic mental representations of self and
Bleiberg 2001; Levy 2005). Because genetic, others are thought to emerge during infancy
interpersonal, and early trauma risk factors through affective experiences that character-
are all addressed elsewhere (see Cloninger ize the childcaregiver relationship. These
2005 and Jang 2005; see in this volume Chap- preverbal experiences are labeled internal
ter 8, Childhood Experiences and the De- working models and broadly reflect whether
velopment of Maladaptive and Adaptive infants expect caregivers to be available or
Personality Traits, and Chapter 11, Mental- helpful when needed. Young children may
ization-Based Treatment of Borderline Per- also internalize a basic sense of whether they
sonality Disorder), we focus here on distur- are worthy of love and whether other people
bances in how children perceive themselves can be trusted to provide love and emotional
and the people around them and how devel- support. Individuals who have predomi-
opmental changes in cognitive ability may nantly negative self-representations usually
play a role in the formation of personality have anxious attachment styles, and those
disorders. In this context we draw on the the- who have predominantly negative represen-
oretical literature on attachment in infancy tations of others tend to have avoidant at-
and childhood (e.g., Cassidy and Shaver tachment styles. Avoidant attachment is
2008; Fonagy et al. 2003) and identity devel- thought to stem from cool, rejecting, and dis-
opment in adolescence (Erikson 1968) and tant treatment by attachment figures, and
seek to bridge the two using Harters (1998) anxious attachment is traced to inconsistent
work on development of self-representation. and unpredictable treatment by early attach-
Developmental Issues 125

ment figures (Ainsworth et al. 1978; Roth- trajectory leading toward paranoid or schizo-
bard and Shaver 1994). typal disturbances in adolescence and adult-
As a guide to behavior, internal working hood. In the absence of biological vulnerabil-
models influence whether young children ity, negative schemas regarding others and
seek to regulate affective distress by ap- corresponding behaviors may gradually be
proaching or by avoiding attachment figures, modified as young people learn that their
or even by alternating between these oppos- family is not typical of the broader social en-
ing strategies for managing negative emo- vironment. Normal maturation processes
tions. With growth in cognitive capacity, thus may reduce the likelihood that early
these basic mental representations of self and Cluster A disturbances persist over time.
other (schemas) are subject to elaboration, Early maltreatment by caregivers may
refinement, and increasing differentiation produce a serious disturbance called disorga-
from affective experiences that occur in close nized attachment (Solomon and George 1999;
relationships. Nevertheless, these schemas van Ijzendoorn and Bakermans-Kranenburg
appear to have a remarkably enduring im- 2003). Abused infants and children often ex-
pact on interpersonal strategies used to regu- perience sharp conflicts when approaching
late emotional distress (e.g., Waters et al. caregivers for comfort and support when
2000). Developmental changes often reflect they also expect maltreatment from them
heterotypic continuity in how attachment thus provoking unstable fluctuations be-
styles are expressed in different relationships tween conflicting attachment strategies and
across developmental stages. That is, a neg- behavior. Furthermore, children may fear the
ative self-schema may generate anxious loss of the caregivers they depend on, thus
preoccupation with changing attachment fig- limiting their ability to experience or express
uresa primary caregiver in early child- any age-appropriate anger or aggression to-
hood, a peer group in adolescence, a romantic ward that person. This phenomenon proba-
partner in adulthoodbut nevertheless re- bly pertains most to the development of the
flect the same basic difficulty regulating affect marked instability in interpersonal relation-
across the different kinds of relationships. ships associated with borderline psychopa-
Although distorted working models ac- thology (Fonagy et al. 2000; Levy 2005; Wes-
companying insecure attachment are not ten et al. 2006).
pathological by themselves, they may never-
theless contribute to the formation of Axis II Developmental Changes in How the
psychopathology, especially when combined Self and Others Are Perceived
with other risks or biological vulnerability.
Markedly negative representations of others, Harter (1998) described how cognitive devel-
for instance, may explain higher levels of dis- opment in infancy, childhood, and adoles-
trust and suspiciousness in avoidant preado- cence leads to changes in how young people
lescents when compared with more securely experience their sense of self. Harter charac-
attached age peers. From this developmental terized the self as a cognitive structure around
starting point, a variety of pathways leading which behavior is organized, thus anchoring
to pathological and nonpathological out- it squarely within the larger framework of
comes may depend on heritable and environ- personality. As a product of the interaction of
mental risk factors that also influence how biological and social forces, the self under-
personality unfolds. A negative working goes progressive change throughout devel-
model of others, reinforced by a hostile and opment. Despite these changes, the self pro-
secretive family environment and combined vides a sense of continuity and a source for
with a biological vulnerability to Cluster A scripts to organize behavior, thereby creating
disturbances, may foster a developmental a foundation for later identity.
126 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

Harter emphasized how self-representa- ble later symptoms of narcissistic personality

tion often reflects self-evaluation, a process of disorder. If unrealistically positive self-repre-
comparing oneself with other people or with sentations become inflexible and persist over
an ideal self, which evolves over time as new time, they may limit the young persons abil-
cognitive abilities emerge during develop- ity to abandon immature self-representa-
ment. Self-evaluation may be filled with in- tions as their cognitive resources and percep-
flated self-worth at one end of the spectrum or tual skills increase during the course of
laden with self-contempt at the other, and normal development. When serving defen-
both ends of the spectrum may play an early sive functions against childhood adversity,
role in the formation of personality disorders. early grandiosity may persist and harden
Even at these extremes, cognitive develop- into personality disorder symptoms. Fur-
ment may nevertheless allow gradual move- thermore, grandiosity may be pathological
ment toward a more accurate self-representa- when asserted aggressively as a way to pre-
tion with a balanced integration of positive vent all good self-representations from
and negative attributes. Although not ad- shifting and suddenly becoming all bad.
dressed in self-perception literature, per se, This defensive style may lead young people
changes in cognitive development probably with narcissistic disturbances to have dis-
influence how others are perceived in an anal- missing or derogatory perceptions of others.
ogous manner. In BPD, young people may lack sufficient in-
One aspect of the cognitive development ternal defenses to prevent self-representa-
of the self can be seen in how children de- tions from alternating frequently between
scribe themselves at different ages. At early the extremes of all good and all bad. Their
ages children usually describe themselves by perception of others similarly alternates be-
their physical characteristics, typical behav- tween extremes of idealization and devalua-
iors, or material possessions. Self-perception tion, thus constituting the clinical phenome-
thus lacks much coherence or integration and non called splitting. Although they remain
self-representation is organized instead unstable, these mental images of self and
around all-or-none thinking (all good or all others are frequently marked by self-con-
bad). Because negative and positive charac- tempt and attributions of malevolence in
teristics are polar opposites, the child cannot others (Bender and Skodol 2007)
recognize that a single person can have both. Among severely abused children, nega-
Given their inability to distinguish real and tive self-perceptions may predominate over
ideal selves, young children typically have positive self-images and lead those children
unrealistically positive self-perceptions that to feel profoundly unworthy and unlovable
often shade into childhood grandiosity. Older (Fischer and Ayoub 1994). In abusive family
children can admit to negative characteristics environments, caregivers typically reinforce
in one domain while retaining a positive self- negative evaluations of the child that are
representation in another. Vacillation between then incorporated into the childs self-repre-
positive and negative self-image in early ado- sentations. As a result, there may be little
lescence is gradually replaced by a more inte- foundation for any cognitive structure of self
grated sense of self and a greater awareness of that would allow the child to develop and in-
the importance of the context to behavior. tegrate both positive and negative self-eval-
Early grandiosity thus subsides as young peo- uations. Furthermore, negative self-evalua-
ple gain the ability to integrate conflicting self- tions may become automatic (Beck et al 2004;
perceptions into a coherent whole in adoles- Siegler 1991) in ways that make them even
cence and early adulthood. more resistant to change.
Inaccurate but age-appropriate self-per- Formal operational thinking, including
ceptions in young children thus may resem- logical and abstract reasoning abilities, nor-
Developmental Issues 127

mally emerges in adolescence (Keating 1990). cies past developmental stages when they
Dramatic increases in differentiation between normally decline. Linehan (1993) empha-
self and other also occur during adolescence sized how invalidating responses from par-
(Bowlby 1973) and thus increase young peo- ents can contribute substantially to the for-
ples ability to view themselves as distinct mation of BPD, and Bezirganian et al. (1993)
from caregivers. Because differentiation facil- have documented that kind of relationship
itates greater autonomy, dependence on pa- empirically.
rental attachment figures normally declines
during adolescence as young people identify Self-Understanding,
more with peer groups intead. At present we Self-Direction, and Identity
know little about how peer relationships in-
fluence personality maturation during ado- Erikson (1968) argued that once young peo-
lescence, but this important social factor may ple gain greater awareness of themselves
reinforce self-perceptions that are more inter- and more accurate perceptions of others,
nally based and less centered around the par- they often experience a normative crisis of
entchild relationship. Although elevated identity during adolescence and early adult-
dependency might not necessarily be patho- hood. This identity crisis is one of eight age-
logical in early adolescence, it may become a specific normative crises that occur in human
symptom of dependent personality disorder development from infancy through old age.
if it persists past late adolescence and into According to Erikson, a crisis is a turning
early adulthood. point when development must move in one
The advent of formal operational think- direction or another. In adolescence, young
ing provides adolescents with a greater ca- people either move toward consolidating a
pacity to evaluate and compare their relation- secure and stable sense of self or they experi-
ships with different attachment figures, not ence diffuse identities that provide limited
just with one another but also against hypo- direction or sense of continuity over time.
thetical ideals. The adolescents ability to con- When the developmental crisis of identity is
sider attachment relationships in the abstract successfully resolved, it normally leads to in-
may bring with it recognition that parents are creased integration of personality. When an
deficient in some ways (Kobak and Cole identity crisis goes unresolved, it may result
1994), perhaps provoking a dismissive rejec- in potentially pathological delays in matura-
tion of the parents or angry preoccupation tion instead.
with their shortcomings. Gains in adolescent Identity consolidation during adoles-
insight into parentchild relationships will cence primarily entails establishing a clear
ideally lead to greater openness, objectivity, sense of self and finding a place in the com-
and flexibility as young people reevaluate at- munity. Identity is broadly defined in Erik-
tachment relationships and attain a more re- sons theory to encompass self-esteem, satis-
alistic and integrated perception of parents. faction with personal and occupational goals,
This developmental process may reduce dis- and confidence in coping skills. Group mem-
missing behaviors or dramatic outbursts of bership and sexual identity represent other
anger that might appear earlier in adoles- important domains in a young persons iden-
cence to be symptoms of narcissism or histri- tity. Identity diffusion, on the other hand, is
onic personality disorder. If parents respond typically expressed in the inability to select
to these adolescent behaviors in maladaptive clear occupational goals or the adoption of
ways based on their own interpersonal dis- roles deviating from conventional social
turbances, they may not facilitate the resolu- norms. Erikson notes that identity diffusion
tion of parentchild conflicts and may instead may at times include delinquent behavior or
reinforce narcissistic and histrionic tenden- psychotic-like symptoms but cautions that
128 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

these disturbances are often transient during Despite the broad overlap between per-
adolescence. Any significant disturbances in sonality disorder symptoms and Eriksons
social and emotional development may act to construct of identity diffusion, relatively lit-
distort or interfere with normative identity tle research has investigated the association
consolidation, perhaps thereby contributing between the two. It thus remains unclear
to the persistence of early personality disor- whether identity disturbances contribute to
der symptoms that otherwise might resolve the emergence of personality disorders in ad-
through normal maturational processes. olescence and later persistence into adult-
As defined by Erikson, identity diffusion hood or if personality disorder symptoms
shares many characteristics with Axis II delay the consolidation of identity. Consis-
symptoms (Cloninger et al. 1993; Kernberg tent with Eriksons epigenetic theory of de-
1975; Taylor and Goritsas 1994). Indeed, velopment, Cluster B symptoms do appear
identity disturbances are explicitly included to interfere with the formation of lasting and
in diagnostic criteria for BPD. Identity dis- committed romantic relationships that repre-
turbances are strongly implied in the sug- sent the key developmental task of early
gestibility to other peoples influence in his- adulthood (Crawford et al. 2004). Identity
trionic personality disorder, idealized but consolidation may occur at a critical stage in
unrealistic self-perceptions in narcissistic the development of personality disorders be-
personality disorder, marked worry about cause it coincides with a period when paren-
other peoples criticism in avoidant person- tal influence declines and youths increas-
ality disorder, and difficulties in making ev- ingly assert their independence. As young
eryday decisions in dependent personality people gradually separate from the family,
disorder. Moreover, identity diffusion and the identity they choose plays that much
personality disorder symptoms share similar greater a role in defining their personality. If
developmental trajectories: both decline adolescents are unable to clearly differentiate
with age during adolescence and early adult- themselves from their parents or to resolve
hood (Johnson et al. 2000a; Meeus et al. 1999). any ongoing disturbances in their relation-
Distinctions between normal identity dif- ship, they are likely to carry internalized ver-
fusion and more enduring personality disor- sions of those difficulties with them in how
der symptoms in adolescence may be diffi- they perceive themselves and others, per-
cult to make in clinical evaluations. Erikson haps even recapitulating those disturbances
noted that it is normal for young people to in new relationships in adulthood. In other
try on different identities during adoles- words, if identity remains poorly differenti-
cence and later abandon them when they fail ated as young people separate from their
to fit comfortably with their sense of identity. family, any corresponding interpersonal dis-
For instance, early attempts to express sexu- turbances may become self-perpetuating
ality may manifest in provocative dress dur- during adulthood.
ing adolescence without necessarily being a
symptom of histrionic personality disorder
(Crawford and Cohen 2007). After trying out
that overt expression of sexuality, young PERSONALITY MATURATION AND
people may subsequently opt for less pro- AXIS II PSYCHOPATHOLOGY
vocative attire that corresponds more with
their internal sense of self. On the other hand, Personality traits reflect a complex adaptive
if provocative dressing co-occurs with poorly system to internal and environmental condi-
regulated affect and maladaptive preoccupa- tions, including changes in affective and cog-
tion with interpersonal relationships, it may nitive structures during development (Caspi
signal a more lasting disturbance of person- 1998; Caspi et al. 2005). Specific affects, be-
ality. haviors, and cognitions that are age appro-
Developmental Issues 129

priate or normative at one stage of personal- ality disorders (paranoid, schizoid, and
ity development may reflect immaturity or schizotypal personality disorders) but also to
psychopathology at subsequent ages. As avoidant personality disorder. Parent re-
young people gain emotional and cognitive ports on individual CBCL items associated
skills, they usually abandon immature ways with Cluster B personality disorders (border-
of experiencing and interacting with the line, histrionic, and narcissistic personality
world around them. On the other hand, disorders and conduct disorder that pre-
when young people continue to experience cedes ASPD) similarly indicate declines in
affects, behaviors, or cognitions that their bragging, showing off, demanding attention,
peers have outgrown, they may encounter getting into fights, lying, cheating, having a
interpersonal difficulties that in some cases hot temper, crying a lot, feeling excessive de-
accumulate over time. Interpersonal difficul- pendence, having problems with peers, and
ties may be traced to various deficits in the experiencing jealousy (Achenbach 1991). De-
development of affect regulation during in- spite broad normative reductions in emo-
fancy, the formation of conscience during tional and behavioral problems, parents re-
early childhood, the establishment of age- port average increases in some childhood
appropriate impulse control in childhood and adolescent symptoms such as being sus-
and adolescence, or the consolidation of picious, secretive, and obsessively preoccu-
identity in late adolescence and early adult- pied with certain thoughts. Furthermore, age
hood. Although deficits or delays in emo- changes in symptoms in normative samples
tional development do not necessarily sig- do not necessarily follow the same trajecto-
nify Axis II pathology, they may indicate that ries as children and adolescents brought in
an individual is on a deviant pathway with for clinical evaluation or treatment (Achen-
increased risk for further maladaptive be- bach 1991). Parents may take normative age
havior. Persistence on a deviant pathway is changes into account when assessing the
related to increasing difficulty in returning to well-being of their children and thus become
a more normal developmental trajectory. concerned only when the expected norma-
Maturational change remains possible, but tive decline does not appear, at least for some
given the organizational function of person- problems.
ality, developmental change will be con- It may be useful to put these changes in
strained by the individuals previous history. symptom levels in the context of changes in
(For a further discussion of developmental the prevalence of Axis I symptoms and dis-
considerations, see Bleiberg 2001; Cohen orders over childhood and adolescence. Dis-
2008; De Clercq and De Fruyt 2007; Freeman ruptive behavior shows a standard inverted
and Reinecke 2007; Geiger and Crick 2001; U-shaped prevalence distribution in a wide
and Kernberg et al. 2000.) range of studies (see Moffitt 2007 for a re-
view), with large increases from childhood to
Emotional and Behavioral Problems adolescence and a sharp drop in young
and Personality Disorder Symptoms adulthood. Based on parent reports, there are
different curvilinear trajectories for mean
Children and adolescents appear to outgrow levels of anxious and depressive symptoms
many problem behaviors that are reflected in for boys and girls over the full age range
current symptom criteria for personality dis- from age 4 to 18 years, with higher rates of
order diagnoses. For instance, parent reports problems for boys in childhood followed by
on the Child Behavior Checklist (CBCL; an adolescent decline (Bongers et al. 2003).
Achenbach 1991) show a significant linear For girls these problems increase until the
decline in withdrawn behavior from age 4 to transitional stage of puberty and then level
age 18 (Bongers et al. 2003). Withdrawn be- off. Using teachers as informants in a large
havior pertains not just to Cluster A person- national epidemiological study, McDermott
130 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

(1996) found means on every symptom clus- age correlations were sometimes different in
ter changing with age between ages 5 and 17 direction in this study, with some neuroti-
years, often quite differently for males and cism facets increasing with age, some stable
females. over the age span, and some declining.
Although knowledge of normative age
Normal and Abnormal changes in personality disorder symptoms is
Personality Traits limited, the available data clearly show that
nearly every Axis II disorder has a gradual
Many researchers view personality disorder linear decrease in average symptom levels
symptoms as extreme variants of personality between ages 10 and 35. When tracked across
traits that are continuously distributed in the 20 years in approximately 800 youths in the
population (Costa and Widiger 2002; Livesley Children in the Community random sample
and Jang 2000). We currently have only par- (Cohen et al. 2005; Crawford et al. 2005),
tial information on developmental changes in mean personality disorder symptoms were
trait levels and no information in childhood highest in early adolescence (Johnson et al.
because of problems in conceptualizing and 2000a) and steadily declined with age well
measuring personality before adolescence into adulthood (Crawford et al. 2006). Since
(Shiner and Caspi 2003). mean levels are higher in early adolescence,
Perhaps the most complete information more youths are likely to meet fixed diagnos-
available on age changes is based on the five- tic criteria that do not take changing age
factor model (FFM) and its component facets norms into account, thereby possibly in-
that attempt to measure variation in the full creasing the rate of false-positive diagnoses
normative range of personality. These broad during adolescence. These data suggest that
dimensions and more narrowly defined fac- age-specific norms may be desirable for an
ets have also been proposed as a way of un- assessment instrument in this age range.
derstanding and potentially measuring per- However, they also make clear that the prob-
sonality disorder (Costa and Widiger 2002). lem of changing normative symptom levels
The dimension labeled neuroticism is espe- is not limited to childhood and adolescence.
cially pertinent because it is hypothesized to Despite changes in symptom level norms,
reflect many of the criteria for personality adolescent psychiatric disorders warrant
disorder. Age changes in self-reported scores clinical attention even in developmental pe-
on the NEO Personality InventoryRevised riods when they are most prevalent.
(NEO-PI-R; Costa and McCrae 1992) were On the whole, normative data do not
evaluated in a longitudinal sample of gifted show any clear congruence between average
children between age 12 and 16 years and in age changes in normal personality dimen-
a much larger cross-sectional sample of sions measured on the NEO-PI-R and nor-
Flemish adolescents between age 14 and 18 mative changes observed using personality
years (McCrae et al. 2002). Most age changes disorder measures based on combined youth
were very small and did not follow previ- and parent reports or parent-reported mea-
ously established adult trajectories that show sures of clinically relevant emotional and be-
a gradual decline in neuroticism from the havior symptoms. Some of the discrepancy
college years to age 30 years (e.g., McCrae et may reflect how the NEO-PI-R assesses per-
al. 1999). When evaluated in younger adoles- sonality traits in the normal range and may
cent samples, neuroticism showed an ele- be limited by ceiling effects at the extreme
vation in girls up to about age 14 years and range of functioning assessed by personality
stability thereafter, and there were no signif- disorder measures or the CBCL. Once again,
icant age changes in boys. Within each the direction of age-related changes in symp-
broader factor of the FFM, individual facet toms in normative community samples may
Developmental Issues 131

not always correspond to age-related change mental interactions also may be an important
in symptoms in children from clinical popu- source of variation in early childhood, when
lations. strong behavioral habits have not yet become
firmly established. There may be more error
in observation-based measures of tempera-
CHANGE AND STABILITY IN ment in infants and preschool children than
in older children due to their greater reactiv-
CHILDHOOD ity to fatigue, hunger, and other temporary
influences. Increased measurement error
Given our knowledge that the prevalence of
thus may contribute to lower stability esti-
particular behaviors changes with age, what
mates in younger children.
can be said about the correlational or rank-
An additional problem has been an ab-
order stability of personality or its tempera-
sence of consensus on how to define the ma-
mental precursors over childhood and ado-
jor temperament dimensions, an issue that
lescence? In particular, to what extent may
has only recently begun to be resolved. Some
we expect that the same individuals who
dimensions of temperament and the mea-
manifest the most extreme personality prob-
sures devised to assess them have been theo-
lems at one age will be among those who do
retically derived (Rothbart et al. 2001; Telle-
so at another age?
gen 1985), including predicted relationships
with specific personality disorders (Clonin-
Temperament ger 1987; Cloninger et al. 1993). However,
Temperament is often regarded as an early theoretically derived dimension names
precursor to personality (Shiner 2005) that sometimes do not clearly correspond to the
reflects basic biological differences in child- content of items. Despite an array of unique
hood characteristics such as activity level, construct-derived names, certain dimen-
fearful withdrawal, ability to be soothed, re- sions of temperament can be viewed in the
sponsiveness to stimuli, and affective inten- frame of the FFM (Shiner and Caspi 2003).
sity. Temperament is typically measured by This frame has the advantage of uniting the
observational ratings or maternal report. personality and temperament literatures but
Cloninger and his colleagues have devel- does only partial justice to the original con-
oped a measure assessing temperament di- ceptions of the important individual differ-
mensions in preschool children that are pos- ences in early childhood such as executive
ited to be related to later personality disorder control and emotional reactivity. Research
(Constantino et al. 2002). Mervielde et al. linking temperament measured in early
(2005) and De Clercq et al. (2006) have also childhood to later personality disorder is
conceptualized and developed measures to only beginning to appear (Constantino et al.
track continuities between childhood tem- 2002).
perament, personality, and personality pa-
thology. Personality Dimensions
In general, temperament shows signifi-
cant but low stability in early childhood There is clear evidence that, on average, a
(Rothbart and Bates 1998) for reasons that personality dimension assessed by a self-
may be intrinsic to the developmental pro- report instrument will show lower correla-
cess. For example, there may be effects spe- tion over equivalent time for younger per-
cific to maturational levels due to genetic or sons than for older persons (Roberts and
other constitutional influences. The impact DelVecchio 2000). Because instability in per-
of contextual factors may vary at different sonality dimensions continues throughout
maturational levels. Constitutionalenviron- life, correlations measuring stability reach a
132 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

maximum at about age 50 years. Stabilities 2005). Of adolescents diagnosed with Cluster
for symptom measures of personality disor- A disorders in a community sample, 35%
der are very likely to show a similar pattern. had disruptive disorders, 25% had anxiety
Although we have no preadolescent data, disorders, and 20% had depressive disorders
correlations measuring stability are moder- (Kasen et al. 1999). A full 47% of those with
ate in adolescence (Crawford et al. 2001a), Cluster B disorders had a comorbid disrup-
and it is not clear that they increase from ad- tive behavior disorder, 38% had an anxiety
olescence into young adulthood (Johnson et disorder, and 28% had a depressive disorder.
al. 2000a). General stability may not be the is- Over half (51%) of those with Cluster C diag-
sue if disorders are outcomes of geneenvi- noses had an anxiety disorder, 34% had a dis-
ronment interactions, so that it is expected ruptive behavior disorder, and 23% had a de-
that there may be lower stability on extreme pressive disorder.
scores than over the full range of relevant di- Co-occurrence of Axis I and Axis II disor-
mensions (DiLalla et al. 2000). ders is important because of its impact on
Axis II diagnoses, on the other hand, tend prognosis. When adolescent Cluster A disor-
to be much less stable in clinical and commu- ders co-occurred with disruptive behavior
nity samples of adolescents (Bernstein et al. disorders, the risk for a subsequent Cluster A
1993;Mattanah et al. 1995). Axis II diagnoses diagnosis in early adulthood was 24.6 times
in adults similarly tend to be unstable even in higher than the risk for adolescents with no
clinical samples (Shea et al. 2002; Zanarini et psychiatric disorder (Kasen et al. 1999). By
al. 2003). This instability is likely to be a con- comparison, the risk for Cluster A disorders
sequence not only of the generally poorer in early adulthood was only 2.76 times higher
measurement quality of dichotomized con- when Cluster A disorders in adolescence
tinuous measures (MacCallum et al. 2002) were diagnosed without any disruptive be-
but also of somewhat arbitrary decisions havior disorder. Similarly, the risk for Cluster
about diagnostic cut-points. Despite the rela- B diagnoses in early adulthood was 19 times
tive instability of categorically defined diag- higher when earlier Cluster B disorders co-
noses, there is increasing evidence of long- occurred with major depression, but only 3.83
term impairment and poor prognosis associ- times higher when these personality disor-
ated with adolescent personality disorder or ders occurred without major depression.
a high level of personality disorder symp- Recent research shows that the effects of
toms independent of Axis I disorders or comorbid Axis I and Axis II disorders con-
symptoms (Chen et al. 2006a, 2006b; Craw- tinue well beyond early adulthood. Com-
ford et al. 2008; Johnson et al. 1999, 2000b; pared with adolescents with no psychiatric
Kasen et al. 1999, 2007; Lofgren et al. 1991; disorder, adolescents with diagnoses on both
Rey et al. 1997; Skodol et al. 2007; Winograd axes had an almost nine-fold increase in risk
et al. 2008). We are thus accumulating evi- for psychiatric disorders 20 years later, when
dence that the criteria for adult disorders assessed at mean age 33 years (Crawford et
may be useful indicators of Axis II pathology al. 2008). Adolescents with co-occurring Axis
even in young samples and not necessarily I and II disorders also were more likely to
less predictive at times when they are more have been in treatment or take psychotropic
normative (also see Cohen et al. 2007; Ehren- drugs when subsequently assessed as adults
saft et al. 2006; Kasen et al. 2001). (Kasen et al. 2007). Impairment and lower
functioning were still evident 20 years later
Comorbidity With Axis I Disorders among adolescents with Axis II diagnoses
alone, but prognosis was consistently worse
Axis II disorders in adolescence frequently for academic, occupational, interpersonal,
co-occur with Axis I disorders (Cohen et al. and psychiatric functioning when Axis I and
Developmental Issues 133

Axis II disorders co-occurred in adolescence peared to be reinforced by defensive re-

(Crawford et al. 2008). Given effects that per- actions to one parents overinvolvement
sisted across two decades, it is important for and also by identification with the other
parents angry and dismissive interper-
clinicians to evaluate and treat adolescent
sonal style. Despite poor social adjust-
personality disorders, especially when they ment with peers, this youth nevertheless
co-occur with Axis I disorders. reported having a positive relationship
with his parents. Given this protective
Case Examples factor and an absence of key risk factors
such as childhood trauma, his personal-
Given what we know about developmental ity disorder symptoms may well sub-
side over time as he gains maturity.
trajectories of personality disorder symp-
Treatment in this case focused on assist-
toms in adolescents, it appears unwise to
ing the youth to regulate angry affect
make categorical diagnoses during initial more adaptively and to articulate how
clinical assessment except perhaps in ex- peers upset him so much. Treatment
treme cases. Nevertheless, it is meaningful to also helped the parents to become more
consider youths at risk for subsequent per- aware of how their own anxiety and
sonality disorder based on how they present mood disturbances contributed to mal-
for treatment and based on collateral risk fac- adaptive parentchild interactions in
ways that inhibited the youths gradual
tors in close family members, as illustrated
movement toward independent adult
by the following vignette. functioning.

A 15-year-old boy from an intact family Just as maturational factors appear to

was brought for individual psychother- confound the assessment of personality dis-
apy to address uncontrolled anger, a
orders in adolescents, they may also obscure
pervasive hatred toward age peers, and
oppositional and self-defeating behav- the presence of personality disorder in adults
ior (e.g., threatening to drop out of high who have outgrown some earlier manifesta-
school). His self-image reflected grandi- tions of the disorder. Knowledge of norma-
osity or self-contempt depending on tive trajectories of personality disorder thus
different social contexts, and he tended can inform the assessment and treatment of
to be avoidant and dismissive of others. personality disturbances in adult patients
During treatment it became evident
whose symptoms fall short of current diag-
that one parent tried to cope with exces-
sive anxiety by becoming overinvolved nostic criteria:
in the patients long-term plans in ways
that provoked stubborn and opposi- A 42-year-old woman presented for
tional behavior. The other parent had treatment with complaints of loneliness
recurrent depressive episodes, an ex- and an enduring inability to establish a
plosive and unpredictable temper, stable romantic relationship. She re-
identity diffusion, and a dismissive in- called adolescence and early adulthood
terpersonal style. as p erio ds characterized by labile
On initial evaluation, this 15-year- mood, frequent abandonment fears,
old youth appeared at increased risk for volatile and unpredictable romantic
lasting BPD based on persistent symp- relationships, reactive anger toward
toms of uncontrolled anger, marked an- parents and peers, and a series of sui-
tipathy for others, and poorly integrated cidal gestures. During childhood she
representations of self and others. The witnessed violent conflicts between
long-term risk may be compounded by mother and father before they divorced
the presence of parental anxiety and and experienced significant emotional
mood disturbances, especially as they neglect afterward. When presenting for
are woven into parentchild relation- treatment, this patient denied any
ships. Personality disturbances thus ap- mood lability or suicidal ideation and
134 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

functioned well at work. However, she diagnosis of personality disorders might lead
experienced abandonment fears, reac- to premature labeling.
tive anger, and mood lability whenever
she became involved in new romantic
relationships, thereby prompting a gen-
eral avoidance of close relationships CHOICE OF INFORMANT
and reinforcing unwanted feelings of
loneliness. In clinical assessment of adults the patient is
usually the primary informant, although the
Even allowing for inaccuracies that dis- corroboration and independent perspectives
tort retrospective clinical reports, this patient of knowledgeable sources such as family
undoubtedly met full criteria for BPD during members are welcome and often useful. Re-
adolescence and early adulthood. Although search measures designed for use with adults
her symptoms have since fallen below diag- in clinical or other samples generally are con-
nostic threshold, either due to maturational fined to self-report. In the child mental health
factors or the effects of prior individual psy- field, there is still little consensus about which
chotherapies, she nevertheless continues to informants should be considered primary
experience significant borderline psychopa- and which should be seen as auxiliary at dif-
thology in ways that become painfully ap- ferent points in childhood development. It is
parent whenever new romantic attachments often assumed that the primary caretaker
evoke long-standing abandonment fears she should be the principal informant for pre-
otherwise seeks to avoid. Treatment in this school children. As such, instruments mea-
case focused on clarifying and then reducing suring temperament, personality, and psy-
abandonment fears and addressing how her chopathology in children younger than age 9
reactive anger undermined the stability of or 10 years are most often based on informa-
her romantic relationships. Treatment also tion supplied by parentsfor example, the
addressed the disorganizing effects of child- Childrens Behavior Questionnaire (Rothbart
hood trauma and neglect. et al. 2001), the Personality Inventory for Chil-
dren (PIC-2; Lachar 1999a), and the CBCL. For
young children in preschool or day care,
WHAT IS THE BEST MEASUREMENT teachers or childcare leaders may provide the
best data based on greater familiarity with
STRATEGY FOR ASSESSING EARLY normative patterns of behavior for the age.
PERSONALITY DISORDERS? Similarly, parents and teachers are both
thought to provide relevant data for older
Three overall issues pose problems for the as- children in elementary school, although
sessment of personality disorders in children agreement on the childs problems is often
and adolescents. First, DSM-IV-TR does not poor across these informants (Achenbach et
take into account normative developmental al. 1987). Sometime in childhood, at least by
changes in the prevalence of certain problems age 9 or 10, most children are able to provide
and the consequent possibility that certain di- data on their own characteristics and prob-
agnostic criteria should not be seen as abnor- lems. For instance, the youth-reported ver-
mal behavior at some ages. Second, there is a sion of the Diagnostic Interview Schedule for
shortage of appropriate and validated diag- Children Version IV (Shaffer et al. 2000) as-
nostic instruments for children and adoles- sesses psychopathology in children as young
cents and unresolved questions about the as age 9 years. Similarly, the Personality In-
best measurement strategy. Third, aspects of ventory for Youth (PIY; Lachar and Gruber
personality are less stable in childhood and 1995) uses self-reports to assess personality
adolescence, thus raising concern that early in children starting at the same age. Most
Developmental Issues 135

self-report measures of normal personality, personality disorder instruments for adults

however, are used only with adolescents and have been specifically adapted to be age ap-
adults (Shiner and Caspi 2003). propriate for child or adolescent respon-
Agreement between parent and offspring dents. Among self- and parent-reported
on youth behaviors and problems is often instruments, Axis II scales are typically com-
poor (Achenbach et al. 1987). Age changes in bined with various measures of Axis I distur-
prevalence based on youth self-report look bances and thus facilitate the assessment of
different from those based on parent report or co-occurrence between these psychiatric con-
teacher report. Agreement between self and structs.
informant reports on personality disorder
measures for adults is similarly poor (Klon- Structured Clinical Interviews
sky et al. 2002). Despite these difficulties, the
research field generally agrees that all infor- Of the structured interviews designed to as-
mants add usefully to the assessment of Axis sess DSM-defined personality disorders in
I disorders in children and adolescents. As a adults, the Personality Disorder Examina-
consequence, the preferred strategy is to ob- tion (Loranger 1988) has been most thor-
tain data from at least two informants and oughly evaluated and appears to be a valid
consider any symptomatic report to be valid measure of Axis II disturbance in adoles-
providing there is evidence of associated im- cents. Nevertheless, more work is needed to
pairment. However, evidence indicates that identify age-related differences in adolescent
adolescents are better informants about emo- and adult manifestations of personality dis-
tions and often acknowledge disruptive or orders. In a longitudinal comparison of ado-
antisocial behaviors that may be unknown to lescent and adult inpatient samples, person-
the parent when only a single informant is ality disorders assessed with the Personality
used. Parents or teachers are thought to be Disorder Examination were less stable over a
better informants on issues where normative 2-year interval in adolescents than in adults
comparisons are relevant. (Mattanah et al. 1995). However, threshold
Where does this leave us when deciding effects often add unreliability to stability es-
which informants to assess for personality timates of categorically defined personality
disturbances in children and adolescents? At disorders, thus making comparisons across
present we do not know whether diagnostic age groups more difficult to interpret. When
criteria for Axis II are intrinsically more dif- assessed as dimensional constructs, stability
ficult to assess in youth than most criteria for estimates for personality disorders usually
Axis I. All things considered, it appears pru- appear higher than when assessed as cate-
dent to gather data from multiple sources gorical constructs.
whenever possible when assessing children
and adolescents. Self-Report Instruments
Although self-report instruments are easier
and more cost-efficient to administer than
structured interviews, questions are raised
ADOLESCENT PERSONALITY about whether respondents have sufficient
DISORDER self-awareness or willingness to acknowl-
edge Axis II symptoms that might stigmatize
As interest in early Axis II disturbances in them. Given problems in setting reliable
young people has grown, a variety of instru- thresholds, the available instruments for
ments designed to measure normal and ab- children and adolescents tend to assess Axis
normal personality in adults have been eval- II symptoms using continuous scales instead
uated for use in adolescent samples. Certain of making formal diagnoses.
136 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

The Millon Adolescent Clinical Inventory disturbances. Despite substantial similarity

(MACI) is a well-known instrument modeled on the PIY and PIC-2, youth and parent in-
on the Millon Clinical Multiaxial Inventory formants show moderate agreement (me-
designed for adults (see Davis et al. 1999). In- dian correlation= 0.43, range 0.280.53) (La-
tended for adolescents as young as age 13, the char 1999b). These concordance rates appear
MACI uses 160 self-report items to measure better than the 0.25 correlation between
personality disorder constructs congruent youth and parent reports for comparable age
with DSM-defined personality disorders but groups on the CBCL (Achenbach et al. 1987).
also reflecting Millons (1990) theory of per-
sonality. The MACI thus measures 12 person- Additional Instruments for
ality styles labeled Introversive, Inhibited,
Doleful, Submissive, Dramatizing, Egotistic,
Personality Disorders
Unruly, Forceful, Conforming, Oppositional,
Self-Demeaning, and Borderline Tendency. The Adolescent Psychopathology Scale
Computer-generated scores on the MACI (Reynolds 1998) is a self-report measure de-
make adjustments for age and gender differ- signed for adolescents age 1219 years. It
ences in Axis II disturbances based on norms measures five of the 10 DSM-IV personality
from separate samples of normal and dis- disorders (borderline, avoidant, obsessive-
turbed adolescents. Standardized scores are compulsive, paranoid, and schizotypal per-
further adjusted to take estimated base rates sonality disorders) along with conduct disor-
of psych opathology into account ev en der as the childhood precursor of ASPD. The
though the prevalence of Axis II disturbances Coolidge Personality and Neuropsychologi-
in adolescents has yet to be established. These cal Inventory for Children (CPNI; Coolidge
built-in adjustments are not readily transpar- 1998) assesses DSM-IV Axis II disorders in
ent and effectively preclude their use for in- children and adolescents from age 517 years
vestigating the population prevalence of ado- by parent report. The CPNI assesses symp-
lescent Axis II disturbances or for assessing toms of all DSM-IV personality disorders as
developmental change. well as conduct disorder symptoms. The
Shedler-Westen Assessment Procedure200
Parallel Parent- and for Adolescents (SWAP-200-A; Westen et al.
Youth-Reported Instruments 2003, 2005) is a Q-sort instrument designed
for use by skilled clinical observers to assess
The PIC-2 and PIY were both originally mod- Axis II pathology in adolescent patients they
eled on the Minnesota Multiphasic Personal- see in treatment. Q-sort is a method by which
ity Inventory and thus do not correspond di- items are arrayed by the clinician from most
rectly to DSM-IV (American Psychiatric descriptive of the adolescent to least descrip-
Association 1994) Axis II disorders. The par- tive of the adolescent. This measure was
ent-reported PIC-2 uses 275 forced-choice adapted from the Shedler-Westen Assess-
items to measure constructs labeled Cogni- ment Procedure200, a Q-sort designed for
tive Impairment, Impulsivity and Distracti- adults that has shown evidence of validity,
bility, Delinquency, Family Dysfunction, reliability, and utility in taxonomic research
Reality Distortion, Somatic Concern, Psycho- with adult samples (e.g., Shedler and Westen
logical Discomfort, Social Withdrawal, and 1998). At present none of the available instru-
Social Skills Deficits. The youth-reported PIY ments has demonstrated clear superiority in
uses 270 forced-choice items to measure the clinical and research applications, and there
same constructs. As suggested by the labels, are limited validity data available for newly
the PIY and PIC-2 measure constructs that developed scales measuring DSM-IV and
probably tap a mixture of Axis I and Axis II DSM-IV-TR personality disorders.
Developmental Issues 137

Since the late 1980s, there has been accumu- Achenbach TM: Manual for the Child Behavior
Checklist, Ages 418 and 1991 Profile. Burling-
lating evidence that clinically meaningful
ton, VT, University of Vermont, Department of
personality disorders occur in adolescents. Psychiatry, 1991
Adolescent personality disorders are associ- Achenbach TM, McConaughy SH, Howell CT:
ated with emotional distress and psychoso- Child/adolescent behavioral and emotional
cial impairment in community samples problems: implications of cross-informant cor-
relations for situational specificity. Psychol
(Bernstein et al. 1993, 1996; Golombek et al.
Bull 101:213232, 1987
1987; Lewinsohn et al. 1997; Marton et al. Ainsworth MDS, Blehar MC, Waters E, et al: Pat-
1987; Rogosch and Cicchetti 2005; Stein et al. terns of Attachment: A Psychological Study of
1987) and clinical samples in inpatient and the Strange Situation. Hillsdale, NJ, Erlbaum,
outpatient settings (Becker and Grilo 2006; 1978
Brent et al. 1994; Grilo et al. 1996; Pinto et al. Alexander PC: Application of attachment theory
to the study of sexual abuse. J Consult Clin
1996; Westen et al. 2003). When adolescent
Psychol 60:185195, 1992
and adult personality disorders are com- American Psychiatric Association: Diagnostic and
pared, many similarities can be observed. Statistical Manual of Mental Disorders, 4th
Just as adult personality disorders are associ- Edition. Washington, DC, American Psychiat-
ated with co-occurring Axis I disturbances ric Association, 1994
American Psychiatric Association: Diagnostic and
(Bienvenu and Stein 2003; Skodol 2005), so
Statistical Manual of Mental Disorders, 4th
have ad olescent personality disorders Edition, Text Revision. Washington, DC,
(Becker et al. 2000; Crawford et al. 2001a, American Psychiatric Association, 2000
2008). However, more work is needed on age- Beck AT, Freeman A, Davis DD, et al: Cognitive
related differences in adolescent and adult Therapy of Personality Disorders, 2nd Edi-
tion. New York, Guilford Press, 2004
manifestations of personality disorders.
Becker DF, Grilo CM: Validation studies of the bor-
Although elevated personality disorder derline personality disorder construct in ado-
scores in adolescence represent a risk for later lescents. Adolesc Psychiatry 29:217235, 2006
psychiatric and psychosocial disturbances, Becker DF, Grilo CM, Edell WS, et al: Comorbidity
early personality disorder symptoms are of borderline personality disorder with other
likely to decline over time. Caution is war- personality disorders in hospitalized adoles-
cents and adults. Am J Psychiatry 157:2011
ranted in evaluating the clinical significance 2016, 2000
of symptoms during adolescence. Yet, there is Bender DS, Skodol AE: Borderline personality as
almost universal agreement that prevention self-other representational disturbance. J Per-
of mental disorders is best accomplished at a sonal Disord 21:500517, 2007
very young age and in collaboration with Bernstein DP, Cohen P, Velez CN, et al: Prevalence
and stability of the DSM-III-R personality dis-
parents, particularly when other risk factors
orders in a community-based survey of ado-
are also present. Therefore, children with ele- lescents. Am J Psychiatry 150:12371243, 1993
vated symptoms or their precursors are Bernstein DP, Cohen P, Skodol A, et al: Childhood
prime targets for secondary preventionthat antecedents of adolescent personality disor-
is, prevention of further developmental de- ders. Am J Psychiatry 153:907913, 1996
Bezirganian S, Cohen P, Brook JS: The impact of
lays or elevation to frank disorder. Regardless
mother-child interaction on the development
of how childhood and adolescent personality of borderline personality disorder. Am J Psy-
disorders are definedeither as an early ver- chiatry 150:18361842, 1993
sion of adult disorder or an early indicator of Bienvenu OJ, Stein MB: Personality and anxiety
elevated risk of adult disorderit appears disorders: a review. J Personal Disord 17:139
151, 2003
appropriate to consider interventions to alle-
Bleiberg E: Treating Personality Disorders in Chil-
viate Axis II disturbances when they manifest dren and Adolescents: A Relational Approach.
in young people. New York, Guilford, 2001
138 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

Bongers IL, Koot HM, van der Ende J, et al: The sonality to early childhood. Psychiatr Res
normative development of child and adoles- 109:229244, 2002
cent problem behavior. J Abnorm Psychol Coolidge FL: Coolidge Personality and Neuropsy-
112:179192, 2003 chological Inventory for Children Manual:
Bowlby J: Attachment and Loss, Vol 1: Attach- CPNI. Colorado Springs, CO, FL Coolidge,
ment. New York, Basic Books, 1969 1998
Bowlby J: Attachment and Loss, Vol 2: Separation. Coolidge FL, Thede LL, Jang KL: Heritability of
New York, Basic Books, 1973 personality disorders in childhood: a prelimi-
Brent DA, Johnson B, Perper J, et al: Personality nary investigation. J Personal Disord 15:3340,
disorder, personality traits, impulsive vio- 2001
lence and completed suicide in adolescents. Costa PT Jr, McCrae RR: Revised NEO Personality
J Am Acad Child Adolesc Psychiatry 33:1080 Inventory (NEO-PI-R) and NEO Five-Factor
1086, 1994 Inventory (NEO-FFI) Professional Manual.
Caspi A: Personality development across the life Odessa, FL, Psychological Assessment Re-
course, in Handbook of Child Psychology, Vol sources, 1992
3: Social, Emotional, and Personality Develop- Costa PT Jr, Widiger TA (eds): Personality Disor-
ment, 5th Edition. Edited by Damon W, Eisen- ders and the Five-Factor Model of Personality,
berg N. New York, Wiley, 1998, pp 311388 2nd Edition. Washington, DC, American Psy-
Caspi A, Roberts BW, Shiner RL: Personality de- chological Association, 2002
velopment: stability and Change. Ann Rev Crawford TN, Cohen P: Histrionic personality
Psychology 56: 453484, 2005 disorder, in Personality Disorders in Children
Cassidy J, Shaver PR: Handbook of Attachment: and Adolescents. Edited by Freeman A, Rei-
Theory, Research, and Clinical Applications, necke M. New York, John Wiley & Sons, 2007,
2nd Edition. New York, Guilford, 2008 pp 495532
Chen H, Cohen P, Kasen S, et al: Adolescent Axis Crawford TN, Cohen P, Brook JS: Dramatic erratic
I and personality disorders predict quality of personality disorder symptoms, I: continuity
life during young adulthood. J Adolesc Health from early adolescence into adulthood. J Per-
39:1419, 2006a sonal Disord 15:319335, 2001a
Chen H, Cohen P, Kasen S, et al: Impact of adoles- Crawford TN, Cohen P, Brook JS: Dramatic erratic
cent mental disorders and physical illnesses personality disorder symptoms, II: develop-
on quality of life 17 years later. Arch Pediatr mental pathways from early adolescence to
Adolesc Med 160:9399, 2006b adulthood. J Personal Disord 15:336350,
Cloninger CR: A systematic method for clinical 2001b
description and classification of personality Crawford TN, Cohen P, Johnson JG, et al: The
variants: a proposal. Arch Gen Psychiatry course and psychosocial correlates of person-
44:573588, 1987 ality disorder symptoms in adolescence: Erik-
Cloninger CR: Genetics, in The American Psychi- sons developmental theory revisited. J Youth
atric Publishing Textbook of Personality Dis- Adolesc 33:373387, 2004
orders. Edited by Oldham JM, Skodol AE, Crawford TN, Cohen P, Johnson JG, et al: Self-re-
Bender D. Washington, DC, American Psychi- ported personality disorder in the Children in
atric Publishing, 2005, pp 143154 the Community sample: convergent validity
Cloninger CR, Svrakic DM, Przybeck TR: A psy- and prospective validity in late adolescence
chobiological model of temperament and char- and adulthood. J Personal Disord 19:3052,
acter. Arch Gen Psychiatry 50:975990, 1993 2005
Cohen P: Child development and personality dis- Crawford TN, Shaver PR, Cohen P et al: Self-re-
order. Psychiatr Clin North Am 31:477493, ported attachment, interpersonal aggression,
vii, 2008 and personality disorder in a prospective com-
Cohen P, Crawford TN, Johnson JG, et al: The munity sample of adolescents and adults.
Children in the Community Study of develop- J Personal Disord 20:331353, 2006
mental course of personality disorder. J Per- Crawford TN, Cohen P, First MB, et al: Comorbid
sonal Disord 19:466486, 2005 Axis I and Axis II disorders in early adoles-
Cohen P, Chen H, Crawford TN, et al: Personality cence: outcomes 20 years later. Arch Gen Psy-
disorders in early adolescence and the devel- chiatry 65:641648, 2008
opment of later substance use disorders in the Davis RD, Woodward M, Goncalves A, et al: Treat-
general population. Drug and Alcohol Addic- ment planning and outcome in adults: the Mil-
tion 88S:S71S84, 2007 lon Clinical Multiaxial Inventory-III, in The
Constantino JN, Cloninger CR, Clarke AR, et al: Use of Psychological Testing for Treatment
Application of the seven-factor model of per- Planning and Outcomes Assessment, 2nd Edi-
Developmental Issues 139

tion. Edited by Maruish ME. Mahwah, NJ, Er- Jang KL: The Behavioral Genetics of Psychopa-
lbaum, 1999, pp 10511081 thology: A Clinical Guide. Mahwah, NJ, Er-
De Clercq B, De Fruyt: Childhood antecedents of lbaum, 2005
personality disorder. Curr Opin Psychiatry Johnson JG, Cohen P, Skodol AE, et al: Personality
20:5761, 2007 disorders in adolescence and risk of major
De Clercq B, De Fruyt F, Van Leeuwen K, et al: The mental disorders and suicidality during adult-
structure of maladaptive personality traits in hood. Arch Gen Psychiatry 56: 805811, 1999
childhood: a step toward an integrative devel- Johnson JG, Cohen P, Kasen S, et al: Age-related
opmental perspective for DSM-V. J Abnorm change in personality disorder trait levels be-
Psychol 115:639657 2006 tween early adolescence and adulthood: a
DiLalla DL, Gottesman II, Carey G: Madness be- community-based longitudinal investigation.
yond the threshold? Associations between Acta Psychiatr Scand 102:265275, 2000a
personality and psychopathology, in Temper- Johnson JG, Cohen P, Smailes E, et al: Adolescent
ament and Personality Development Across personality disorders associated with violence
the Life Span. Edited by Molfese VJ, Molfese and criminal behavior during adolescence and
DL. Mahwah, NJ, Erlbaum, 2000, pp 177210 early adulthood. Am J Psychiatry 157:1406
Ehrensaft M, Cohen P, Johnson JG: Development 1412, 2000b
of personality disorder symptoms and the risk Kasen S, Cohen P, Skodol AE, et al: Influence of
for partner violence. J Abnormal Psychol child and adolescent psychiatric disorders on
115:474483, 2006 young adult personality disorder. Am J Psy-
Erikson EH: Identity: Youth and Crisis. New York, chiatry 156:15291535, 1999
WW Norton, 1968 Kasen S, Cohen P, Skodol AE, et al: Childhood de-
Fischer KW, Ayoub C: Affective splitting and disso- pression and adult personality disorder: alter-
ciation in normal and maltreated children: de- native pathways of continuity. Arch Gen Psy-
velopmental pathways for self in relationships, chiatry 58:231236, 2001
in Disorders and Dysfunctions of the Self: Kasen S, Cohen P, Skodol AE, et al: Comorbid per-
Rochester Symposium on Developmental Psy- sonality disorder and treatment use in a com-
chopathology, Vol 5. Edited by Cicchetti D, Toth munity sample of youths: a 20-year follow-up.
SL. New York, Plenum Press, 1994, pp 149222 Acta Psychiatr Scand 115:5665, 2007
Fonagy P, Target M, Gergely G: Attachment and Keating D: Adolescent thinking, in At the Thresh-
borderline personality disorder: a theory and old: The Developing Adolescent. Edited by
some evidence. Psychiatr Clin North Am Feldman SS, Elliott G. Cambridge, MA, Har-
23:103122, 2000 vard University Press, 1990, pp 5490
Fonagy P, Target M, Gergely G, et al: The develop- Kernberg OF: Borderline Conditions and Patho-
mental roots of borderline personality disor- logical Narcissism. Northvale, NJ, Jason Aron-
der in early attachment relationships: a theory son, 1975
and some evidence. Psychoanalytic Inquiry Kernberg PF, Weiner AS, Bardenstein KK: Person-
23:412459, 2003 ality Disorders in Children and Adolescents.
Freeman A, Reinecke M (eds): Personality Disor- New York, Basic Books, 2000
ders in Children and Adolescents. New York, Klonsky ED, Oltmanns TF, Turkheirmer E: Infor-
John Wiley & Sons, 2007 mant reports of personality disorders: relation
Geiger TC, Crick NR: A developmental psychopa- to self-reports and future research directions.
thology perspective on vulnerability to per- Clinical Psychology: Science and Practice
sonality disorders, in Vulnerability to Psycho- 9:300311, 2002
pathology. Edited by Ingram RE, Price JM. Kobak R, Cole C: Attachment and metamonitor-
New York, Guilford, 2001, pp 57102 ing: implications for adolescent autonomy
Golombek H, Marton P, Stein B, et al: Personality and psychopathology, in Rochester Sympo-
functioning status during early and middle ad- sium on Development and Psychopathology,
olescence. Adolesc Psychiatry 14:365377, 1987 Vol 5: Disorders of the Self. Edited by Cicchetti
Grilo CM, Becker DF, Fehon DC, et al: Gender dif- D. Rochester, NY, University of Rochester
ferences in personality disorders in psychiatri- Press, 1994, pp 267297
cally hospitalized adolescents. Am J Psychia- Lachar D: Personality Inventory for Children-2
try 153:10891091, 1996 (PIC-2) Manual. Los Angeles, CA, Western
Harter S: The development of self-representa- Psychological Services, 1999a
tions, in Handbook of Child Psychology, Vol 3: Lachar D: Personality Inventory for Children, 2nd
Social, Emotional, and Personality Develop- Edition (PIC-2), Personality Inventory for
ment, 5th Edition. Edited by Damon W, Eisen- Youth (PIY), and Student Behavior Survey
berg N. New York, Wiley, 1998, pp 553617 (SBS), in The Use of Psychological Testing for
140 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

Treatment Planning and Outcomes Assess- literature and longitudinal analysis. Dev Rev
ment, 2nd Edition. Edited by Maruish ME. 19:419461, 1999
Mahwah, NJ, Lawrence Erlbaum Associates, Mervielde I, De Clercq B, De Fruyt F, et al: Tem-
1999b, pp 399427 perament, personality, and developmental
Lachar D, Gruber CP: Personality Inventory for psychopathology as childhood antecedents of
Youth (PIY) Manual. Los Angeles, CA, West- personality disorders. J Person Disord 19:171
ern Psychological Services, 1995 201, 2005
Levy KN: The implications of attachment theory Mikulincer M, Shaver PR, Pereg D: Attachment
and research for understanding borderline theory and affect regulation: the dynamics, de-
personality disorder. Developmental Psycho- velopment, and cognitive consequences of at-
pathology 17:959986, 2005 tachment-related strategies. Motiv Emot 2:77
Lewinsohn PM, Rohde P, Seeley JR, et al: Axis II 102, 2003
psychopathology as a function of Axis I disor- Millon T: Toward a New Personology: An Evolu-
ders in childhood and adolescence. J Am Acad tionary Model. New York, Wiley, 1990
Child Adolesc Psychiatry 36:17521759, 1997 Moffitt TE: A review of research on the taxonomy
Linehan MM: Cognitive Behavioral Treatment of of life-course persistent versus adolescence-
Borderline Personality Disorder. New York, limited antisocial behavior, in The Cambridge
Guilford, 1993 Handbook of Violent Behavior and Aggres-
Livesley WJ, Jang KL: Toward an empirically sion. Edited by Flannery DJ, Vazsonyi AT,
based classification of personality disorder. Waldman ID. New York, Cambridge Univer-
J Person Disord 14:137151, 2000 sity Press, 2007, pp 4974
Lofgren DP, Bemporad J, King J, et al: A follow-up Pine F: On the development of the borderline-
study of so-called borderline children. Am J child-to-be. Am J Orthopsychiatry 56:450
Psychiatry 148:15411547, 1991 457, 1986
Loranger AW: Personality Disorder Examination Pinto A, Grapentine WL, Francis G, et al: Border-
(PDE) Manual. Yonkers, NY, DV Communica- line personality disorder in adolescents: affec-
tions, 1988 tive and cognitive features. J Am Acad Child
Lyddon WJ, Alford DJ: Personality disorders: a Adolesc Psychiatry 35:13381343, 1996
cognitive behavioral perspective, in Personal- Rey JM, Singh M, Morris-Yates A, et al: Referred
ity Disorders in Children and Adolescents. adolescents as young adults: the relationship
Edited by Freeman A, Reinecke M. New York, between psychosocial functioning and person-
John Wiley & Sons, 2007, pp. 99130 ality disorder. Aust N Z J Psychiatry 31:219
MacCallum RC, Zhang S, Preacher KJ, et al: On 226, 1997
the practice of dichotomization of quantitative Reynolds WM: Adolescent Psychopathology
variables. Psychol Methods 7:4163, 2002 Scale. Odessa, FL, Psychological Assessment
Marton P, Golombek H, Stein B, et al: Behavior Resources, 1998
disturbance and changes in personality dys- Roberts BW, DelVecchio WF: The rank-order con-
function from early to middle adolescence. sistency of personality traits from childhood
Adolesc Psychiatry 14:394406, 1987 to old age: a quantitative review of longitudi-
Mattanah JJF, Becker DF, Levy KN, et al: Diagnos- nal studies. Psychol Bull 126:325, 2000
tic stability in adolescents followed 2 years af- Rogosch FA, Cicchetti D: Child maltreatment, at-
ter hospitalization. Am J Psychiatry 152:889 tention networks, and potential precursors to
894, 1995 borderline personality disorder. Dev Psycho-
McCrae RR, Costa PT, de Lima MP, et al: Age dif- pathol 17:10711089, 2005
ferences in personality across the adult life Rothbard JC, Shaver PR: Continuity of attachment
span: parallels in five cultures. Dev Psychol across the life span, in Attachment in Adults:
35:466477, 1999 Clinical and Developmental Perspectives. Ed-
McCrae RR, Costa PT, Terracciano A, et al: Person- ited by Sperling MB, Berman WH. New York,
ality trait development from age 12 to age 18: Guilford, 1994, pp 3171
longitudinal, cross-sectional, and cross-cul- Rothbart MK, Bates JE: Temperament, in Hand-
tural analyses. J Pers Soc Psychol 83:1456 book of Child Psychology, Vol 3: Social, Emo-
1468, 2002 tional, and Personality Development, 5th Edi-
McDermott PA: A nationwide study of develop- tion. Edited by Damon M, Eisenberg N. New
mental and gender prevalence for psychopa- York, Wiley, 1998, pp 105176
thology in childhood and adolescence. J Ab- Rothbart MK, Ahadi SA, Hershey KL, et al: Inves-
norm Child Psychol 24:5366, 1996 tigations of temperament at three to seven
Meeus W, Iedema J, Helsen M, et al: Patterns of years: the Childrens Behavior Questionnaire.
adolescent identity development: review of Child Dev 72:13941408, 2001
Developmental Issues 141

Shaffer D, Fisher P, Lucas CP, et al: NIMH Diag- Stein B, Golombek H, Marton P, et al: Personality
nostic Interview Schedule for Children Ver- functioning and change in clinical presenta-
sion IV (NIMH DISC-IV): description, differ- tion from early to middle adolescence. Ado-
ences from previous versions, and reliability lesc Psychiatry 14:378393, 1987
of some common diagnoses. J Am Acad Child Taylor S, Goritsas E: Dimensions of identity diffu-
Adolesc Psychiatry 39:2838, 2000 sion. J Personal Disord 8:229239, 1994
Shea MT, Stout R, Gunderson J, et al: Short-term Tellegen A: Structure of mood and personality and
diagnostic stability of schizotypal, borderline, their relevance to assessing anxiety, with an
avoidant, and obsessive-compulsive personal- emphasis on self-report, in Anxiety and the
ity disorders. Am J Psychiatry 159:20362041, Anxiety Disorders. Edited by Tuma AH, Maser
2002 JD. Hillsdale, NJ, Erlbaum, 1985, pp 681706
Shedler J, Westen D: Refining the measurement of van Ijzendoorn MH, Bakermans-Kranenburg MJ:
Axis II: a Q-sort procedure for assessing per- Attachment disorders and disorganized at-
sonality pathology. Assessment 5:335355, tachment: similar and different. Attach Hum
1998 Dev 5:313320, 2003
Shiner R: A developmental perspective on person- Waters E, Merrick S, Treboux D, et al: Attachment
ality disorders: lessons from research on nor- security in infancy and early adulthood: a lon-
mal personality development in childhood gitudinal study. Child Dev 71:684689, 2000
and adolescence. J Personal Disord 19:202 Westen D, Shedler J, Glass S, et al: Personality di-
210, 2005 agnoses in adolescence: DSM-IV Axis II diag-
Shiner R, Caspi A: Personality differences in child- noses and an empirically derived alternative.
hood and adolescence: measurement, devel- Am J Psychiatry 160:952966, 2003
opment, and consequences. J Child Psychol Westen D, Dutra L, Shedler J: Assessing adoles-
Psychiatry 44:232, 2003 cent personality pathology. Br J Psychiatry
Siegler RS: Childrens Thinking, 2nd Edition. En- 186:227238, 2005
glewood Cliffs, NJ, Prentice Hall, 1991 Westen D, Nakash O, Thomas C, et al: Clinical as-
Skodol AE: Manifestations, clinical diagnosis, and sessment of attachment patterns and person-
comorbidity, in The American Psychiatric ality disorder in adolescents and adults. J Con-
Publishing Textbook of Personality Disorders. sult Clin Psychol 74:10651085, 2006
Edited by Oldham JM, Skodol AE, Bender D. Widiger TA, Sankis L: Adult psychopathology: is-
Washington, DC, American Psychiatric Pub- sues and controversies. Ann Rev Psychol
lishing, 2005, pp 5787 51:377404, 2000
Skodol AE, Johnson JG, Cohen P, et al: Personality Winograd G, Cohen P, Chen H: Adolescent bor-
disorder and impaired functioning from ado- derline symptoms in the community: progno-
lescence to adulthood. Br J Psychiatry 190:415 sis for functioning over 20 years. J Child Psy-
420, 2007 chol Psychiatry 49:933941, 2008
Solomon J, George C: Attachment Disorganiza- Zanarini MC, Frankenburg FR, Hennen J, et al:
tion. New York, Guilford, 1999 The longitudinal course of borderline psycho-
Sroufe LA: Emotional Development: The Organi- pathology: 6-year prospective follow-up of the
zation of Emotional Life in the Early Years. phenomenology of borderline personality dis-
New York, Cambridge University Press, 1996 order. Am J Psychiatry 160:274283, 2003
This page intentionally left blank
Childhood Experiences and
Development of Maladaptive
and Adaptive Personality Traits
Jeffrey G. Johnson, Ph.D.
Elizabeth Bromley, M.D.
Pamela G. McGeoch, M.A.

During the past century, clinical experience warmth and support, may be likely to pro-
and research have provided considerable mote the development of adaptive traits
support for the hypothesis that interpersonal such as trust, altruism, and optimism, due to
experiences during childhood and adoles- social learning processes and development
cence play an important role in personality of a secure attachment style during child-
development (e.g., Erikson 1963). Childhood hood (e.g., Erikson 1963; Sroufe et al. 1999).
adversities such as maladaptive parenting This chapter presents a summary of research
and childhood abuse and neglect may be findings that are currently available regard-
likely to have an adverse impact on person- ing the role that childhood experiences may
ality development because they interfere play in the development of maladaptive and
with or alter the trajectory of normative so- adaptive personality traits. We begin by sum-
cialization processes during childhood and marizing the evidence that is currently avail-
adolescence (Cohen 1999; Johnson et al. able from retrospective and prospective stud-
2001a). Positive experiences during child- ies regarding the hypothesized association
hood and adolescence, such as parental between childhood adversities and personal-

144 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

ity disorders. The association of childhood findings supporting the validity of retrospec-
abuse and neglect with risk for the develop- tive reports of childhood adversities (e.g., Bi-
ment of personality disorders is examined in fulco et al. 1997; Robins et al. 1985), and al-
particular depth because this association is of though retrospective studies have promoted
considerable interest to clinicians and because the formulation of developmental hypothe-
comparatively little information is available ses, it is nevertheless problematic to make
regarding other adversities that may contrib- strong causal inferences based on retrospec-
ute to the development of maladaptive traits. tive data. In addition, a number of studies
The chapter concludes with an examination of have supported the hypothesis that genetic
and prenatal factors may play an important
the available evidence regarding the associa-
role in the development of behavioral and
tion of positive childhood experiences with the
emotional problems that may become evi-
development of adaptive traits.
dent during childhood (Livesley et al. 1993;
Neugebauer et al. 1999; Thomas and Chess
1984). Furthermore, research has indicated
CHILDHOOD ADVERSITIES that maladaptive childhood traits may have
ASSOCIATED WITH DEVELOPMENT an adverse influence on parenting behavior,
OF PERSONALITY DISORDERS potentially increasing risk for childhood
maltreatment (Kendler 1996). Such findings
A large body of research has provided find- have contributed to skepticism about the hy-
ings that are indirectly consistent with the pothesis that childhood adversities play an
hypothesis that some types of childhood ad- important role in the development of mal-
versities may contribute to the development adaptive personality traits and personality
of maladaptive personality traits and per- disorders.
sonality disorders. Retrospective studies, the However, in recent years, investigations
majority of which have been conducted with utilizing a number of different research para-
clinical samples, have demonstrated that in- digms have provided new and compelling
dividuals with personality disorders tend to evidence in support of the hypothesis that
be more likely than individuals without per- childhood experiences have an important in-
sonality disorders to report a history of child- fluence on personality development. Re-
hood maltreatment and other traumatic search has indicated that maladaptive per-
childhood experiences. These findings have sonality traits are likely to be caused by the
also been of interest because they have pro- interaction of genetic and environmental risk
vided evidence of specificity, indicating that factors (Caspi et al. 2002), including maternal
patients with personality disorders are par- behavior, health, and environmental charac-
ticularly likely to report having experienced teristics affecting prenatal development
specific types of childhood abuse or neglect. (Neugebauer et al. 1999). Epidemiological
However, retrospective studies cannot rule studies and co-twin analyses that have con-
out the alternative hypotheses that the asso- trolled for genetic factors have indicated that
ciation of childhood adversities with mal- childhood abuse is likely to be causally re-
adaptive personality traits is attributable to lated to an increased risk for a broad spec-
recall bias or to preexisting childhood traits trum of psychiatric symptoms (Kendler et al.
that may contribute to the onset of some 2000). Neurobiological studies have provided
types of childhood adversities (Maughan considerable evidence suggesting that child-
and Rutter 1997; Paris 1997). hood maltreatment may cause persistent def-
Both of these alternative hypotheses have icits in brain activity associated with a wide
presented significant challenges to research- range of psychiatric symptoms (Teicher et al.
ers in this field. Although there have been 2003). Prospective longitudinal studies and
Childhood Experiences and Development of Personality Traits 145

investigations that obtained evidence of depressive, passive-aggressive, and schizoid

childhood maltreatment from official records personality disorder traits after parental ed-
have supported the hypothesis that child- ucation and parental psychopathology were
hood abuse and neglect may contribute to in- controlled statistically (Johnson et al. 1999a).
creased risk for the development of personal- Antisocial and depressive personality disor-
ity disorders. (Drake et al. 1988; Guzder et al. der traits remained significantly associated
1996; Johnson et al. 1999a, 2000, 2001a, 2001b; with documented physical abuse after other
Luntz and Widom 1994). The findings of personality disorder traits were controlled
these studies and those that have provided statistically. Evidence of physical abuse,
relevant retrospective data are described in obtained from either official records or retro-
greater detail below. spective self-reports, was associated with
elevated antisocial, borderline, passive-ag-
Childhood Physical Abuse gressive, and schizotypal personality disor-
der traits after controlling for parental educa-
Research conducted with clinical, forensic, tion, parental psychopathology, sexual
and epidemiological samples has indicated abuse, and neglect (Johnson et al. 1999a).
that indices of childhood physical abuse may In summary, prospective epidemiologi-
be associated with antisocial, borderline, and cal studies and retrospective clinical studies
other personality disorder traits. Patients have provided considerable evidence in sup-
with antisocial personality disorder (ASPD) port of the hypothesis that childhood physi-
have been found to be more likely than pa- cal abuse may contribute to the onset of
tients with other psychiatric disorders to re- ASPD, independent of the effects of other
port a history of physical abuse during child- types of childhood maltreatment. Epidemio-
hood (e.g., Bierer et al. 2003; Bradley et al. logical studies that relied on prospective and
2005; Norden et al. 1995; see also Bernstein et retrospective data and retrospective clinical
al. 1998). Confirmatory findings have indi- findings have suggested that childhood phys-
cated that individuals identified as having ical abuse may be associated with elevated
experienced childhood physical abuse are risk for the development of borderline and
likely to have problems with aggressive, schizotypal personality disorders after other
criminal, or antisocial behavior (Pollock et. kinds of childhood maltreatment are ac-
al. 1990; Widom 1989). Patients with border- counted for. In addition, prospective or retro-
line personality disorder (BPD) have also spective studies have provided evidence
been found to be likely to report a history of suggesting that childhood physical abuse
physical abuse during childhood (e.g., may be associated with the development of
Brown and Anderson 1991; Goldman et al. depressive, paranoid, passive-aggressive,
1992). Other studies have yielded findings and schizoid personality disorder traits.
indicating that paranoid, schizoid, and
schizotypal traits were associated with re- Childhood Sexual Abuse
ports of childhood physical abuse in clinical
samples (Bernstein et al. 1998; Bierer et al. Many studies have provided evidence indi-
2003; Yen et al. 2002). cating that patients with personality disor-
Data from the Children in the Commu- ders are significantly more likely than pa-
nity study (CICS; for detailed information tients without personality disorders to report
about the study methodology, please see a history of sexual abuse. Patients with BPD, a community- are more likely than other patients to report a
based longitudinal study, indicated that doc- history of childhood sexual abuse (Brown
umented physical abuse was associated with and Anderson 1991; Laporte and Guttman
elevated antisocial, borderline, dependent, 1996; Westen et al. 1990). Evidence also has
146 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

suggested that patients with BPD may tend of the effects of other types of childhood mal-
to have experienced chronic (as opposed to treatment. Epidemiological studies that re-
episodic) sexual abuse during childhood lied on prospective and retrospective data
(Weaver and Clum 1993). Systematic studies and retrospective clinical findings have sug-
have suggested that other personality disor- gested that childhood sexual abuse may be
ders may also be associated with a history of associated with elevated risk for the develop-
reported sexual abuse. Norden et al. (1995) ment of histrionic and depressive personal-
found that reports of childhood sexual abuse ity disorder traits after other kinds of child-
were associated with borderline, histrionic, hood maltreatment are accounted for. In
narcissistic, and schizotypal personality dis- addition, prospective or retrospective stud-
orders. Bierer et al. (2003) reported that ies have provided evidence suggesting that
childhood sexual abuse was associated with childhood sexual abuse may be associated
paranoid personality disorder in a sample of with the development of antisocial, avoidant,
psychiatric outpatients. Shea et al. (1999) ob- narcissistic, paranoid, passive-aggressive,
tained convergent findings from both inpa- schizoid, and schizotypal personality disor-
tient and outpatient samples indicating that der traits.
reported childhood sexual abuse in both
samples was associate d w it h elev ate d Childhood Emotional Abuse
avoidant, paranoid, and schizotypal person-
ality disorder symptom levels. Ruggiero et Research has suggested that emotional abuse
al. (1999) found that military veterans who (including verbal abuse, humiliation, and
reported severe childhood sexual abuse had other psychological maltreatment) may con-
higher antisocial, avoidant, passive-aggres- tribute, independently, to the development of
sive, schizoid, and schizotypal personality personality disorder traits during childhood
disorder symptom levels than did men who and adolescence. Childhood emotional abuse
reported minimal childhood maltreatment. may increase risk for the development of per-
Reports of childhood sexual abuse have also sonality disorders in part by increasing the
been found to be associated with depressive likelihood that youths will experience mal-
symptoms and interpersonal difficulties adaptive thoughts and feelings, such as ex-
(Browne and Finkelhor 1986). cessive guilt, resentment, social anxiety,
Community-based research findings shame, and mistrust of others during their
have indicated that documented sexual most critical years of psychosocial develop-
abuse was associated with elevated BPD ment. Severe childhood verbal abuse may
traits after parental education and parental leave deeper scars than other types of abuse,
psychopathology were controlled statisti- because children tend to internalize verbally
cally (Johnson et al. 1999a). Evidence of sex- abusive statements and to self-inflict these
ual abuse, obtained from either official abusive thoughts throughout their lives (Ney
records or retrospective self-reports, was as- 1987).
sociated with elevated borderline, depres- Clinical studies have indicated that many
sive, and histrionic personality disorder patients with BPD and other personality dis-
traits after controlling for parental education, orders report a history of childhood verbal
parental psychopathology, physical abuse, abuse (Laporte and Guttman 1996; Oldham
and neglect (Johnson et al. 1999a). et al. 1996; Zanarini et al. 1997). Childhood
In summary, prospective epidemiologi- emotional abuse and intolerant, shaming pa-
cal studies and retrospective clinical studies rental behavior have been found to be associ-
have provided considerable support for the ated with avoidant personality traits among
hypothesis that childhood sexual abuse may psychiatric patients (Grilo and Masheb 2002;
contribute to the onset of BPD, independent Stravynski et al. 1989). Bierer et al. (2003) re-
Childhood Experiences and Development of Personality Traits 147

ported that childhood emotional abuse was Childhood Neglect

associated with paranoid personality disor-
der in a sample of psychiatric outpatients. The association of childhood neglect with the
Bernstein et al. (1998) reported that Cluster B development of personality disorders has
and C personality disorders were signifi- been investigated less extensively than the
cantly associated with retrospective reports corresponding associations regarding child-
of childhood emotional abuse. When sexual hood sexual and physical abuse. Neverthe-
and physical abuse have been statistically less, the available evidence suggests that
controlled, emotional abuse has also been childhood neglect may contribute as or more
found to be independently associated with strongly than physical and sexual abuse to
depressive personality traits, including poor the development of personality disorder
self-esteem and suicidality (Briere and Runtz symptoms and other maladaptive personal-
1990; Mullen et al. 1996). ity traits (Gauthier et al. 1996; Johnson et al.
Longitudinal research has indicated that 1999a). Many patients with personality disor-
verbal abuse during childhood, assessed in a ders report a history of childhood neglect
series of maternal interviews, was associated (Oldham et al. 1996), and reports of a lack of
with increased risk for borderline, narcissis- parental affection during childhood have
tic, obsessive-compulsive, and paranoid per- been found in clinical samples to be associ-
sonality disorders and with elevated border- ated with antisocial, avoidant, borderline, de-
line, narcissistic, paranoid, schizoid, and pendent, paranoid, and schizoid personality
schizotypal personality disorder symptom disorder symptoms (Arbel and Stravynski
levels during adolescence and early adult- 1991; Carter et al. 1999; Norden et al. 1995).
hood (Johnson et al. 2001b). These findings Reports of severe childhood neglect have
were obtained after behavioral and emo- been found to be associated with elevated
tional problems during childhood, physical avoidant, schizoid, and schizotypal personal-
abuse, sexual abuse, neglect, physical pun- ity disorder symptom levels among military
ishment, parental education, parental psy- veterans (Ruggiero et al. 1999). Childhood
chopathology, and co-occurring psychiatric neglect has also been found to be associated
disorders were controlled statistically. Such with a wide range of other maladaptive per-
findings have suggested that childhood ver- sonality traits, including antisocial or avoid-
bal abuse may contribute to the development ant behavior, attachment difficulties, hostil-
of some types of personality disorders, inde- ity, paranoia, and self-destructive behavior
pendent of the effects of other types of child- (Dubo et al. 1997; Gauthier et al. 1996; Robins
hood maltreatment. 1966; Sroufe et al. 1999).
When the available data from prospec- Longitudinal research has suggested that
tive epidemiological studies and retrospec- childhood neglect may increase risk for the
tive clinical studies are considered together, development of personality disorders. Evi-
considerable evidence supports the hypothe- dence of childhood neglect (i.e., lack of paren-
sis that childhood emotional abuse may con- tal affection and supervision during early ad-
tribute to the onset of BPD, independent of olescence) has been found to be associated
the effects of other types of childhood mal- with risk for dependent and passive-aggres-
treatment. In addition, prospective or retro- sive personality disorders during adulthood
spective studies have provided evidence (Drake et al. 1988). CICS findings indicated
suggesting that childhood emotional abuse that documented childhood neglect was as-
may be associated with the development of sociated with elevated antisocial, avoidant,
avoidant, depressive, narcissistic, obsessive- borderline, dependent, narcissistic, paranoid,
compulsive, paranoid, schizoid, and schizo- passive-aggressive, and schizotypal person-
typal personality disorder traits. ality disorder traits after parental education
148 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

and parental psychopathology were con- emotional, physical, and supervision neglect
trolled statistically (Johnson et al. 1999a). An- were independently associated with ele-
tisocial, avoidant, borderline, narcissistic, vated overall personality disorder symptom
and passive-aggressive personality disorder levels and overall risk for personality disor-
traits remained significantly associated with ders during adolescence or early adulthood
documented neglect after other personality after other types of childhood maltreatment
disorder traits were controlled statistically. were accounted for (Johnson et al. 2000).
Evidence of childhood neglect, obtained from These findings suggest that specific types of
either official records or retrospective self- childhood neglect may contribute in unique
reports, was associated with elevated antiso- ways, in combination with other childhood
cial, avoidant, borderline, dependent, narcis- adversities, to the development of different
sistic, passive-aggressive, and schizotypal types of personality disorder symptoms.
personality disorder traits after controlling In summary, prospective epidemiologi-
for parental education, parental psychopa- cal studies and retrospective clinical studies
thology, sexual abuse, and neglect (Johnson et have provided considerable evidence in sup-
al. 1999a). port of the hypothesis that childhood neglect
In recent years, research has suggested may contribute to the onset of avoidant, bor-
that specific types of childhood neglect may derline, passive-aggressive, and schizotypal
be differentially associated with specific personality disorders, independent of the ef-
types of personality disorder traits. Patients fects of other types of childhood maltreat-
with BPD have been found to be more likely ment. In addition, epidemiological studies
than patients with other personality disor- that relied on prospective and retrospective
ders to report a history of childhood emo- data and retrospective clinical findings have
tional, physical, and supervision neglect (Za- suggested that childhood neglect may be as-
narini et al. 1997). Patients with antisocial, sociated with risk for ASPD after other kinds
avoidant, dependent, and paranoid person- of childhood maltreatment are accounted for.
ality disorders have been found to be more Furthermore, prospective or retrospective
likely than other patients to report a history studies have provided evidence suggesting
of childhood emotional neglect (Carter et al. that childhood neglect may be associated
1999). Patients with elevated schizoid per- with the development of dependent, narcis-
sonality disorder symptom levels have been sistic, paranoid, and schizoid personality
found to be particularly likely to report a his- disorder traits.
tory of childhood emotional neglect (Bern- Prospective epidemiological studies
stein et al. 1998). Dubo et al. (1997) found that have suggested that specific types of child-
symptoms of self-mutilation and suicidality hood neglect may be differentially associated
were associated with retrospective reports of with elevated risk for specific types of per-
childhood emotional neglect among patients sonality disorder symptoms.
with BPD. Johnson et al. (2000) reported that
childhood emotional neglect was indepen- Case Example
dently associated with increased risk for Ms. D was a 19-year-old psychiatric out-
avoidant personality disorder and elevated patient diagnosed with depressive, ob-
paranoid personality disorder symptom lev- sessive-compulsive, and borderline per-
els, physical neglect was independently as- sonality disorder symptoms and severe
sociated with elevated schizotypal symptom narcissistic personality disorder symp-
levels, and supervision neglect was indepen- toms. During psychotherapy sessions,
she reported that she had been molested
dently associated with elevated borderline,
and sexually abused by an older half-
paranoid, and passive-aggressive personal- brother from age 4 years until early ad-
ity disorder symptom levels. In addition, olescence, when she became aware of
Childhood Experiences and Development of Personality Traits 149

the meaning of sexual activity. The sex- ample, she reported that her mother fre-
ual abuse began with episodic molesta- quently forced her to wait until the other
tion but became more severe from age 7 family members had finished their meals
onward, occurring during lengthy peri- before allowing her to eat. She also re-
ods of time when she and her half- ported that her mother made her spend
brother were left alone and unsuper- substantially more of her time than her sib-
vised by her parents. Ms. D tried to tell lings doing housework and other chores.
her mother about the sexual abuseshe Ms. E reported that her mother often hu-
miliated her by doing things such as mak-
wrote a suicide note that her mother
ing her wear boys clothing to school, and
foundbut her mother did not put an
that her mother punished her severely for
end to her half-brothers behavior.
any appearance of sexual behavior. In ad-
Ms. D also reported that her father
dition, Ms. E reported that her mother ne-
frequently relied on harsh physical pun- glected her emotionally and that her
ishment to discipline her, for example, mother rarely, if ever, was affectionate,
using a belt to whip her when she re- nurturing, or supportive toward her.
ceived poor grades in school. At times, Ms. E reported that she was forced to
this punishment was so severe that it re- leave the family home at age 17 and soon
sulted in bruises or lacerations, includ- afterward became homeless. She became
ing an open gash in her leg, indicative of involved in an abusive relationship with a
physical abuse. Ms. D also reported that man who beat and raped her repeatedly.
her father abused her emotionally, call- She ran away from him and asked her
ing her names like stupid, lazy, and mother to take her in, but her mother told
a whore. In addition, she reported her, You chose your bed. Now you can lie
that her mother was emotionally and in it. Ms. E went to a homeless shelter,
physically abusive, although this abuse where she was beaten by some other
was not as severe as that perpetrated by young women on the day she arrived.
her father. During her stay at the shelter, Ms. E began
Ms. D frequently witnessed physi- receiving psychiatric treatment for the first
cal violence between her parents, who time. However, her sense of well-being
were often verbally or physically com- was frequently threatened while she lived
bative. Her father often drank heavily there, and she reported that she constantly
and had a number of extramarital af- felt endangered during that time. She re-
fairs. Ms. D informed her mother about ported that on one occasion she was raped
one of her father s indiscretions, and by a stranger while at the shelter. Prior to
her mother became so enraged that she her initial hospitalization, Ms. E reported
shot Ms. Ds father. There was abun- having cut her wrists on several occasions
dant evidence indicating that Ms. Ds when she found the adversities of life in
history of maltreatment and problem- the homeless shelter to be overwhelming.
atic parenting contributed to the devel- She received several years of treatment
opment and persistence of her depres- that enabled her to recover sufficiently to
sive, obsessive-compulsive, borderline, be able to live semi-independently. How-
and narcissistic personality disorder ever, many of her symptoms were so se-
traits, which were associated with con- vere and unremitting that she was eventu-
siderable impairment and distress. ally classified as chronically disabled.

Case Example
Ms. E was a 22-year-old psychiatric inpa- CLINICAL IMPLICATIONS OF
tient with severe BPD, dependent person- RESEARCH ON CHILDHOOD
ality disorder traits, and posttraumatic
stress disorder, with a history of dissocia- MALTREATMENT AND RISK FOR
tive symptoms, severe insomnia, de- PERSONALITY DISORDER
pressed mood, and psychotic episodes.
Ms. E reported that she had been emotion-
It may be possible to prevent the onset of
ally abused and scapegoated by her
mother throughout her childhood. For ex- chronic personality disorders among some
150 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

youths by providing high-risk parents with ality disorder symptoms, and paternal over-
services that assist them in developing more protection was associated with schizoid per-
adaptive parenting behaviors. Research has sonality disorder symptoms (Norden et al.
indicated that it is possible to reduce the like- 1995). Community-based longitudinal re-
lihood that children will develop psychiatric search has also indicated that a wide range of
disorders by helping parents to learn more problematic parenting behaviors may be as-
effective child-rearing techniques (Redmond sociated with risk for personality disorders
et al. 1999). In addition, because treatment of (Drake et al. 1988; Johnson et al. 2001a, 2006).
parental disorders may help to reduce the In addition, research has indicated that a
likelihood of childhood maltreatment and number of other childhood adversities in-
problematic parenting, it may be possible to cluding parental death, parental separation
decrease offspring risk for personality disor- or divorce, socioeconomic adversities, trau-
ders by improving the recognition and treat- matic life events, and victimization (e.g., as-
ment of psychiatric disorders among parents sault, bullying) may be associated with ele-
in the community (Chilcoat et al. 1996). vated risk for personality disorders (Coid
1999; Johnson et al. 1999b, 2006; Zanarini and
Frankenburg 1997).


Although childhood abuse and neglect are Research has identified a wide variety of ex-
likely to play a particularly important role in periences, relationships, and community re-
the development of personality disorder sources that may promote the development
symptoms, a number of studies have in- of adaptive personality traits, such as hardi-
dicated that problematic parenting (e.g., ness or resiliency during childhood and ado-
parenting behavior that, although problem- lescence. Familial warmth, extrafamilial sup-
atic, is not sufficiently severe to be classified port, and other facilitative environmental
as abuse or neglect) is likely to be associ- characteristics have been found to be associ-
ated with the development of maladaptive ated with the development of adaptive traits.
personality traits and personality disorders These traits, in turn, are likely to play an im-
(see Chapter 7, Developmental Issues). Re- portant mediating role in determining
search has indicated that a lack of parental whether individuals are able to adapt effec-
affection during childhood, low family com- tively to adversities during adulthood (Gar-
munication and expressiveness, a lack of pa- mezy 1985; Shiner 2000; Werner and Smith
rental time with the child, and harsh, control- 1982).
ling parenting behavior are associated with
elevated personality disorder traits among
Familial Protective Factors
adolescent nonpatients and adult psychiatric
patients (Baker et al. 1996; Head et al. 1991; A wide range of parenting behaviors (e.g., af-
Johnson et al. 1997, 2006; Parker et al. 1999; fection, communication, time spent with
Stravynski et al. 1989). Retrospective reports children) and characteristics of the family
of a lack of parental affection during child- and home environment play an important
hood by patients with personality disorders role in healthy child development (Johnson
have been found, in a patient sample, to be et al. 2001a; see also Chapter 7, Develop-
associated with ASPD and schizoid person- mental Issues). Parental empathy, support,
Childhood Experiences and Development of Personality Traits 151

and warmth have been found to help chil- community organizations that help young
dren and adolescents cope effectively with people to develop ethics and values; mentors
many types of adversities (Cowen et al. 1997; such as teachers, godparents, and adult role
Jaffee et al. 2007; Luthar and Zigler 1991; Wy- models; and confidants in the form of highly
man et al. 1991). Research has also indicated functioning and supportive peers may facili-
that children who develop a close, strong, tate the development of adaptive personality
and mutually respectful relationship with traits (Werner 1989). The presence of a men-
their parents tend to be particularly resilient tor during adolescence has been found to be
and to have adaptive coping skills (Kobak associated with improved academic achieve-
and Sceery 1988). Furthermore, strong and ment, attitudes about school, insight, rela-
supportive relationships with parents and tionships with parents and peers, and self-
family members tend to be associated with esteem, and with reductions in aggressive
healthy interpersonal functioning during behavior and psychoactive substance use
adulthood and successful adaptation to (Wolkow and Ferguson 2001; Zimmerman et
adult responsibilities (Werner and Smith al. 2002).
1982; see also Chapter 7, Developmental Is- Participation in community activities and
sues). Young adults who perceive their fam- organizations may also have a variety of ben-
ily as warm and supportive tend to be rela- eficial consequences. Community involve-
tively confident and adaptable and to have ment may help to provide a sense of purpose,
high self-esteem, whereas those who per- to increase the availability of social support,
ceive their parents as authoritarian tend to be and to foster resiliency (Vaillant 1977).
more uncertain about themselves and the fu- Youths who have a strong sense of member-
ture (Strage 1998). ship in and identification with the commu-
It is important to note that a variety of nity may adapt more effectively to stressful
parenting styles may lead to positive out- life events (Heath et al. 1999). Extracurricular
comes (Baldwin et al. 1990) and that the activities, such as participation in athletic ac-
child-rearing behavior of the parent is deter- tivities, arts and crafts, hobbies, musical en-
mined, in part, by the disposition or temper- sembles, and organized recreational activi-
ament of the child (Cohen 1999; Kendler ties may also promote healthy adaptation to
1996). For example, youths with externaliz- adversity (Bell and Suggs 1998). Thus, com-
ing behavior problems may need extra pa- munities and schools that provide young
rental supervision, and youths with internal- people with a wide range of opportunities to
izing problems may be in particular need of engage in such activities may help to pro-
parental warmth and support. However, re- mote healthy personality development dur-
search has indicated that most youths benefit ing childhood and adolescence.
from having responsible, nurturing, sup-
portive parents who gradually encourage
them to function in an increasingly autono- ADAPTIVE PERSONALITY TRAITS
mous manner as they mature (Cowen et al.
1997; Luthar and Zigler 1991; Rikhye et al. The research findings cited earlier are consis-
2008; Strage 1998; Wyman et al. 1991). tent with Erik Eriksons (1963) hypothesis
that personality development during child-
Extrafamilial Protective Factors hood and adolescence is determined, in large
measure, by the childs upbringing, chronic
Many different types of community and adversities, and other important interper-
neighborhood resources may help to pro- sonal experiences. Although each child be-
mote healthy personality development dur- gins life with behavioral tendencies that are
ing childhood and adolescence. Supportive influenced by genetic and prenatal factors
152 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

(Livesley et al. 1993; Neugebauer et al. 1999; view themselves as capable of changing the
Thomas and Chess 1984), life experiences ap- events and circumstances in their lives, and
pear to play a critical role in determining how to believe that planning and preparation can
these temperamental characteristics are ex- mitigate or prevent future problems (Kobasa
pressed (Caspi et al. 2002; Cohen 1999; Wil- 1979). In addition, hardy individuals have
son et al. 2006). Research has also supported been found to view changing circumstances
Eriksons hypothesis that children who grow as opportunities for growth (Pengilly and
up in a supportive environment are more Dowd 2000; Werner 1989, 1992).
likely to develop character strengths such as
trust in others, autonomy, industriousness, Self-Efficacy
and self-esteem (see Chapter 7, Develop-
mental Issues). These and other personality Children and adults who believe that they are
traits have been found to promote the devel- in control of their lives tend to remain well in
opment of strong, supportive relationships the face of adversity. Longitudinal research
with others to decrease risk for the develop- has indicated that children with high self-ef-
ment of psychiatric disorders (Bromley et al. ficacy scores who experienced a high level of
2006) and to facilitate adaptation to adversi- family stress were more likely than other chil-
ties later in life (Garmezy 1985; Rutter 1987; dren to have positive outcomes, such as being
Shiner 2000; Werner and Smith 1982). competent and caring (Werner 1989, 1992).
Similarly, self-mastery and an internal locus
Resiliency of control have also been found to be associ-
ated with positive outcomes (Wyman et al.
Research has identified personality traits, 1991). Longitudinal research has indicated
such as optimism and productivity, that tend that successful peer and school adaptation
to be associated with an adaptive, resilient re- are particularly evident among children who
sponse to stress (Pengilly and Dowd 2000; work enthusiastically, creatively, and persis-
Rutter 1987). The development of resiliency tently and who strive to achieve high stan-
may stem in part from experiences that teach dards (Shiner 2000).
individuals how to cope effectively with dif-
ficulties, thereby inoculating them so that Pro-Social Traits
they are able to deal with future adversities
more effectively (Rutter 1987). Adaptive Pro-social traits, including communication
traits referred to as ego resiliency (confi- skills, confidence, empathy, perceptiveness,
dent optimism, insight and warmth, produc- and warmth, appear to play an important
tive activity, and skilled expressiveness) have role in the development of adaptive func-
been found to be associated with positive out- tioning during childhood and adolescence
comes, such as the ability to arouse liking and (Shiner 2000). Resilient youths tend to inter-
acceptance by others (Block and Gjerde 1990; act with and reach out to others, rather than
Klohnen 1996; Klohnen et al. 1996). withdrawing, in both adverse and normal
circumstances. Longitudinal research has
Hardiness shown that adolescents with pro-social ten-
dencies tend to have better long-term psy-
Kobasa (1979) identified a similar set of chosocial outcomes (Shiner 2000). This may
adaptive traits as being indicative of hardi- be attributable in part to the familial and ex-
ness. Individuals with a high level of hardi- trafamilial support that may be made partic-
ness tend to view stressful events as being ularly abundant to individuals with pro-
potentially meaningful and interesting, to social personality traits (Garmezy 1985).
Childhood Experiences and Development of Personality Traits 153

Other Adaptive Traits grew closer to her father and learned

more about his illness experience. She
Conscientiousness, impulse control, integ- took a long-wished-for trip to Italy to
rity, and persistence have been found to be visit her high school art teacher. She
committed herself to finishing her work
associated with the development of resil-
on a large painting exhibit with a friend,
iency during adolescence (Funder and Block saying I just knew my work and my
1989; Klohnen et al. 1996; Luthar and Zigler community would keep me upbeat and
1991; Rutter 1990; Shiner 2000). The ability to give my days hope. Referring to how
respond to humor and to share it with others she coped with a friends death as a
has been found to promote positive out- teen, she said she learned at the time
comes (Klohnen et al. 1996; Luthar and Zig- that I had the strength to find some
meaning in hardship. Ms. F was also
ler 1991; Vaillant 1977). Humor is viewed as
able to explore her anger, sense of loss,
an adaptive defense or coping style by psy-
and sorrow for those she would leave
choanalytic theorists, as are altruism, sup- behind.
pression (i.e., the conscious postponement of
attention to disturbing circumstances), antic- Ms. Fs story illustrates how community,
ipation (i.e., consciously planning how to activity, self-efficacy, parents, and loved ones
cope with stressful circumstances), self-regu- are called on in times of profound stress. She
lation (Gardner et al. 2008; Martel et al. 2007), not only coped with the difficulties of cancer
and sublimation (Vaillant 1977). treatment but also found generative ways to
help others. The love and encouragement
that she received from her parents and men-
Case Example
tors, her insight into her own feelings, her
Ms. F was a 31-year-old woman with perseverance, and her ability to rally others
metastatic adenocarcinoma. At the time around her were important aspects of her
of her diagnosis, she was living with her successful coping.
boyfriend and working as a painter
with some commercial success. She had
graduated from a prestigious univer-
sity and attended graduate art school. CLINICAL IMPLICATIONS OF
The oncologist described Ms. F as a
real fighter, all the way through the RESEARCH ON PROTECTIVE FACTORS
c hem o t he ra p y a nd s ur ge ry. She AND ADAPTIVE PERSONALITY TRAITS
learned what she could about the illness
and treatment. Ms. F brought small,
The present literature review supports the
bright paintings for the nurses and pa-
tients every week. She came to the hos- recommendation to assess protective factors
pital each week with a variety of close and personality strengths as well as symp-
friends and loved ones. Her parents toms and maladaptive traits. Assessing pro-
were helpful with the treatment ar- tective factors and adaptive personality traits
rangements. may increase the effectiveness of a clinical in-
When seen by the psychiatrist, Ms. tervention, in part, by making it clear to the
Fs boyfriend, mother, and father were
patient that the clinician is interested in de-
sitting around her bed. Flowers and
small paintings were on the bedside ta-
veloping a well-rounded understanding of
ble. Ms. F, bald and thin, smiled weakly. the patients strengths and weaknesses,
She said that at the time of her diagno- thereby fostering the development of a strong
sis, my career was really taking off, therapeutic alliance. Moreover, recent evi-
and my boyfriend and I were engaged. dence suggests that patients with some inter-
It was such a shock, but I had no choice personal problems who can cite positive de-
but to learn what I could from the hor-
velopmental experiences may have a better
ror. She described how she coped. She
prognosis than those patients who cannot
154 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

(Skodol et al. 2007). In addition, there are Bifulco A, Brown GW, Lillie A, et al: Memories of
many ways that clinicians and other profes- childhood neglect and abuse: corroboration in
a series of sisters. J Child Psychol Psychiatry
sionals who work with young people can
38:365374, 1997
help to promote the development of adaptive Block J, Gjerde PF: Depressive symptoms in late
personality traits during childhood and ado- adolescence: a longitudinal perspective on
lescence. Some examples include promoting personality antecedents, in Risk and Protec-
extracurricular activities, encouraging youths tive Factors in the Development of Psychopa-
to take on age-appropriate challenges and re- thology. Edited by Rolf J, Masten AS, Cicchetti
D, et al. Cambridge, England, Cambridge Uni-
sponsibilities, facilitating the development of versity Press, 1990, pp 334360
appropriate relationships with suitable adult Bradley R, Jenei J, Westen D: Etiology of border-
mentors, and assisting the youth and family line personality disorder: disentangling the
in taking advantage of community resources contributions of intercorrelated antecedents.
and in participating actively in the life of the J Nerv Ment Dis 193:2431, 2005
Briere J, Runtz M: Differential adult symptomatol-
community. Furthermore, it may be impor-
ogy associated with three types of child abuse
tant to assist parents in improving their child- histories. Child Abuse Neglect 14:357364, 1990
rearing skills, developing a closer and more Bromley E, Johnson JG, Cohen P: Personality
supportive relationship with their children, strengths in adolescence and decreased risk of
and encouraging parents to minimize reli- developing mental health problems in early
adulthood. Compr Psychiatry 47:315324, 2006
ance on disciplinary methods that have been
Brown GR, Anderson B: Psychiatric morbidity in
found to be problematic if utilized too fre- adult inpatients with childhood histories of
quently (e.g., harshly controlling, punishing, sexual and physical abuse. Am J Psychiatry
or shaming offspring; see Redmond et al. 148:5561, 1991
1999). Browne A, Finkelhor D: Impact of child sexual
abuse: a review of the research. Psychol Bull
99:6677, 1986
Carter JD, Joyce PR, Mulder RT, et al: Early defi-
REFERENCES cient parenting in depressed outpatients is as-
sociated with personality dysfunction and not
Arbel N, Stravynski A: A retrospective study of with depression subtypes. J Affect Disord
sepa ra tion in the development of adult 54:2937, 1999
avoidant personality disorder. Acta Psychiatr Caspi A, McClay J, Moffitt TE, et al: Role of geno-
Scand 83:174178, 1991 type in the cycle of violence in maltreated chil-
Baker JD, Capron EW, Azorlosa J: Family environ- dren. Science 297:851854, 2002
ment characteristics of persons with histrionic Chilcoat HD, Breslau N, Anthony JC: Potential
and dependent personality disorders. J Per- barriers to parent monitoring: social disad-
sonal Disord 10:8287, 1996 vantage, marital status, and maternal psychi-
Baldwin AL, Baldwin C, Cole RE: Stress-resistant atric disorder. J Am Acad Child Adolesc Psy-
families and stress-resistant children, in Risk chiatry 35:16731682, 1996
and Protective Factors in the Development of Cohen P: Personality development in childhood:
Psychopathology. Edited by Rolf J, Masten AS, old and new findings, in Personality and Psy-
Cicchetti D, et al. Cambridge, England, Cam- chopathology. Edited by Cloninger CR. Wash-
bridge University Press, 1990, pp 257280 ington, DC, American Psychiatric Press, 1999,
Bell CC, Suggs H: Using sports to strengthen resil- pp 101127
iency in children: training heart. Child Ado- Coid JW: Aetiological risk factors for personality
lesc Psychiatr Clin N Am 7:859865, 1998 disorders. Br J Psychiatry 174:530538, 1999
Bernstein DP, Stein JA, Handelsman L: Predicting Cowen EL, Wyman PA, Work WC, et al: Follow-
personality pathology among adult patients up study of young stress-affected and stress-
with substance use disorders: effects of child- resilient urban children. Dev Psychopathol
hood maltreatment. Addict Behav 23:855868, 9:565577, 1997
1998 Drake RE, Adler DA, Vaillant GE: Antecedents of
Bierer LM, Yehuda R, Schmeidler J, et al: Abuse personality disorders in a community sample
and neglect in childhood: relationship to per- of men. J Personal Disord 2:6068, 1988
sonality disorder diagnoses. CNS Spectrums Dubo ED, Zanarini MC, Lewis RE, et al: Child-
8:737754, 2003 hood antecedents of self-destructiveness in
Childhood Experiences and Development of Personality Traits 155

borderline personality disorder. Can J Psychi- psychiatric disorders. J Abnorm Psychol

atry 42:6369, 1997 108:490499, 1999b
Erikson EH: Childhood and Society, 2nd Edition. Johnson JG, Smailes EM, Cohen P, et al: Associa-
New York, WW Norton, 1963 tions between four types of childhood neglect
Funder DC, Block J: The role of ego-control, ego-re- and personality disorder symptoms during
siliency, and IQ in delay of gratification in ado- adolescence and early adulthood: findings of a
lescence. J Pers Soc Psychol 57:10411050, 1989 community-based longitudinal study. J Per-
Gardner TW, Dishion TJ, Connell AM: Adolescent sonal Disord 14:171187, 2000
self-regulation as resilience: resistance to anti- Johnson JG, Cohen P, Kasen S, et al: Association of
social behavior within the deviant peer con- maladaptive parental behavior with psychiat-
text. J Abnorm Child Psychol 36:273284, 2008 ric disorder among parents and their offspring.
Garmezy N: Stress-resistant children: the search Arch Gen Psychiatry 58: 453460, 2001a
for protective factors, in Recent Research in Johnson JG, Cohen P, Smailes EM, et al: Childhood
Developmental Psychopathology. Edited by verbal abuse and risk for personality disorders
Stevenson JE. Oxford, England, Pergamon during adolescence and early adulthood.
Press, 1985, pp 213233 Compr Psychiatry 42:1623, 2001b
Gauthier L, Stollak G, Mess L, et al: Recall of Johnson JG, Cohen P, Chen H, et al: Parenting be-
childhood neglect and physical abuse as differ- haviors associated with risk for offspring per-
ential predictors of current psychological func- sonality disorder during adulthood. Arch Gen
tioning. Child Abuse Neglect 20:549559, 1996 Psychiatry 63:597587, 2006
Goldman SJ, DAngelo EJ, DeMaso DR, et al: Kendler KS: Parenting: a genetic-epidemiologic
Physical and sexual abuse histories among perspective. Am J Psychiatry 153:1120, 1996
children with borderline personality disorder. Kendler KS, Bulik CM, Silberg J, et al: Childhood
Am J Psychiatry 149:17231726, 1992 sexual abuse and adult psychiatric and sub-
Grilo C, Masheb RM: Childhood maltreatment stance use disorders in women: an epidemio-
and personality disorders in adult patients logical and co-twin control analysis. Arch Gen
with binge eating disorder. Acta Psychiatr Psychiatry 57:953959, 2000
Scand 106:183188, 2002 Klohnen EC: Conceptual analysis and measure-
Guzder J, Paris J, Zelkowitz P, et al: Risk factors for ment of the construct of ego-resiliency. J Pers
borderline pathology in children. J Am Acad Soc Psychol 70:10671079, 1996
Child Adolesc Psychiatry 35:2633, 1996 Klohnen EC, Vandewater EA, Young A: Negotiat-
Head SB, Baker JD, Williamson DA: Family envi- ing the middle years: ego-resiliency and suc-
ronment characteristics and dependent person- cessful midlife adjustment in women. Psychol
ality disorder. J Personal Disord 5:256263, 1991 Aging 11:431442, 1996
Heath AC, Madden PA, Grant JD, et al: Resiliency Kobak RR, Sceery A: Attachment in late adoles-
factors protecting against teenage alcohol use cence: working models, affect regulation, and
and smoking: influences of religion, religious representations of self and others. Child Dev
involvement and values, and ethnicity in the 59:135146, 1988
Missouri Adolescent Female Twin Study. Twin Kobasa SC: Stressful life events, personality and
Res 2:145155, 1999 health: an inquiry into hardiness. J Pers Soc
Jaffee SR, Caspi A, Moffitt TE, et al: Individual, Psychol 37:111, 1979
family, and neighborhood factors distinguish Laporte L, Guttman H: Traumatic childhood ex-
resilient from non-resilient maltreated chil- periences as risk factors for borderline and
dren: a cumulative stressors model. Child other personality disorders. J Personal Disord
Abuse Negl 31:231253, 2007 10:247259, 1996
Johnson JG, Quigley JF, Sherman MF: Adolescent Livesley WJ, Jang KL, Jackson DN, et al: Genetic
personality disorder symptoms mediate the and environmental contributions to dimen-
relationship between perceived parental be- sions of personality disorder. Am J Psychiatry
havior and Axis I symptomatology. J Personal 150:18261831, 1993
Disord 11:381390, 1997 Luntz BK, Widom CS: Antisocial personality dis-
Johnson JG, Cohen P, Brown J, et al: Childhood order in abused and neglected children grown
maltreatment increases risk for personality up. Am J Psychiatry 151:670674, 1994
disorders during early adulthood. Arch Gen Luthar SS, Zigler E: Vulnerability and competence:
Psychiatry 56:600606, 1999a a review of research on resilience in childhood.
Johnson JG, Cohen P, Dohrenwend BP, et al: A lon- Am J Orthopsychiatry 61:622, 1991
gitudinal investigation of social causation and Martel MM, Nigg JT, Wong MM, et al: Childhood
social selection processes involved in the asso- and adolescent resiliency, regulation, and ex-
ciation between socioeconomic status and ecutive functioning in relation to adolescent
156 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

problems and competence in a high-risk sam- Ruggiero J, Bernstein DP, Handelsman L: Trau-
ple. Dev Psychopathol 19:541563, 2007 matic stress in childhood and later personality
Maughan B, Rutter M: Retrospective reporting of disorders: a retrospective study of male pa-
childhood adversity: issues in assessing long- tients with substance dependence. Psychiatr
term recall. J Personal Disord 11:1933, 1997 Ann 29:713721, 1999
Mullen PE, Martin JL, Anderson JC, et al: The Rutter M: Psychosocial resiliency and protective
long-term impact of the physical, emotional mechanisms. Am J Orthopsychiatry 57:316
and sexual abuse of children: a community 329, 1987
study. Child Abuse Neglect 20:721, 1996 Rutter M: Psychosocial resilience and protective
Neugebauer R, Hoek HW, Susser E: Prenatal ex- mechanisms, in Risk and Protective Factors in
posure to wartime famine and development of the Development of Psychopathology. Edited
antisocial personality disorder in early adult- by Rolf J, Masten AS, Cicchetti D, et al. Cam-
hood. JAMA 282:455462, 1999 bridge, England, Cambridge University Press,
Ney PG: Does verbal abuse leave deeper scars: a 1990, pp 181214
study of children and their parents. Can J Psy- Shea MT, Zlotnick C, Weisberg RB: Commonality
chiatry 32:371378, 1987 and specificity of personality disorder profiles
Norden KA, Klein DN, Donaldson SK, et al: Re- in subjects with trauma histories. J Personal
ports of the early home environment in DSM- Disord 13:199210, 1999
III-R personality disorders. J Personal Disord Shiner RL: Linking childhood personality with
9:213223, 1995 adaptation: evidence for continuity and
Oldham JM, Skodol AE, Gallaher PE, et al: Rela- change across time into late adolescence. J Pers
tionship of borderline symptoms to histories Soc Psychol 78:310325, 2000
of abuse and neglect: a pilot study. Psychiatr Q Skodol AE, Bender DS, Pagano ME, et al: Positive
67:287295, 1996 childhood experiences: resilience and recov-
Paris J: Childhood trauma as an etiological factor ery from personality disorder in early adult-
in the personality disorders. J Personal Disord hood. J Clin Psychiatry 68:11021108, 2007
11:3449, 1997 Sroufe LA, Carlson EA, Levy AK, et al: Implica-
Parker G, Roy K, Wilhelm K, et al: An exploration tions of attachment theory for developmental
of links between early parenting experiences psychopathology. Dev Psychopathol 11:113,
and personality disorder type and disordered 1999
personality functioning. J Personal Disord Strage AA: Family context variables and the de-
13:361374, 1999 velopment of self-regulation in college stu-
Pengilly JW, Dowd ET: Hardiness and social sup- dents. Adolescence 33:1731, 1998
port as moderators of stress. J Clin Psychol Stravynski A, Elie R, Franche RL: Perception of
56:813820, 2000 early parenting by patients diagnosed with
Pollock VE, Briere J, Schneider L, et al: Childhood avoidant personality disorder: a test of the
antecedents of antisocial behavior: parental al- overprotection hypothesis. Acta Psychiatr
coholism and physical abusiveness. Am J Psy- Scand 80:415420, 1989
chiatry 147:12901293, 1990 Teicher MH, Andersen SL, Polcari A, et al: The
Redmond C, Spoth R, Shin C, et al: Modeling neurobiological consequences of early stress
long-term parent outcomes of two universal and childhood maltreatment. Neurosci Biobe-
family focused preventive interventions: one- hav Rev 27:3344, 2003
year follow-up results. J Consult Clin Psychol Thomas A, Chess S: Genesis and evolution of be-
67:975984, 1999 havioral disorders: from infancy to early adult
Rikhye K, Tyrka AR, Kelly MM, et al: Interplay be- life. Am J Orthopsychiatry 141:19, 1984
tween childhood maltreatment, parental Vaillant GE: Adaptation to Life. Cambridge, MA,
bonding, and gender effects: impact on quality Harvard University Press, 1977
of life. Child Abuse Negl 32:1934, 2008 Weaver TL, Clum GA: Early family environments
Robins LN: Deviant Children Grow Up: A Socio- and traumatic experiences associated with
logical and Psychiatric Study of Sociopathic borderline personality disorder. J Consult Clin
Personality. Balt imore, MD, Williams & Psychol 61:10681075, 1993
Wilkins, 1966 Werner EE: High-risk children in young adult-
Robins LN, Schoenberg SP, Holmes SJ, et al: Early hood: a longitudinal study from birth to 32
home environment and retrospective recall: a years. Am J Orthopsychiatry 59:7281, 1989
test for concordance between siblings with Werner EE: The children of Kauai: resiliency and
and without psychiatric disorders. Am J Or- recovery in adolescence and adulthood. J Ad-
thopsychiatry 55:2741, 1985 olesc Health 13:262268, 1992
Childhood Experiences and Development of Personality Traits 157

Werner EE, Smith RS: Vulnerable But Invincible: A major life stress. Am J Community Psychol
Study of Resilient Children. New York, Mc- 19:405426, 1991
Graw-Hill, 1982 Yen S, Shea MT, Battle CL, et al: Traumatic exposure
Westen D, Ludolph P, Misle B, et al: Physical and and posttraumatic stress disorder in borderline,
sexual abuse in adolescent girls with border- schizotypal, avoidant, and obsessive-compul-
line personality disorder. Am J Orthopsychia- sive personality disorders: findings from the
try 60:5566, 1990 Collaborative Longitudinal Personality Disor-
Widom CS: The cycle of violence. Science 244:160 ders Study. J Nerv Ment Dis 190:510518, 2002
166, 1989 Zanarini MC, Frankenburg FR: Pathways to the
Wilson RS, Krueger KR, Arnold SE, et al: Child- development of borderline personality disor-
hood adversity and psychosocial adjustment der. J Personal Disord 11:93104, 1997
in old age. Am J Geriatr Psychiatry 14:307315, Zanarini MC, Williams AA, Lewis RE, et al: Re-
2006 ported pathological childhood experiences as-
Wolkow KE, Ferguson HB: Community factors in sociated with the development of borderline
the development of resiliency: considerations p er so n a l it y d is o rd e r. A m J Ps y c h i a t ry
and future directions. Community Ment 154:11011106, 1997
Health J 37:489498, 2001 Zimmerman MA, Bingenheimer JB, Notaro PC:
Wyman PA, Cowen EL, Work WC, et al: Develop- Natural mentors and adolescent resiliency: a
mental and family milieu correlates of resil- study with urban youth. Am J Community
iency in urban children who have experienced Psychol 30:221243, 2002
This page intentionally left blank
Part IV
This page intentionally left blank
Levels of Care in Treatment

John G. Gunderson, M.D.

Kim L. Gratz, Ph.D.
Edmund C. Neuhaus, Ph.D.
George W. Smith, M.S.W.

In this chapter, we describe the indications, The presence of a comorbid personality dis-
goals, structures, and empirical evidence re- order often complicates the treatment of an
lated to the use of four different and decreas- Axis I disorder (Tyrer et al. 1997); for exam-
ingly intensive levels of care: IVhospital; ple, the patient with avoidant personality
IIIpartial hospital/day treatment; IIin- disorder may not attend group sessions; the
tensive outpatient; and Ioutpatient. Al- patient with borderline personality disorder
though we are concerned primarily with the (BPD) may refuse family contacts; the pa-
roles these levels of care play in treating per- tient with histrionic personality disorder
sonality disorders, we recognize that pa- may express sensitivity to the side effects of
tients who have a personality disorder often medications, and so on. In such ways, per-
will be placed in levels of care due to a treat- sonality disorders may diminish the progno-
ment primarily directed at comorbid Axis I sis for the treatment of Axis I disorders.
conditions. For example, major depression
has a comorbidity of about 50% with Cluster
B and C disorders (Dolan-Sewell et al. 2001), DEFINITIONS
anxiety disorders have a comorbidity of
about 25% with Cluster C disorders (Dyck et Treatments can be organized according to the
al. 2001), and substance abuse is associated four different levels of care (see Table 91).
with a comorbid Cluster B disorder more Those personality disorders associated with
than 50% of the time (Oldham et al. 1995). the most severe crises and highest levels of

162 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

Table 91. Levels of care

IV Hospital; 24 hours/day with option of locked doors and seclusion

III Partial hospital/day treatment; 28 hours/day, 35 days/week, 630 hours/week;
usually heavily involving group therapies
II Intensive outpatient; 310 hours/week of specifically prescribed, scheduled, and
integrated therapies
I Outpatient; 13 hours/week of scheduled therapies

dysfunction are more apt to require higher search involving specified levels of care is
levels of care. The four levels of care are hier- shown in Table 92. Many of these studies
archical in terms of containment, intensity, have examined personality disorders in gen-
structure, and costs per day, and they are in- eral, and these studies vary in the extent to
versely related to usual length of stay. which they control for or examine the specific
impact of particular personality disorders or a
personality disorder cluster on treatment effi-
GENERAL PRINCIPLES cacy. Most research examining treatment for
GOVERNING LEVELS OF CARE specific personality disorders involves BPD
and antisocial personality disorder (ASPD),
The least restrictive level of care possible is with a growing number of studies examining
usually bestthis maximizes and requires the effectiveness of treatments at different
use of ones personal strengths, increases the levels of care on Cluster C disordersin par-
likelihood of being able to apply new capa- ticular, avoidant personality disorder. Cluster
bilities and skills to community settings (i.e., A disorders have received the least systematic
generalization), and decreases the likelihood attention from researchers. Furthermore, em-
of regressive aspects of treatment (e.g., rein- pirical support for the comparative efficacy of
forcement of dysfunctional behavior). different levels of care is rare.
Availability of all levels is clinically desir-
able (most treatment settings include only IV
and I) and possibly cost beneficial (Quayt-
man and Scharfstein 1997). The availability
of levels II and III decreases use of hospital- USE OF DIFFERENT
izations, decreases dropouts, and increases LEVELS OF CARE
social rehabilitation interventions.
With each decrease in level of care, the The intended goals for therapeutic change are
treatments become more specific for differ- often classified within four domains: subjec-
ent types of personality disorders. tive distress, maladaptive behaviors, interper-
sonal, and intrapsychic (i.e., psychological).
These domains are identified in the sequence
in which change can be expected (Gabbard et
EVIDENTIARY BASE al. 2002; Gunderson and Gabbard 1999;
Several meta-analytic reports have affirmed Howard et al. 1986; Kopta et al. 1994; Lanktree
the value of psychotherapy, which really in- and Briere 1995) as well as a sequence that
cludes all psychosocial therapies, for person- generally should be prioritized in treatment
ality disorders (Leichsenring and Leibing planning. Table 93 indicates the relative ca-
2003; Perry and Bond 2000). The extant re- pacity for the different levels of care to effect
Table 92. Studies on level of care for personality disorders

Hospital Partial hospital Intensive outpatient Outpatient

All personality Dolan et al. 1992, 1997 Karterud et al. 1992, 2003 NA NA
disorders Mehlum et al. 1991
Vaglum et al. 1990
Wilberg et al. 1998b, 1999
Cluster A NA NA NA NA
Cluster B NA NA NA NA
BPD Barley et al. 1993 Bateman and Fonagy 1999, Linehan et al. 1991, 1993, 1994, Blum et al. 2002, 2008

Levels of Care in Treatment

Bohus et al. 2000, 2004 2001, 2008 2006 Brown et al. 2004
Silk et al. 1994 McQuillan et al. 2005 Davidson et al. 2006
Giesen-Bloo et al. 2006
Gratz and Gunderson 2006
Munroe-Blum and Marziali 1995
Stevenson and Meares 1992
Wilberg et al. 1998a
ASPD Cacciola et al. 1995 Cacciola et al. 1995 NA Brooner et al. 1998
Gabbard and Coyne 1987 Compton et al. 1998
Harris et al. 1994) Messina et al. 2003
Hildebrand et al. 2004
Messina et al. 1999, 2002
Ogloff et al. 1990
Reiss et al. 1999
Rice et al. 1992
Richards et al. 2003
Clusters B and C NA Krawitz 1997 NA Hoglend 1993
Winston et al. 1994
Cluster C Gude and Vaglum 2001 NA NA Hardy et al. 1995
AVPD NA NA NA Alden 1989

Note. ASPD= antisocial personality disorder; AVPD=avoidant personality disorder; BPD =borderline personality disorder; NA =not available.
Source. Search engine used: PsycINFO.
164 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

Table 93. Goals and their relationship to levels of care

Distress Behavioral Interpersonal Intrapsychic

Hospital ++ +
Partial hospital ++ ++ +

Intensive outpatient + ++ ++ +
Outpatient + + ++ ++

Note. Effectiveness: ++ =strong; += possible; =unlikely.

change in these four domains. Of course, this itive college. However, her completion of
profile is tied to the expected lengths of stay. college was repeatedly delayed due to an-
gry conflicts with peers and teachers usu-
Another way to classify goals is by the
ally followed by self-endangering behav-
priority they should be assigned in planning iors of variable seriousness. For the past
treatment interventions (Gunderson 2001). several years, she has been living at home
Here, for example, goals include crisis man- doing minimal work.
agement, behavioral stabilization, social re- On arrival, Ms. G was angry at her
habilitation, and psychological growth (Table parents for bringing her but desperate
about needing help. She refused to enter
94). These goals naturally map onto those in the residential program, insisting that she
Table 93 and vary in the extent to which they wasnt that sick. When advised it would
can and should be addressed within each provide a way to develop peer relation-
level of care. For example, crises are often ships, she angrily denied the need for that
managed by hospitalizations because of the and claimed sh e could mak e friends
whenever she wanted. Her parents effort
hospitals role in providing containment, asy-
to correct that claim only made her resis-
lum from stress, and the potential for rapid tance to entering the partial hospital more
medication changes that offer immediate resolute. She nonetheless reiterated that
symptom relief. Behavioral stabilization and she was desperate to receive an intensive
social rehabilitation are achieved through outpatient program (IOP, level II) that had
been unavailable at home.
corrective social learning experiences, as well
A clinical decision needed to be made:
as the continued opportunities for the acqui- either accept Ms. G in an IOP (level II) as
sition and generalization of coping skills that she insisted or not (under which circum-
are central to partial hospital (level III) and in- stances she claimed returning home was
tensive outpatient (level II) programs. Psy- not an option, and she would go to the
streets of Boston).
chological growth change requires longer-
At her parents urging, Ms. G was ac-
term and often repetitious learning experi- cepted into an IOP and quickly found res-
ences available only in stable longer-term set- idence with someone she met there. She
tingsthat is, in level I outpatient care. got attached to her individual therapist
but became preoccupied with her room-
mate, who had her own problems. After 6
Case Example months, the therapist and group therapy
leaders in the IOP concluded that Ms. G
Ms. G, a 26-year-old, single white woman, was making no progressthat she had in-
was referred from another state for treat- sufficient social supports or structure to
ment of personality disorder not other- use treatment for other than crisis man-
wise specified with borderline, schizo- agement. The therapist felt that to insist
typal, and avoidant features. With an on residential care would be experienced
excellent high school grade point average, as rejection and might precipitate suicidal
she had been accepted into a very compet- danger.
Levels of Care in Treatment 165

Ms. Gs case illustrates several common months) suggests that they can be useful in
and difficult problems involved in selecting ways that are not feasible in 214 days. That
or changing a level of care. One is making a research is described at the end of this section.
concession on level of care to accommodate a Potential risks associated with inpatient
patients insistence. Making a concession is hospitalization include reinforcement of
sometimes necessary but is best done with maladaptive behaviors (e.g., parasuicidal, at-
the proviso that the patient agrees to accept tention seeking, control struggles) and/or
your recommendations if he or she has not passive problem solving. Clinicians should
achieved some reasonable progress in some consider these issues for the particular pa-
agreed-upon time (e.g., has not achieved a tient. Nonetheless, the role of hospitalization
job, has not attended therapies regularly, has should be appreciated, because personality
not established a social support system, or disorder patients treated in community-ori-
has not diminished high-risk activities). A ented treatments (level I or II) with an ag-
second problem illustrated by this case is gressive emphasis on keeping patients out of
that once treatments are under way, it can be the hospital have worse outcomes than pa-
very difficult to change them without break- tients for whom hospitals were used as
ing the relational alliance with the patient needed (Tyrer and Simmonds 2003).
and/or precipitating a potentially dangerous
flight. In Ms. Gs case, the IOP team called for Indications
a consultation. Use of outside consultants to
oversee changes helps depersonalize what Hospitalization may be clinically indicated
often looms as a very difficult confrontation. during acute crises and in response to increas-
To do this required time-consuming commu- ingly severe behavioral dysfunction, espe-
nications by the treaters. It is more easily cially with regard to suicidal behaviors and
done within hospitals. How to make such violence toward others. Perhaps also worth
confrontations (advising patients of informa- noting is that hospitalization may be used as
tion that they do not want to hear) and how an asylum to permit patients with personality
to impose limits (prohibiting behavior[s] that disorders to leave abusive or otherwise harm-
a patient wishes to continue) without pa- ful situations or relationships. Hospitaliza-
tients becoming self-destructive or leaving tions can also enable patients to leave treat-
treatment is not easy (see Gunderson 2001 ments that were not helpful but that they
for discussion). would otherwise have had difficulty leaving.
Many people with personality disorders
enter hospitals because of comorbid psychiat-
LEVELS OF CARE ric disorders to which the personality disor-
der may predispose them. Indeed, more than
What follows is an elaboration of indications, half of psychiatrically hospitalized patients
goals, and structures for each of the four lev- have a personality disorder, with the most
els of care (see Table 94). Where available, common types being borderline, avoidant,
relevant empirical evidence is noted. and dependent personality disorders (Lor-
anger 1990). Attention to and consideration
Level IV: Hospital of the personality disorder may affect treat-
ment efficacy for the Axis I disorders and
Given that hospitalizations are almost always should influence treatment decisions.
214 days in duration in practice, the follow- Given that BPD is the predominant per-
ing discussion is geared to those lengths of sonality disorder found in inpatient hospital
stay. Still, research that has been done on hos- services (constituting about 15% of hospital-
pitalizations of longer durations (e.g., 23 izations; Koenigsberg et al. 1985; Loranger
Table 94. Levels of care: modalities, goals, duration, and therapeutic processes

Level Modalities Goals and procedures Duration Therapeutic processes

IV. Hospital Medication Crisis management 214 days Containment

Milieu Decrease distress 24 hours/day Support
Group Decrease suicide risk
Case management
Neurological evaluations

E s s e n t i a l s o f P e r so n a l i t y D i s o r d e r s
Psychological evaluations

Plan/change treatment
Develop treatment plan
Identify primary therapist
Initiate medication changes
Expert consultation

III. Partial hospital Milieu Skills training 12 weeks Structure

Group Stabilize daily living skills 620 hours/week Support
Case management Structure daily activities Involvement
Family Identify maladaptive patterns
Behavioral stabilization
Decrease impulsive behavior 312 weeks
Increase coping skills 610 hours/week

Social rehabilitation 16+ weeks

Improve social functioning
Vocational rehabilitation
Community reentry
Table 94. Levels of care: modalities, goals, duration, and therapeutic processes

Level Modalities Goals and procedures Duration Therapeutic processes

II. Intensive outpatient Group Social (behavioral) adaptation 318 months Support
Family Vocational 310 hours/week Involvement
Individual Behavioral

I. Outpatient Individual Psychological growth 1236 months Involvement

Group Interpersonal 13 hours/week Validation

Levels of Care in Treatment


168 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

1990; Widiger and Weissman 1991), it is worth Although level IV care is rarely indicated
noting a common clinical situation for which for Cluster C personality disorders, it is
hospitalizations should be used with caution. worth noting that hospitalization may offer
Hospitalization should rarely be used in re- useful exposures for patients with avoidant
sponse to self-injurious behavior without sui- personality disorder. For instance, hospital-
cidal intent. Furthermore, although inpatient ization may expose avoidant patients to typ-
stays may be warranted for BPD patients who ically avoided social situations (e.g., seeing
are acutely suicidal, such hospitalizations others in distress) or internal experiences
may not decrease the likelihood of future sui- (e.g., feeling helpless or anxious). Although
cide attempts (Van der Sande et al. 1997). hospitalization may result in initial improve-
Despite the widespread belief that hospi- ments in symptoms for these patients, how-
talization is generally contraindicated for pa- ever, research suggests that patients with
tients with ASPD, research on the effective- pure Cluster C personality disorders (espe-
ness of inpatient therapeutic community cially avoidant personality disorder) may be
drug treatment programs and inpatient sub- at greater risk for relapsing after discharge
stance abuse programs suggests that ASPD than patients with Cluster B personality dis-
patients may respond positively to certain orders (see Gude and Vaglum 2001).
types of inpatient programs (Cacciola et al.
1995; Messina et al. 1999, 2002). Of course, Goals: Crisis Management,
the generalizability of these results to other, Assessments, and Planning and
less specialized inpatient hospital programs
Implementing Treatment Changes
is indeterminable. With regard to contra-
indications to hospitalization, Gabbard and Following are the major goals of hospitaliza-
Coyne (1987) noted that negative responses tions and the usual time required for meeting
to hospitalization are likely for ASPD pa- them:
tients with a history of felony arrests or con-
victions; a history of repeated lying, aliases, Crisis management (26 days): Hospital-
and conning; an unresolved legal situation at ization can diminish acute suicidal or vio-
admission; forced hospitalization as an alter- lent dangers.
native to incarceration; and a history of vio- Extensive neurological or psychological
lence toward others. Moreover, research sug- evaluations (26 days): These evaluations
gests that psychiatric hospitalization is are more easily coordinated, and may
relatively contraindicated for those ASPD only be feasible, in hospital settings.
patients (approximately 25%65%; see Widi- Development of a treatment plan and per-
ger and Corbitt 1996) who are psychopathic sonnel (314 days): Such plans usually re-
(i.e., display a lack of remorse, lack of empa- quire arranging for continuity through
thy, and shallow affect; Harris et al. 1994; appropriate step-downs and assessing the
Hart and Hare 1997; Hildebrand et al. 2004; suitability of new therapy personnel. An
Reiss et al. 1999; Richards et al. 2003; see also essential part of these processes is to iden-
Salekin 2002). In contrast, there is evidence to tify the primary clinician who will be re-
suggest that ASPD patients without psych- sponsible for the patients treatment. For
opathy may respond positively to forensic primary clinicians, an essential first step is
hospitals with therapeutic community pro- to define roles and goalsthat is, estab-
grams (Rice et al. 1992) and that the presence lish a contractual allianceand to con-
of comorbid anxiety and/or depression may tract with the patient about participation
also be associated with a positive response to in aftercare services.
hospitalization among ASPD patients (see Changes in prior therapies (314 days):
Gabbard and Coyne 1987). These changes are often indicated, but
Levels of Care in Treatment 169

they may require expert consultation and The psychiatrist should oversee medica-
the introduction of new therapists. If the tions and involve the patient in any changes.
changes are considered undesirable by As important, the psychiatrist should caution
the patient, working through resistance patients about the relatively modest benefits
may be possible only in the hospital, they can expect. In addition, the psychiatrist
where the options for flight from the pro- should evaluate coexisting Axis I disorders
posed changes are reduced. and give them appropriate priority in after-
Hospitalizations may allow therapists to care planning. For example, comorbid sub-
review prior impasses or establish a stance or alcohol abuse almost always should
clearer framework for their ongoing work. be assigned high priority in aftercare, and co-
For many patients with personality disor- existing depression may not respond well to
der, hospitalization serves as an environ- medications (Gunderson et al. 2004; Koenigs-
ment to initiate medication changes and berg et al. 1999; Kool et al. 2003; Shea et al.
evaluate medication benefits. 1987; Soloff 1998) and may require further
monitoring at a less-restrictive level of care.
Longer-Term Hospitalizations:
A businesslike, practical, supportive, and task- Empirical Lessons
oriented atmosphere and orientation is useful.
Harmful is a milieu that encourages long one- Longer-term hospitalizations may address
to-one talks or the development of personal re- additional goals. Dolan et al. (1997) found
lationships with staff or other patients. Com- that long-term (i.e., average of 7 months)
munity meetings or group therapies that em- therapeutic community inpatient treatment
phasize cohesion or bonding among patients was associated with decreased borderline
are relatively contraindicated. Care should be psychopathology. Furthermore, hospitaliza-
taken not to reinforce maladaptive behaviors tions of 23 months may result in behavioral
or increasingly intense/severe expressions of stabilization (a goal usually reached in level
distress and suicidality. Furthermore, empha- III and generally not addressed in the 210
sizing the short-term nature of the treatment, day hospitalizations described above). Dia-
retaining a focus on impending discharge, and lectical behavior therapy (DBT)based inpa-
making aftercare plans for less-restrictive lev- tient programs in particular may be more ef-
els of care may be useful strategies for prevent- fective in fulfilling this goal than other
ing lengthy or contraindicated hospitaliza- inpatient programs with equally long stays
tions (see Bohus et al. 2000; Silk et al. 1994). (see Barley et al. 1993; Bohus et al. 2004). The
Given that it is easy for staff to feel trapped by advantages shown for such programs are
escalating suicidal ideation as patients ap- presumably due to the emphasis on skills
proach discharge, a consultant can help allevi- training (in particular, distress tolerance
ate unrealistic liability fears. skills), behavioral analyses of problem be-
In regard to staffing, a case manager haviors (with the goal of identifying precipi-
should be assigned and should keep the pa- tants and consequences of problem behaviors
tient oriented toward the problems preceding so as to determine appropriate interven-
the hospitalization that the patient will need tions), and the use of contingency strategies,
to cope with on discharge. A primary task of potentially minimizing unintended rein-
the case manager is to bring in relatives or forcement of maladaptive behaviors (Barley
other significant people to help understand et al. 1993; Bohus et al. 2000, 2004). Moreover,
precipitating events and diminish the likeli- the milieu can be used to practice, and begin
hood of their recurrence after discharge, to re- to generalize, the skills being learned.
ceive psychoeducation about the personality These studies of 2- to 3-month hospitaliza-
disorder, and to coordinate aftercare plans. tions can be used to inform the development
170 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

of hospital programs with shorter lengths of hospital programs may be indicated for pa-
stay. For instance, Silk et al. (1994) developed tients who lack either the social supports or
a DBT-based inpatient treatment for BPD vocational options to make community living
with an average length of stay of 1017 days. viable. Whether as an alternative to or step-
In assessing patients perceptions of improve- down from hospitalization, partial hospital
ment at the time of discharge, they found that programs are needed for personality disorder
patients in the DBT-based program (com- patients with marked social or behavioral im-
pared with patients assigned to a non-DBT pairment. Karterud et al. (2003) operational-
discussion group) felt that the lessons they ized the need for partial hospitalization by
learned would help them to better handle dif- suggesting that it was indicated for patients
ficult or painful situations. Although the with Global Assessment of Functioning
same level of change found with longer-term scores below 50. Partial hospital programs
hospital stays would not be expected with can assess the types of, and reasons for, social
short-term hospitalizations, a similar struc- disability and introduce rehabilitative efforts.
ture can be utilized with beneficial results. This level of care can also be used to intro-
duce or stabilize new treatments when close
Level III: Partial Hospital/Day supervision and evaluations are needed.
Treatment As with level IV, BPD is the personality
disorder most likely to use level III. Al-
The literature presents a diverse picture of though patients with ASPD are also socially
partial hospital treatments for personality impaired, there is some evidence to suggest
disorders, varying extensively with regard to that they may be likely to drop out of partial
treatment duration (i.e., length of stay) and hospital programs (Karterud et al. 2003; Wil-
treatment intensity (i.e., hours per week). berg et al. 1998b). However, given evidence
These programs offer more structure, con- of a positive response to level III therapeutic
tainment, and intensive treatment than out- com m un it y d ru g t re at m en t p rog ram s
patient care while providing the opportunity among ASPD patients (Cacciola et al. 1995),
for behavioral stabilizationand for skills it is possible that certain specialized pro-
building and generalizationthat are not grams may be effective in the treatment of
usually available in typical short-term inpa- ASPD patients. Patients with schizotypal
tient settings. Our discussion of partial hospi- personality disorder are unlikely to be
tal treatment is oriented toward the levels of helped by partial hospital programs (see
duration and intensitythat is, the shorter- Karterud et al. 1992; Vaglum et al. 1990). Sur-
term partial hospital programs (24 weeks in prisingly, however, patients with paranoid
duration) that are usually available in the cur- personality disorder treated in an 18-week
rent health care system. A discussion of the day treatment program evidenced signifi-
lessons that can be learned from the longer- cant improvements in global functioning,
term partial programs that have been the pri- symptom severity, and interpersonal func-
mary recipients of empirical attention can be tioningimprovements comparable with
found at the end of this section. those seen among patients with other per-
sonality disorders (Wilberg et al. 1998b).
For some patients with a personality disor- Goals: Skills Training, Stabilization, and
der, partial hospital programs offer an opti- Social Rehabilitation
mal level of care. They can diminish the like-
lihood of substance/alcohol abuse relapses A pragmatic and theoretically grounded ap-
or suicide attempts (although they do not proach conceptualizes treatment in stages,
have enough monitoring or containment to each of which has its own goals and interven-
prevent these behaviors). Moreover, partial tions. Stages are organized with respect to
Levels of Care in Treatment 171

their duration and intensity as well as to the Stage 2 (38 Weeks)

changes that can reasonably be expected to
occur within a given time frame. Based on Stabilize behavior: This goal involves at-
clinical experience, a high-intensity front- taining better impulse control, resulting in
loaded treatment is optimal initially, fol- a decrease in the frequency and severity of
lowed by treatment at a titrated intensity impulsive and self-destructive behaviors
over several months. Given that most pro- (e.g., self-harm, suicide attempts, sub-
grams can offer only short-term treatment stance use). This stabilization is often a
(what we refer to here as stage 1 of treat- nonspecific effect of asylum, structure,
ment), this titration is particularly relevant. and support.
One partial hospitalization program that em- Initiate vocational rehabilitation: This
phasizes stage 1 targets found that patients goal is not easily accomplished and typi-
can improve in the brief time frame of 12 cally gets overlooked due to the fact that
weeks (Neuhaus 2006; Neuhaus et al. 2007). patients with a personality disorder rarely
The following are goals for each stage: seek vocational rehabilitation. Young or
inexperienced staff are unlikely to give
Stage 1 (12 Weeks) this goal adequate value and importance.
Program administrators or staff involved
Develop a therapeutic alliance and the pa- with families are more apt to determine
tient attaches to treatment: This process is whether it is addressed.
assisted by defining goals and establish- Reevaluate the treatment contract on the
ing an initial treatment contract that gets basis of the extent to which patients are
reviewed and refined over time. working collaboratively and responsibly
Psychoeducation: This helps patients in treatment: In stage 2 treatment, after pa-
frame goals and understand treatment tients have achieved more competency
options, and enlists family support. with regard to coping skills (e.g., better
Stabilize or teach daily living skills (eat- impulse control, improved anxiety man-
ing, sleeping, hygiene): The need for this agement), a greater emphasis is placed on
goal varies, as does the optimal approach interpersonal relationships.
to achieving it. Most patients need consis-
tent monitoring and education about the
importance of eating and sleeping in reg-
ular patterns. Sleep medications may Partial hospital and day treatment programs
prove useful for patients who have trou- rely heavily on group therapy and the posi-
ble falling asleep because of fearfulness. tive effects of a therapeutic milieu. Such
Schedule and structure time and activi- structure must be actively constructed, pro-
ties: This assists with the management actively maintained, and updated as needed.
and planning of daily activities, promotes With a 1- or 2-week length of stay, patients
self-care routines (e.g., sleep, hygiene, do not have the luxury of time to settle in
good eating habits), and aids in overall to treatment. Clear and concise information
stress management. (e.g., description of the program and treat-
Begin to identify typical patterns (e.g., in- ment philosophy, expectations of patients,
terpersonal conflicts, loneliness) that lead roles of treatment team members) provided
to maladaptive behaviors: Although mal- upon entry can diminish a patients anxiety
adaptive patterns cannot be expected to and facilitate immediate involvement in
change in stage 1, it is often useful to in- treatment. Handouts with daily schedules
troduce basic behavioral strategies at this and brief descriptions of groups may further
point, including chain analyses and im- orient patients and promote memory of what
pulse control skills. to expect. The structure of the partial hospital
172 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

program may become a template for struc- havioral treatment, an intensive DBT ap-
turing their lives outside the program. proach has also been shown to be effective in
As with hospitals, a case manager respon- treating borderline patients in a 3- to 4-week
sible for implementing the treatment and time frame (McQuillan et al. 2005). Notably,
monitoring progress should be assigned. This even the psychoanalytic partial hospital pro-
person should work directly with the patient, gram developed by Bateman and Fonagy
treatment team members both within and (2004) is heavily cognitive-behavioral; espe-
outside of the program, and family members. cially in early phases. As such, a core constel-
To be an effective case manager with person- lation of groups within stage 1 treatment may
ality disorder patients requires clinical savvy include treatment contracting, community
to set limits, confront when necessary (e.g., meeting (for addressing therapy-interfering
when patients are missing groups), and stay behaviors), personality disorder psychoedu-
connected with a patient despite being vili- cation (with information about the influence
fied as not understanding or caring. An effec- of Axis I disorders), basic behavioral skills
tive case manager must also be willing and groups (including behavioral scheduling, im-
able to negotiate with treatment team mem- pulse control, distress tolerance, and anxiety
bers outside of the program (e.g., the patients management), self-assessment groups that
outpatient therapist). In short-term treatment teach patients how to identify maladaptive
the case manager may be invaluable to the behavioral patterns (including the emotional
family system in providing information, sup- and cognitive precipitants of maladaptive be-
port, and the framework of an overall treat- haviors), and rudimentary interpersonal
ment plan (e.g., anticipating the issues in- groups (e.g., assertiveness training, interper-
volved in less intensive levels of care). sonal effectiveness, and the impact of person-
Given the potential for treatment noncom- ality styles on relationships).
pliance at this level of care (patients may have Groups during stage 2 of treatment may
difficulty attending treatment when feeling follow the framework of stage 1, although
bad), the structure of the program should in- with greater depth and further opportunities
clude provisions for addressing noncompli- for patients to practice skills and achieve some
ance through program expectations, policies, competency in their use. At this stage of treat-
and an explicit emphasis on therapy-interfer- ment, as patients begin to feel like part of the
ing behaviors. Noncompliance is best ad- milieu, there is a natural progression to focus
dressed by the case manager and then fol- more explicitly on the patients interpersonal
lowed up in groups. It is useful to utilize a relationships. This stage may see an intensifi-
combination of validation, confronting, limit cation of treatment relationships and attach-
setting (e.g., patients cannot stay in the pro- ments, which offers challenges to patients and
gram unless attendance improves), and the treatment providers alike. Any honeymoon
teaching of coping skills to facilitate improved phase would likely be over by this stage of
attendance despite emotional distress. treatment, necessitating that the frustrations,
High-intensity, short-term treatment at disappointments, and realities of treatment be
this level of care should be front-loaded with addressed. For effective treatment to con-
skills training and psychoeducation to pro- tinue, the therapeutic alliance must be strong
mote stabilization and safety. It is essential to enough to endure these obstacles.
select and prioritize elemental skills that are
feasible for patients to learn quickly. This Longer-Term Partial Hospital/Day
bottom-up approach draws heavily from Treatment Programs: Empirical Lessons
cognitive-behavioral therapy (CBT) princi-
ples and simplifies the treatment program Research on longer-term partial hospital and
for patients (Levendusky et al. 1994; Neu- day treatment programs suggests that they
haus 2006). As an extension of cognitive-be- may provide the opportunity for behavioral
Levels of Care in Treatment 173

stabilization (including decreased parasui- from 816 hours/week) did not evidence any
cidal behavior) and symptom improvement improvements in work functioning at 1-year
and, unlike shorter-term programs, can also follow-up (Karterud et al. 2003).
positively affect social and interpersonal Finally, although it is often assumed that
functioning. Bateman and Fonagy (1999) more treatment (i.e., greater intensity) will re-
found that BPD patients treated in a partial sult in greater improvements, research sug-
hospital program organized around a men- gests that when it comes to the day treatment
talization-based model of treatment (Bate- of patients with BPD, less is more. Karterud
man and Fonagy 2004), compared with BPD et al. (2003) found that lower-intensity (i.e., 8
patients in standard outpatient care, evi- 10 hours per week) partial hospital treatment
denced significantly fewer suicide attempts was more effective (i.e., fewer dropouts and
after 6 months, significant reductions in de- greater improvements in global functioning
pression and anxiety after approximately and symptom severity) for patients with BPD
9 months, and significantly fewer acts of self- than high-intensity treatments (approxi-
harm after 12 months. Moreover, partial hos- mately 16 hours per week). This finding is
pital program patients not only maintained consistent with the level of intensity found to
their gains but reported further improve- be effective in Bateman and Fonagys (1999,
ments at follow-up (Bateman and Fonagy 2001, 2008) randomized controlled trial.
2001, 2008).
Global symptom severity of patients with Level II: Intensive
Cluster B and C disorders has been shown to Outpatient Program
decrease significantly after approximately
4 months of level III treatment (Karterud et When described as a level of care, an IOP dif-
al. 1992; Krawitz 1997; Vaglum et al. 1990; fers from an intensive schedule of individual
Wilberg et al. 1998b). These same improve- psychotherapy. For purposes of this review,
ments may also be found for patients with we define IOP as an integration of two or
paranoid personality disorder (Wilberg et al. more modalities in which efforts are coordi-
1998b). Research also suggests that patients nated and patients receive 310 hours of ser-
with Cluster C personality disorders may ex- vices per week. Thus, at its higher end, IOPs
perience improvements in social functioning overlap with low-intensity partial hospital
after 4 months (Karterud et al. 1992; Vaglum programs. When patients need fewer than 3
et al. 1990). For patients with BPD, on the hours of service per week and/or the ser-
other hand, significant improvements in so- vices do not need to regularly coordinate
cial adjustment and interpersonal function- their efforts, the treatment becomes level I.
ing may require up to 18 months of treatment IOP or level II care is of particular value for
(Bateman and Fonagy 1999). BPD patients (Gunderson 2001; Smith et al.
Moreover, even in these longer-term par- 2001). Although Linehans (1993) DBT treat-
tial hospital/day treatment programs, voca- ment has been identified as an outpatient
tional rehabilitation is difficult to achieve. Fol- service, it involves both individual and
lowing a 4-month day treatment program, group therapy provided by collaborating cli-
unemployment rates of patients with Cluster nicians at a level of 3.5 hours per week
B and C disorders did not change during the thus, it could be classified as an IOP.
2-year follow-up period despite other signifi-
cant improvements in global functioning and Indications
symptom severity (Krawitz 1997). Similarly,
in another study, patients with personality IOPs are indicated for personality disorder
disorder (primarily borderline, avoidant, and patients whose problems with living in the
paranoid) treated in a variety of day treatment community are not acutely self-endangering
programs (ranging from 1841 weeks and but are sufficiently severe that only daily, or
174 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

otherwise unusually intensive, care can Behavioral (412 months): Improve abili-
bring about changes. IOPs can provide a ties to control impulsive behaviors and
gradual transition from higher levels of care out-of-control (i.e., ineffective) expres-
(i.e., inpatient and partial hospital pro- sions of feelings. Improve ability to en-
grams). Because patients often experience gage in goal-directed, valued behaviors.
this transition as happening too quickly or Affective (652 weeks): Recognize feeling
before they are ready, treatment modalities states in self and others and learn to asso-
at this level of care should recognize and val- ciate them with behaviors (what Fonagy
idate patients subjective distress while si- [1991] has termed mentalization). This rec-
multaneously encouraging the use of their ognition is especially important for the
personal strengths and skills. feelings of fear and anger in avoidant per-
sonality disorder and BPD, respectively.
Case Example Increase emotional acceptance.
Interpersonal (6 months2 years): Recog-
Ms. H, a 38-year-old, single white
nize dependent needs as part of self and
woman with BPD and posttraumatic
stress disorder (PTSD), was hospital-
others. This goal initially involves recogni-
ized for increased PTSD symptoms and tion of such needs; becoming comfortable
reckless behavior following a serious with them follows (Gunderson et al. 1993).
accident in which her son was injured. Increase interpersonal effectiveness.
Previous hospitalizations had lasted
weeks longer than anticipated and were Structures
followed by 23 months in partial hos-
pital without obvious benefit. On the in- In addition to individual psychotherapy,
patient unit, the patient was disap- IOPs offer complementary groups that meet
pointed not to have the case manager three to five times per week and promote the
and psychiatrist she had worked with
resumption of functional capacities within
in the past and also was eager to go
home to spend Christmas with her son.
the community. Groups are best offered in
Her therapist was leaving for vacation the morning or late afternoon to allow time in
and seemed worried about the patient the patients schedules to pursue nonclinical
returning to outpatient care in her ab- activities such as work or volunteering. Of-
sence. The patient and therapist negoti- fering several different types of groups, each
ated with the leaders of the IOP groups with a different format and function, will
about how she could use the groups to help patients meet the various goals of this
manage emotional distress and to mon-
level of care.
itor any reckless behavior. The patient
was discharged to an IOP, resumed her For instance, daily self-assessment groups
functioning as a mother, and survived can provide a structured format for each pa-
her therapists absence. tient, in turn, to discuss the transition, identify
maladaptive behaviors and/or interpersonal
This vignette illustrates both the value of difficulties, and obtain support. Daily groups,
an IOP as a step-down from hospital (level by virtue of their frequency, enable patients to
IV) care and the ways it can assist during cri- become familiar with the details of each oth-
ses (in this case, the absence of a therapist) ers lives, contributing to their sense of being
while in outpatient (level I) care. heard and understood. The format of these
groups also enables group members to hold
Goals: Social (Behavioral) Adaptation each other accountable for managing mal-
adaptive behaviors and interpersonal prob-
Vocational (412 weeks): Enlist in needed lems. DBT or CBT skills groups in an IOP pro-
vocational training or develop skills and vide patients with the opportunity to further
initiative required to obtain work. develop and refine skills learned in more in-
Levels of Care in Treatment 175

tensive levels of care (e.g., skills to control im- Level I: Outpatient

pulses, regulate emotions, tolerate distress,
and improve their capacity to negotiate rela- Level I is the level of care in which most of
tionships). the treatment for personality disorders oc-
These groups are more rigorous than curs. In a previous generation, psychoana-
supportive groups (such as self-assessment lytic therapies were considered the treatment
ones) and require a serious commitment to of choice for all personality disorders, even
homework assignments and behavioral though the literature mainly consisted of
change. Interpersonal groups are process- negative accounts about the resistances en-
oriented psychotherapy groups that have a countered. Some of the problems tradition-
longer-term focus. These groups encourage ally encountered in outpatient care can be
patients to examine how they relate to others, addressed by utilizing higher levels of care
both within the group and in their lives. Con- especially IOPs. Other problems were due to
flicts between members are expected to be applying psychoanalysis to patients who
frequent and may reflect ambivalence about needed more structure, more of a here-and-
dependency needs and competitiveness for now focus, and more support. To some ex-
attention. Mentalization groupsderived tent, psychoanalytic theory has been modi-
from mentalization-based treatmentare fied to address these problems, acknowledg-
designed to help members develop more ing the importance of the real relationships
awareness of the mental states of other per- and putting insight into perspective. More-
sons and to consider how that awareness af- over, outpatient treatments for personality
fects their own mental state (Bateman and disorders have diversified to include a much
Fonagy 2006). stronger place for CBTs and medications.
The nature and function of these groups
provide useful information as to how best to Indications
incorporate them within the structure of the There are no specific indications or contrain-
IOP. Whereas self-assessment groups may be dications for outpatient care. There are ge-
invaluable during the initial period of transi- neric issues; because outpatient care requires
tion, lasting as little as a few weeks, interper- conscious willful effort, such care is limited
sonal and DBT/CBT skills groups have a de facto to those who seek it and who can be
longer-term focus and may be expected to sufficiently reliable to attend scheduled ses-
continue after the patient leaves the IOP for sions. Beyond these considerations, some pa-
outpatient care (level I), thereby assisting in tients primarily seek support or direction,
this new transition. and their motivation to work on changing
Finally, it is essential that the groups of their personality may not be present. The ab-
the IOP be coordinated and integrated with sence of this motivation may be a relative
the overall goals of the patients treatment. contraindication for outpatient treatment.
When a patients primary clinician works Research on the effectiveness of outpa-
outside the IOP setting, the responsibility for tient treatments for personality disorders has
coordination and implementation of the focused primarily on relatively high-func-
treatment plan can be handicapped. Regular tioning clients. For instance, Winston et al.
communication between the patients psy- (1994) required that clients with a personality
chotherapist and group leaders, as well as disorder have no suicidal behavior, no his-
among the group leaders themselves, is crit- tory of destructive impulse control prob-
ical. For instance, Linehans (1993) DBT in- lems, no use of psychotropic medications in
cludes a weekly 2-hour consultation team the past year, and the presence of one close
meeting, considered to be one of the neces- interpersonal relationship; Alden (1989) ex-
sary components of the treatment. cluded participants who had ever been hos-
176 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

pitalized for psychiatric difficulties, and the primary targets of psychodynamic

most of the participants were employed or in (and psychoanalytic) psychotherapies.
school; in Hoglends (1993) study, most par- Although the attention given to diagnoses
ticipants were employed and none had se- is often limited and idiosyncratic, there is
vere acting-out behaviors. Because we do not a body of evidence relevant to their effec-
know what fraction of the personality disor- tiveness with personality disorders (see
der patients in outpatient services meet these Gunderson and Gabbard 1999). The re-
requirements, the generalizability of this re- sults published by Knight (1941) indi-
search is unclear. What may be concluded is cated that psychoanalytic psychothera-
that the results will not apply to patients with pies are more effective for neurosis
severe social dysfunction. (63%) and character disorders (57%) than
for psychosis (25%). A review of available
Goals: Interpersonal and literature suggests that such therapy is
Intrapsychic Growth particularly likely to help patients with
obsessive-compulsive, narcissistic, and
Goals for level I are often the same as those dependent personality disorders (Gun-
identified for an IOP. Indeed, because most derson 2003).
treatments for patients with a personality The effectiveness of psychodynamic
disorder are delivered in this nonintensive psychotherapy for BPD has received the
outpatient level of care, the goals identified most attention. Although a study con-
for an IOP are usually initiated in level I. ducted at McLean Hospital found that it
However, in outpatient care, achieving these was rare for a patient with BPD to remain
goals is more apt to involve active selection in long-term treatment and get dramati-
and motivation by the particular patient. cally better (Waldinger and Gunderson
An important discontinuity with the IOP 1989), such cases could be identified and
is that outpatient care is rarely directed at vo- the processes of change seemed to occur
cational rehabilitative needs; this may be es- in a predictable sequence (Gunderson et
pecially true for psychodynamic individual al. 1993). Moreover, three studies with
therapies. CBT- or DBT-based approaches larger samples of BPD patients have
may be more likely to accommodate this lack added credibility to the claim that long-
of vocational rehabilitation through role term psychodynamic psychotherapy can
playing and problem-solving issues related be effective for patients with BPD and
to applying for school or work. Still, because have significant cost offsets (Hoke LA:
outpatient therapies depend on what pa- Longitudinal Patterns of Behaviors in
tients identify as goals, and because voca- Borderline Personality Disorder, Doc-
tional rehabilitation is rarely a reason for toral dissertation, Boston University, 1989;
which patients with a personality disorder Howard et al. 1986; Stevenson and Meares
seek therapy, this arena is often neglected. 1992). Of particular note, the psychoana-
lytic psychotherapy offered in the Steven-
Enhance social involvement: Improve son and Meares (1992) study (a 1-year
level of, and satisfaction from, social and manualized treatment conducted by
recreational activities (as demonstrated trainees) was followed by continued im-
by Winston et al. 1994). provement. However, it is important to
Improve impulse control: Although often note that these studies do not show that
begun at higher levels of care, impulse the psychoanalytic components distin-
control can also occur in, or be strength- guish effective psychotherapy, nor do
ened by, outpatient care. they indicate to what extent the favorable
Work on changes in interpersonal related- outcomes can be generalized to the larger
ness and intrapsychic structure: These are universe of BPD patients.
Levels of Care in Treatment 177

Structures forts to interpret his calls or to set limits

on them, and she eventually sought con-
Individual psychotherapy assumes the cen- sultation. The consultant suggested an
tral role in most outpatient treatments for interpersonal group therapy be added.
personality disorders. Still, split treatments In the group, his anxieties about rejec-
(i.e., adding a suitable second modality to ac- tion were seen as unrealistic, and the
company the individual psychotherapy) maladaptive nature of his intrusive
wishes for reassurance were confronted.
have advantages (Gunderson 2001; see also
The patient resisted, but after starting
Chapter 17, Collaborative Treatment). At the group, his behavior in therapy
this level of care, the split treatment may in- changed dramatically. He was able to
volve medications (e.g., for schizotypal, bor- clearly see his reactions to his therapist
derline, or avoidant personality disorders; (and wife) as transference phenomena.
see Chapter 14, Somatic Treatments). The
second modality could also include a social Impasses in individual psychotherapy
rehabilitative component, including a CBT with patients with a personality disorder of-
group (Blum et al. 2002, 2008; Gratz and ten derive from the fact that such patients are
Gunderson 2006), an interpersonal therapy often unaware of how they create problems
group (Marziali and Munroe-Blum 1995; for others and then can feel unjustifiably crit-
Munroe-Blum and Marziali 1995), a self-help icized by therapists who point this out. In
group (e.g., Alcoholics Anonymous, Narcot- many cases, this obstacle can be overcome by
ics Anonymous; see Chapter 19, Substance the addition of a second modality. In this
Abuse), and/or some continuation of fam- case, the group therapy diminished the
ily involvement. These therapies are often in- transference and provided a source of feed-
dicated for dependent, borderline, histrionic, back to Mr. I that was less personalized.
avoidant, or schizoid personality disorders. Some evidence suggests that patients with
Moreover, a meta-analysis of treatment stud- a personality disorder may require longer-
ies on psychopathy suggests that the aug- term treatments to reach normative levels of
mentation of individual psychotherapy with functioning or to maintain treatment gains.
group or family therapy may enhance its ef- Alden (1989) found that although three differ-
fectiveness (see Salekin 2002). en t b eha viora lly b ased t rea tme nts fo r
avoidant personality disorder resulted in
Case Example greater improvements than a waiting list con-
trol group, the patients remained significantly
Mr. I was a 34-year-old man who sought more symptomatic than normative samples.
help because he wanted to reconcile Similarly, in a study of the long-term out-
with his wife, who had kicked him out. comes of patients with and without personal-
He was obsessed with his wife and
ity disorders treated with outpatient dynamic
claimed he did not understand why she
psychotherapy, Hoglend (1993) found that for
had rejected him. She had refused cou-
ples therapy, saying that Mr. I needed to the patients with personality disorders, the
change himself. Prior efforts to assuage number of sessions in treatment was signifi-
his agitation and insomnia with medica- cantly related to acquired insight (i.e., new
tions had proven helpfulbut he now emotional self-understanding) and to overall
needs to change himself, although he personality change at 2- and 4-year follow-up.
could offer no ideas about what he
wanted to change. In the ensuing ses-
sions, Mr. I described a very disturbed
childhood with a punitive mother. CONCLUSION
He became quite devoted to the ther-
apist, a woman, and began calling her Because personality disorders are defined by
frequently for what to her seemed trivial enduring social maladaptations, they are in-
reasons. He was deeply hurt by her ef- trinsically tied to social contingencies. To
178 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

bring about change, good treatments across chy of therapeutic processes (Gunderson
all levels of care must embody coherent and 1978, 2001). Viewed through this lens, the
repetitious interventions with a primary ini- levels of care move from most to least con-
tial focus on the here and now. tainment, with an attendant increase in reli-
The interventions offered by any level of ance on internal controls and self-agency.
care will only be effective if patients develop There is also a progressive decrease in the
an alliance with treaters. For patients with level of structure across the levels of care,
personality disorders, forming an alliance such that the organization of time and activ-
will be complicated by the obstacles created ities imposed by the treatment setting is re-
by their personalities, for example, avoid- duced at each step down. Even within the
ance, deceit, or attention seeking. This chap- least-structured treatment settingthat is,
ter emphasized that establishing an alliance outpatient psychotherapiesthere is a hier-
begins with establishing agreed-upon goals, archy in which more directive and active in-
selecting the level of care appropriate to (i.e., terventions give way to lesser ones as pa-
best able to fulfill) these goals, and clarifying tients progress. Support in the form of
what is expected of patients. These activities reassurance, advice, and expressions of con-
establish a contractual alliance. This form of cern are important elements of all levels of
alliance may be sufficient for hospitals but is care. Because support often bonds people, it
also a necessary prerequisite for all other lev- becomes an essential element for interper-
els of care. The relational alliance (i.e., an al- sonal attachments. Such attachments may be
liance based on liking or trusting the treaters) a negative factor in the highest levels of care
also is necessary. Such alliances usually de- because they can form a resistance to leaving,
velop from supportive attention, and in but they become more important, like the re-
some cases respectful listening may be all lational alliance noted earlier, as patients
that is required. When such an alliance is not move into longer-term settings.
formed with clinicians, progress is unlikely. Winnicott (1965) identified a holding en-
Beyond this relational alliance, the value of vironment as a social context that is a neces-
many specific forms of intervention depends sary prerequisite for the development of an
on what has been termed the working alli- internal sense of safety and security. Every-
ancethat is, an alliance that is needed for one requires a feeling of being securely con-
collaborative work toward the patients tained, but the degree to which this feeling
goals (e.g., acquiring new skills and capabil- depends on external factors varies. Although
ities). Although such a mutual task orienta- Winnicott originally conceptualized the hold-
tion is often assumed when contractual alli- ing environment as a function served by
ances are made, for patients with personality mothers early in development, it has been
disorders, a working alliance can be hard to transformed into a term that is used to de-
achieve. Clinicians working at all levels of scribe a function offered by therapies. As one
care need to be vigilant about whether per- moves through the levels of care, the hold-
sonality disorder patients are working with ing action depends progressively less on im-
the therapist for purposes of changing them- posed constraints and structures and more on
selves (i.e., their identity or self). internal resources. This shift has implications
Earlier in this review, we noted a relation- for the levels of care that may be most appro-
ship between the four levels of care and the priate for specific personality disorders.
domains of psychopathology that patients Although the role of hospitalization tends
are best able to change. Table 93 reflects this to be limited for the treatment of personality
progression as it relates to goals. Another disorders, many patients with personality
way to understand the distinctive effective- disorders use hospitalizations to treat comor-
ness of the four levels of care is via a hierar- bid conditions or crises. Hospitalizations may
Levels of Care in Treatment 179

be most useful for those patients whose sense This research has generally shown that BPD
of self is most disorganized or unstable (BPD) can be responsive to well-structured pro-
or whose unintegrated behaviors pose a dan- grams at all levels of institutional services.
ger to self or others (e.g., BPD and nonpsy- The results for ASPD are more complicated.
chopathic ASPD). Such patients may need Specialized programs at inpatient, partial
more containment and structure to feel hospital, and outpatient levels of care have
held. Partial hospital (day treatment) care been shown to be useful for nonpsycho-
(level III) is used primarily for stabilizing pathic ASPD patients. Although the higher
mental states and initiating longer-term ther- levels of care are generally contraindicated
apies within the 2 weeks usually allotted by for psychopathic ASPD patients, the afore-
the modern managed care environment. mentioned meta-analysis of treatments for
However, there is good evidence that longer- psychopathy suggests that long-term, inten-
term stays can add substantial benefits. The sive individual psychotherapy may have
limited durations of stay available in partial positive results for this population (see Sale-
hospitals have helped create the need for a kin 2002). However, the extent to which these
relatively new level of care, IOPs (level II). research findings generalize to standard
This level of care may be needed for at least practices in outpatient settings is unclear,
1 year and requires theoretically and structur- given the potential for problems related to
ally integrated individual and group compo- treatment retention and compliance at this
nents. These programs have been demon- level of care.
strated to be useful for BPD, but in principle Because research suggests that well-
they should be just as applicable to any per- structured and theoretically consistent pro-
sonality disorder with severe social func- grams are more useful for patients with a
tioning handicaps (e.g., schizoid, avoidant, personality disorder than programs that are
dependent, and antisocial personality disor- not, these qualities should become standards
ders). Ironically, this level of care, arguably for care. More attention should now be given
the most useful for dysfunctional personality to whether programs with different theoreti-
disorders, is the least available in the present cal models (e.g., dynamic vs. cognitive-be-
health care system. Outpatient care is the pri- havioral) have different effects. Research
mary setting for the treatment of most person- should also look at the relative cost-effective-
ality disorders. Here, long-term individual ness of different treatments using follow-up
psychotherapy by itself is thought to be the data, and to the extent indicated, reimburse-
treatment of choice for patients with narcissis- ment policies should be changed accord-
tic, histrionic, and obsessive-compulsive per- ingly. Another area that would benefit from
sonality disorders. Although it is thought that research attention is how well the therapeu-
significant personality change may be possi- tic alliance (and what type of alliance) pre-
ble in outpatient care, availability of this treat- dicts patients subsequent benefits from
ment is largely dependent on private pay and treatment. Furthermore, increased research
thus is frequently inaccessible. attention should focus on the intensive out-
There is very little research relevant to patient level of care (level II), a relatively new
systems of health care services for personal- level that seems particularly promising for
ity disorders. What little there is supports the BPD patients. Finally, given the apparent dif-
value of graduated step-down levels of care. ferences in effective treatments for psycho-
Not surprisingly, the two personality disor- pathic and nonpsychopathic ASPD, research
ders for which higher levels of care are most should continue to distinguish between
needed and that have the most public health these two groups when examining treat-
significance, BPD and ASPD, have generated ments for ASPD as well as begin to identify
the most research attention (see Table 92). the extent to which the positive results found
180 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

for specialized drug treatment programs (at Blum N, Pfohl B, St. John D, et al: STEPPS: a cog-
all levels of care) are generalizable to other nitive-behavioral systems-based group treat-
ment for outpatients with borderline person-
treatment programs for ASPD patients.
ality disorder. A preliminary report. Compr
This chapter provides an overview of the Psychiatry 43:301310, 2002
structures and goals of different levels of care Blum N, St. John D, Pfohl B, et al: Systems Train-
as well as their relative appropriateness for ing for Emotional Predictability and Problem
the treatment of different personality disor- Solving (STEPPS) for outpatients with border-
ders. Attention to and consideration of per- line personality disorder: a randomized con-
trolled trial and 1-year follow-up. Am J Psy-
sonality disorder diagnosis will aid clinicians chiatry 165:468478, 2008
in determining the levels of care likely to be Bohus M, Haaf B, Stiglmayr C, et al: Evaluation of
most effective and clinically indicated and inpatient dialectical behavior therapy for bor-
should be used to inform treatment decisions. derline personality disorder: a prospective
In particular, preliminary evidence suggests study. Behav Res Ther 38:875887, 2000
Bohus M, Haaf B, Simms T, et al: Effectiveness of
the value of using graduated, step-down lev-
inpatient dialectical behavioral therapy for
els of care, although more research is needed borderline personality disorder: a controlled
to determine the specific structure and time- trial. Behav Res Therapy 4:487499, 2004
line most likely to be effective within such a Brooner RK, Kidorf M, King VL, et al: Preliminary
step-down system. evidence of good treatment response in antiso-
cial drug abusers. Drug Alcohol Depend
49:249260, 1998
Brown GK, Newman CF, Charlesworth SE, et al:
An open clinical trial of cognitive behavior
REFERENCES therapy for borderline personality disorder.
J Personal Disord 18:257271, 2004
Alden L: Short-term structured treatment for Cacciola JS, Alterman AI, Rutherford MJ, et al:
avoidant personality disorder. J Consult Clin Treatment response of antisocial substance
Psychol 57:756764, 1989 abusers. J Nerv Ment Dis 183:166171, 1995
Barley WD, Buie SE, Peterson EW, et al: Develop- Compton WM, Cottler LB, Spitznagel EL, et al:
ment of an inpatient cognitive-behavioral Cocaine users with antisocial personality im-
treatment program for borderline personality prove HIV risk behaviors as much as those
disorder. J Personal Disord 7: 232240, 1993 without antisocial personality. Drug Alcohol
Bateman A, Fonagy P: Effectiveness of partial hos- Depend 49:239247, 1998
pitalization in the treatment of borderline per- Davidson K, Norrie J, Tyrer P, et al: The effective-
sonality disorder: a randomized controlled ness of cognitive behavior therapy for border-
trial. Am J Psychiatry 156:15631569, 1999 line personality disorder: results from the Bor-
Bateman A, Fonagy P: Treatment of borderline d e rl i n e Pe rs on a li t y D is o rd e r S t u d y o f
personality disorder with psychoanalytically Cognitive Therapy (BOSCOT) trial. J Personal
oriented partial hospitalization: an 18-month Disord 20:450465, 2006
follow-up. Am J Psychiatry 158:3642, 2001 Dolan BM, Evans C, Wilson J: Therapeutic com-
Bateman A, Fonagy P: Psychotherapy for Border- munity treatment for personality disordered
line Personality Disorder: Mentalization- adults: changes in neurotic symptomatology
Based Treatment. New York, Oxford Univer- on follow-up. Int J Soc Psychiatry 38:243250,
sity Press, 2004 1992
Bateman A, Fonagy P: Mentalization-Based Treat- Dolan B, Warren F, Norton K: Change in border-
ment for Borderline Personality Disorder: A line symptoms one year after therapeutic com-
Practical Guide. Oxford, England, Oxford munity treatment for severe personality disor-
University Press, 2006 der. Br J Psychiatry 171:274279, 1997
Bateman A, Fonagy P: Eight-year follow-up of pa- Dolan-Sewell RT, Krueger RF, Shea MT: Co-occur-
tients treated for borderline personality disor- rence with syndrome disorders, in Handbook
der: mentalization-based treatment versus of Personality Disorders: Theory, Research
treatment as usual. Am J Psychiatry 165:631 and Treatment. Edited by Livesley WJ. New
638, 2008 York, Guilford, 2001, pp 84104
Levels of Care in Treatment 181

Dyck IR, Phillips KA, Warshaw MG, et al: Patterns Harris GT, Rice ME, Cormier CA: Psychopaths: is
of personality pathology in patients with gen- a therapeutic community therapeutic? Thera-
eralized anxiety disorders, panic disorder peutic Communities 15:283299, 1994
with and without agoraphobia, and social Hart SD, Hare RD: Psychopathy: assessment and
phobia. J Personal Disord 15:6071, 2001 association with criminal conduct, in Hand-
Fonagy P: Thinking about thinking: some clinical book of Antisocial Behavior. Edited by Stoff
and theoretical considerations in the treatment DM, Breiling J, Maser JD. New York, Wiley,
of a borderline patient. Int J Psychoanal 72:1 1997, pp 2235
18, 1991 Hildebrand M, de Ruiter C, Nijman H: PCL-R
Gabbard GO, Coyne L: Predictors of response of psychopathy predicts disruptive behavior
antisocial patients to hospital treatment. Hosp among offenders in a Dutch forensic psychiat-
Community Psychiatry 38:11811185, 1987 ric hospital. J Interpers Violence 19:1329, 2004
Gabbard GO, Gunderson JG, Fonagy P: The place Hoglend P: Personality disorders and long-term
of psychoanalytic treatments within psychia- outcome after brief dynamic psychotherapy. J
try. Arch Gen Psychiatry 59:505510, 2002 Personal Disord 7:168181, 1993
Giesen-Bloo J, van Dyck R, Spinhoven P, et al: Out- Howard KI, Kopta SM, Krause MS, et al: The dose-
patient psychotherapy for borderline person- response relationship in psychotherapy. Am
ality disorder: randomized trial of schema- Psychol 41:159164, 1986
focused therapy vs. transference-focused ther- Karterud S, Vaglum S, Friis S, et al: Day hospital
apy. Arch Gen Psychiatry 6:649658, 2006 therapeutic community treatment for patients
Gratz KL, Gunderson JG: Preliminary data on an with personality disorders: an empirical eval-
acceptance-based emotional regulation group uation of the containment function. J Nerv
intervention for deliberate self-harm among Ment Dis 180:238243, 1992
women with borderline personality disorder. Karterud S, Pederson G, Bjordal E, et al: Day treat-
Behav Ther 37:2535, 2006 ment of patients with personality disorders:
Gude T, Vaglum P: One-year follow-up of patients experiences from a Norwegian treatment re-
with Cluster C personality disorders: a pro- search network. J Personal Disord 17:243262,
spective study comparing patients with pure 2003
and comorbid conditions within Cluster C, Knight R: Evaluation of the results of psychoana-
and pure C with pure Cluster A or B con- lytic therapy. Am J Psychiatry 98:434446,
ditions. J Personal Disord 15:216228, 2001 1941
Gunderson JG: Defining the therapeutic processes Koenigsberg HW, Kaplan RD, Gilmore MM, et al:
in psychiatric milieus. Psychiatry 41:327335, The relationship between syndrome and per-
1978 sonality disorder in DSM-III: experience with
Gunderson JG: Borderline Personality Disorder: A 2,464 patients. Am J Psychiatry 142:207212,
Clinical Guide. Washington, DC, American 1985
Psychiatric Press, 2001 Koenigsberg HW, Anwunah I, New AS, et al: Re-
Gunderson JG: Treatment of personality disor- lationship between depression and borderline
ders: an overview. Paper presented at the an- personality disorder. Depress Anxiety 10:158
nual meeting of the American Psychiatric As- 167, 1999
sociation, San Francisco, CA, May 2003 Kool S, Dekker J, Duijsens IJ, et al: Efficacy of com-
Gunderson JG, Gabbard GO: Making the case for bined therapy and pharmacotherapy for de-
psychoanalytic therapies in the current psy- pressed patients with or without personality
chiatric world. J Am Psychoanal Assoc 47:679 disorders. Harv Rev Psychiatry 11:133141, 2003
703, 1999 Kopta SM, Howard KI, Lowry JL, et al: Patterns of
Gunderson JG, Waldinger R, Sabo A: Stages of symptomatic recovery in psychotherapy.
change in dynamic psychotherapy with bor- J Clin Consult Psychol 62:10091016, 1994
derline patients: clinical and research implica- Krawitz R: A prospective psychotherapy outcome
tions. J Psychother Pract Res 2:6472, 1993 study. Aust N Z J Psychiatry 31:465473, 1997
Gunderson JG, Morey LC, Stout RL, et al: Major Lanktree CB, Briere J: Outcome of therapy for sex-
depressive disorder and borderline personal- ual abused children: a repeated measures
ity disorder revisited: longitudinal interac- study. Child Abuse Negl 19:11451155, 1995
tions. J Clin Psychiatry 65:10491056, 2004 Leichsenring F, Leibing E: The effectiveness of
Hardy GE, Barkham M, Shapiro DA, et al: Impact psychodynamic therapy in the treatment of
of Cluster C personality disorders on outcome personality disorders: a meta-analysis. Am J
of contrasting brief psychotherapies for de- Psychiatry 160:12231232, 2003
pression. J Consult Clin Psychol 63:9971004, Levendusky PG, Willis BS, Ghinassi FA: The ther-
1995 apeutic contracting program: a comprehen-
182 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

sive continuum of care model. Psychiatr Q Neuhaus EC, Christopher M, Jacob K, et al: Short-
65:189207, 1994 term cognitive behavioral partial hospital
Linehan MM: Cognitive-Behavioral Treatment of treatment: a pilot study. J Psychiatr Pract
Borderline Personality Disorder. New York, 13:298307, 2007
Guilford, 1993 Ogloff JR, Wong S, Greenwood A: Treating crimi-
Linehan MM, Armstrong HE, Suarez A, et al: Cog- nal psychopaths in a therapeutic community
nitive-behavioral treatment of chronically program. Behav Sci Law 8:181190, 1990
parasuicidal borderline patients. Arch Gen Oldham JM, Skodol AE, Kellman HD, et al: Co-
Psychiatry 48:10601064, 1991 morbidity of Axis I and Axis II disorders. Am
Linehan MM, Heard HL, Armstrong HE: Natural- J Psychiatry 152:571578, 1995
istic follow-up of a behavioral treatment for Perry JC, Bond M: Empirical studies of psycho-
chronically parasuicidal borderline patients. therapy for personality disorders, in Psycho-
Arch Gen Psychiatry 50: 971974, 1993 therapy for Personality Disorders. Edited by
Linehan M, Tutek DA, Heard HL, et al: Interper- Gunderson JG, Gabbard GO (Review of Psy-
sonal outcome of cognitive behavioral treat- chiatry Series, Vol 19; Oldham JM and Riba
ment for chronically suicidal borderline pa- MB, series eds). Washington, DC, American
tients. Am J Psychiatry 151:17711776, 1994 Psychiatric Press, 2000, pp 131
Linehan MM, Comtois KA, Murray AM, et al: Quaytman M, Scharfstein SS: Treatment for severe
Two-year randomized controlled trial and fol- borderline personality disorder in 1987 and
low-up of dialectical behavior therapy vs. 1997. Am J Psychiatry 154:11391144, 1997
therapy by experts for suicidal behaviors and Reiss D, Grubin D, Meux C: Institutional perfor-
borderline personality disorder. Arch Gen mance of male psychopaths in a high-secu-
Psychiatry 7:757766, 2006 rity hospital. Journal of Forensic Psychiatry
Loranger AW: The impact of DSM-III on diagnos- 10:290299, 1999
tic practice in a university hospital. Arch Gen Rice ME, Harris GT, Cormier CA: An evaluation
Psychiatry 47:672675, 1990 of maximum security therapeutic community
Marziali E, Munroe-Blum H: An interpersonal ap- for psychopaths and other disordered offend-
proach to group psychotherapy with border- ers. Law Hum Behav 16: 399412, 1992
line personality disorder. J Personal Disord Richards HJ, Casey JO, Lucente SW: Psychopathy
9:179189, 1995 and treatment response in incarcerated female
McQuillan A, Nicastro R, Guenot F, et al: Intensive substance abusers. Crim Justice Behav 30:251
dialectical behavior therapy for outpatients 276, 2003
with borderline personality disorder who are Salekin RT: Psychopathy and therapeutic pessi-
in crisis. Psychiatr Serv 56: 193197, 2005 mism: clinical lore or clinical reality? Clin Psy-
Mehlum L, Friis S, Irion T, et al: Personality disor- chol Rev 22:79112, 2002
ders 25 years after treatment: a prospective Shea MT, Glass DR, Pilkonis PA, et al: Frequency
follow-up study. Acta Psychiatr Scand 84:72 and implications of personality disorders in a
77, 1991 sample of depressed inpatients. J Personal
Messina NP, Wish ED, Nemes S: Therapeutic com- Disord 1:2741, 1987
munity treatment for substance abusers with Silk KR, Eisner W, Allport C, et al: Focused time-
antisocial personality disorder. J Subst Abuse limited inpatient treatment of borderline per-
Treat 17:121128, 1999 sonality disorder. J Personal Disord 8:268278,
Messina NP, Wish ED, Hoffman JA, et al: Antiso- 1994
cial personality disorder and therapeutic com- Smith G, Ruis-Sancho A, Gunderson JG: An inten-
munity treatment outcomes. Am J Drug Alco- sive outpatient program for patients with bor-
hol Abuse 28:197212, 2002 derline personality disorder. Psychiatr Serv
Messina N, Farabee D, Rawson R: Treatment respon- 52:532533, 2001
sivity of cocaine-dependent patients with antiso- Soloff PH: Algorithm for pharmacological treat-
cial personality disorder to cognitive-behavioral ment of personality dimensions: symptom-
and contingency management interventions. specific treatments for cognitive-perceptual,
J Consult Clin Psychol 71:320329, 2003 affective and impulsive-behavioral dysregula-
Munroe-Blum H, Marziali E: A controlled trial of tion. Bull Menninger Clin 62:195214, 1998
short-term group treatment for borderline Stevenson J, Meares R: An outcome study of psy-
personality disorder. J Personal Disord 9:190 chotherapy for patients with borderline per-
198, 1995 sonality disorder. Am J Psychiatry 149:358
Neuhaus EC: Fixed values and a flexible partial 362, 1992
hospital program model. Harv Rev Psychiatry Tyrer P, Simmonds S: Treatment models for those
14:114, 2006 with severe mental illness and comorbid per-
Levels of Care in Treatment 183

sonality disorder. Br J Psychiatry 44(suppl):15 Widiger TA, Weissman MM: Epidemiology of bor-
18, 2003 derline personality disorder. Hosp Commu-
Tyrer P, Gunderson JG, Lyons M, et al: Special fea- nity Psychiatry 42:10151021, 1991
ture: extent of comorbidity between mental Wilberg T, Friis S, Karterud S, et al: Outpatient
state and personality disorders. J Personal Dis- group psychotherapy: a valuable continuation
ord 11:242259, 1997 treatment for patients with borderline person-
Vaglum P, Friis S, Irion T, et al: Treatment response ality disorder treated in a day hospital. A 3-year
of severe and nonsevere personality disorders follow-up study. Nord J Psychiatry 52:213221,
in a therapeutic community day unit. J Per- 1998a
sonal Disord 4:161172, 1990 Wilberg T, Karterud S, Urnes O, et al: Outcomes of
Van der Sande R, van Rooijen L, Buskens E, et al: poorly functioning patients with personality
Intensive in-patient and community interven- disorders in a day treatment program. Psychi-
tion versus routine care after attempted sui- atr Serv 49:15621467, 1998b
cide: a randomised controlled intervention Wilberg T, Urnes O, Friis S, et al: One-year follow-
study. Br J Psychiatry 171:3541, 1997 up of day treatment for poorly functioning pa-
Waldinger RJ, Gunderson JG: Effective Psycho- tients with personality disorders. Psychiatr
therapy With Borderline Patients: Case Stud- Serv 50:13261330, 1999
ies. Washington, DC, American Psychiatric Winnicott DW: The Maturational Process and the
Press, 1989 Facilitating Environment. London, Hogarth
Widiger TA, Corbitt EM: Antisocial personality Press, 1965
disorder, in DSM-IV Source Book, Vol 2. Ed- Winston A, Laikin M, Pollack J, et al: Short-term
ited by Widiger TA, Frances AJ, Pincus HA. psychotherapy of personality disorders. Am J
Washington, DC, American Psychiatric Asso- Psychiatry 151:190194, 1994
ciation, 1996, pp 703716
This page intentionally left blank
Psychoanalysis and
Psychodynamic Psychotherapy
Glen O. Gabbard, M.D.

The field of psychoanalysis emerged in the up a lost object only by identifying with the
last decade of the nineteenth century as a lost person, suggesting that one seminal as-
means of treatment for hysterical symptoms. pect of the development of personality is
However, in a relatively short time the focus identification with parents and others in the
of psychoanalysis shifted toward long- course of development.
standing character pathology. As early as Freuds work was expanded by Karl Abra-
1908, Freud was inaugurating that shift ham (1923/1948) when he developed a sys-
when he wrote his classic paper Character tem of classifying character traits according
and Anal Eroticism (Freud 1908/1959). to their linkage with oral, anal, and genital
Freud linked specific character traits, such as eroticism. However, it was Wilhelm Reich
miserliness, obstinacy, and orderliness, with (1931) who was the true trailblazer in the psy-
the anal psychosexual stage of development. choanalytic understanding of character. He
Whereas he regarded neurotic symptoms as developed the term character armor to de-
reflecting the return of repressed uncon- scribe the unconscious and ego-syntonic de-
scious material, he viewed character traits as fensive style of patients who come to analytic
the end result of the successful use of repres- treatment. He postulated that childhood con-
sion as well as other defenses such as subli- flicts were mastered with specific defense
mation and reaction formation. As he moved mechanisms. These defenses subsequently
in the direction of the structural model, he emerged in the psychoanalytic setting in the
became aware that identification was of way patients entered the office, reclined on
great importance in the formation of charac- the couch, related to the analyst, and resisted
ter. He recognized that some people can give the psychoanalytic process. Although neu-

186 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

rotic symptoms were regarded as compro- agrees with this particular model, psychoan-
mise formations that produced distress, alysts and psychotherapists alike must accept
Reich stressed that character traits were the limitations of the treatment they under-
rarely sources of anxiety or emotional pain. take. Genetically based temperament is un-
This distinction continues into the present, likely to be altered by psychoanalysis or psy-
when clinicians often remark how the charac- chodynamic psychotherapy, but the areas of
ter traits in patients with personality disor- self-development, internal object relations,
ders often cause more distress in others than and the patients effectiveness in coping with
in the patient. However, this generalization the environment may be profoundly affected
understates the great extent to which many by psychoanalytically oriented treatments.
people with personality disorders suffer as a Temperament is highly stable over time,
result of their character pathology. whereas the character dimensions tend to be
malleable and undergo development through-
out life (Svrakic et al. 1993).
CONCEPTUAL MODEL From a psychoanalytic perspective, per-
sonality can be viewed as having five major
Contemporary psychoanalysis is primarily components: 1) a biologically based temper-
geared to address character. The symptom- ament, 2) a set of internalized object rela-
atic neuroses of Freuds day are rare in to- tions, 3) an enduring sense of self, 4) a spe-
days clinical setting. Analysts focus on how cific constellation of defense mechanisms,
distortions of self, compromises between and 5) a characteristic cognitive style. Ana-
wishes and defenses that oppose those lysts and analytically oriented therapists
wishes, and internal representations of self would regard the achievement of a stable
and others have forged the patients person- and positive sense of self and the establish-
ality (Gabbard 2000b). Psychodynamic psy- ment of mutually gratifying and enduring
chotherapy, which is derived from psycho- relationships as perhaps the two fundamen-
analysis, may be used for Axis I conditions, tal tasks of personality development (Blatt
such as panic disorder and major depression, and Ford 1994). These two fundamental fea-
but it is particularly useful to address the tures of character evolve in a synergistic and
problems associated with Axis II disorders. dialectical relationship throughout the life
Psychoanalysts and dynamic therapists cycle. Blatt (1992; Blatt and Ford 1994; Blatt et
who treat personality disorders today must al. 2007) has stressed that character pathol-
be biologically informed (Gabbard 2001). ogy often divides into two broad subgroups:
There is now abundant evidence that some anaclitic and introjective. The anaclitic type is
personality traits are heritable (Cloninger et mainly concerned with relationships with
al. 1993; Livesley et al. 1993; Svrakic et al. others, and these individuals have longings
199 3) . T h e p syc hob iologica l mo de l of to be nurtured, protected, and loved. The in-
Cloninger et al. (1993), for example, suggests trojective subtype, on the other hand, is pri-
that about 50% of personality is genetically marily focused on self-development, and
based temperament, whereas another 50% is these individuals struggle with feelings of
environmentally based character. These in- unworthiness, failure, and inferiority. They
vestigators stressed that the character vari- are highly self-critical, exceedingly perfec-
ables, based on interactions with family tionistic, and competitive.
members and peer groups, traumatic experi- Because the work of psychoanalysis is
ences, intrapsychic fantasy, and the cultural heavily influenced by transference and coun-
setting in which one develops, are highly in- tertransference developments in the treat-
fluential in determining the subtype of per- ment process, psychoanalytic clinicians tend
sonality disorder. Although not everyone to place a great deal of emphasis on how the
Psychoanalysis and Psychodynamic Psychotherapy 187

patients internal object relations are exter- choanalysts and psychodynamic therapists
nalized in interpersonal relationships with understand the clinical setting as one in
others. The psychoanalytic setting is seen as which patients attempt to re-create their in-
a laboratory in which analysts can directly ternal object relationships through the exter-
observe how their patients re-create their in- nalization of these relatedness patterns
ternal object world in the relationships they formed in childhood.
forge in the course of their daily lives. Hence, The wish-fulfilling nature of actualizing
psychoanalytically oriented therapists tend internal object relationships is clear in the ex-
to conceptualize the nature of the patients ample of an internal self wishing to be loved
psychopathology less in terms of DSM-IV-TR and admired by an internal object. The wish
(American Psychiatric Association 2000) cri- is less apparent in those patients who estab-
teria and more in terms of what unfolds in lish one conflictual and self-defeating rela-
the treatment relationship (Gabbard 1997a, tionship after another. However, even a
2001, 2007). bad or tormenting object may provide
The character dimension of personality is safety and affirmation to a patient for a vari-
usefully conceptualized as involving an on- ety of reasons (Gabbard 2001; Sandler 1981).
going attempt to actualize certain patterns of For abused children, for example, an abusive
relatedness that largely reflect unconscious relationship may be safe in the sense that it is
wishes. Through interpersonal behavior, pa- preferable to having no object at all or to be-
tients try to impose on the clinician a particu- ing abandoned. A basic paradoxical situation
lar way of responding and experiencing. arises in the lives of abused children, where
Character traits, therefore, must be viewed as the person to whom they look for safety and
playing a fundamental role in actualizing an protection is also the abuser. They may have
internal object relationship that is central to a no alternative, then, but to seek safety in the
wish-fulfilling fantasy in the patient (Sandler shadow of one who has abused them. They
1981). The key to understanding the patients may also assume that the only way of re-
relationships outside the treatment, then, maining connected to a significant figure of
may be the observation of what develops in safety is to maintain an abuservictim para-
the transference-countertransference dimen- digm in the relationship. These relationships
sions of the treatment process. may be sought out by patients who were
Developmental themes are at the heart of abused as children because they are reliable,
all psychoanalytic theories. Part of the predictable, and provide the patient with an
conceptual model of personality disorders ongoing sense of continuity and meaning.
assumes that a child internalizes a self-rep- The devil one knows is generally perceived
resentation in interaction with an object rep- as better than the devil one does not know.
resentation connected by an affect state. If, Some of the repetitive relationship pat-
for example, a father repeatedly yells at his terns seen in patients with personality disor-
son, the child internalizes an object relations der are approximations of actual relation-
unit involving a critical, angry object, an in- ships these individuals had with real figures
adequate and beleaguered self, and an affect in the past. However, in some cases they in-
of shame and smoldering resentment. At volve wished-for relationships that never ac-
other times, when the father praises his son, tually existed. Patients with severe child-
the little boy may internalize a loving and ad- hood trauma, for example, often develop
miring object, a good and praiseworthy self, elaborate fantasies about a rescuer who will
and an affect state of glowing self-regard. save them from abuse.
These interactions are etched in neural net- The mode of actualization within the
works and become repetitive patterns of re- clinical relationship is often referred to as
latedness (Westen and Gabbard 2002). Psy- projective identification (Gabbard 1995; Ogden
188 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

1979). Within this model, patients behave in a inations, he could have the fantasy of
characterologically driven way that exerts making his dad and his analyst suffer.
interpersonal pressure on the therapist to Eventually the analyst was also able to
point out to Mr. J that his graduate stu-
conform to what is being projected onto him
dent advisor, another male about the
or her. In other words, a patient may nudge age of his father, was also part of this
the analyst into assuming the role of an same pattern of relatedness. Through
abuser in response to the patients victim interpretation, the analyst made the pa-
role. A patient who treats the therapist with tient aware that he had placed this pa-
contempt, for example, may engender coun- ternal figure in the same role as his fa-
tertransference anger or hate and lead the ther and his analyst.
therapist to make sarcastic or devaluing
comments to the patient. In this model the patients self-represen-
tation can elicit a corresponding object repre-
sentation in the analyst. This model of charac-
Case Example
ter is closely related to the role relationship
Mr. J was a 24-year-old graduate stu-
model of Horowitz (1988, 1991, 1998). In his
dent who was struggling in his aca- theory, a persons schemas reflect uncon-
demic setting because he appeared to be scious selfother organizational units. These
threatened by success. He had all of the units are driven by powerful internal motives
primary symptoms of obsessive-com- that lead away from feared outcomes and to-
pulsive personality disorder (OCPD), ward desired ends. Another way to view these
and his highly perfectionistic expecta- schemas is as belief structures that have both
tions of himself led him to feel that he
form and content. They are often character-
was always failing. In the course of
analysis, he talked at great length about ized, however, by conflicting desires and be-
what a harsh taskmaster his father had liefs that become expressed in the transfer-
been and how he was never able to live ence relationship with the analyst.
up to what his father expected of him. In addition to the development of self in
Eventually this pattern emerged in the relation to objects in the formation of charac-
transference, when the patient revealed ter, psychoanalytically oriented therapists
that he feared that his analyst saw him
study the unique set of defense mechanisms
as a failure as well. He felt the analyst
was heavily invested in having him suc- found in each patient as a key to diagnostic
ceed in graduate school, and the patient understanding and treatment. Defenses ward
was feeling that he could not possibly off awareness of unpleasant affect states and
measure up to what his analyst ex- unacceptable aggressive or sexual wishes
pected. The analyst drew Mr. Js atten- and preserve a sense of self-esteem in the
tion to how he had re-created with the face of narcissistic vulnerability. They may
analyst the same relationship he had
also serve to ensure safety when one is feel-
w ith his father. The p atient co uld
readily see the co nnection, and he
ing threatened. A contemporary analytic
thanked the analyst for pointing it out. perspective, however, would recognize that
However, the analyst also recognized defense mechanisms do not merely change
that the patients deferential and ingra- the relationship between an emotional state
tiating quality was a reaction formation and an idea; they also influence the relation-
to a good deal of resentment about feel- ship between self and object (Vaillant and
ing driven by others to succeed. The an- Vaillant 1999). Patients may be able to man-
alyst thus interpreted that the patient
age unresolved conflicts with important fig-
had found a way to indirectly express
his resentment at both his analyst and ures in their lives or with old objects from the
his father by thwarting their perceived past that haunt them in the present through
ho p es fo r hi s s ucce ss in graduate the use of defenses. In patients with person-
school. By failing to pass his oral exam- ality disorder for whom relationship difficul-
Psychoanalysis and Psychodynamic Psychotherapy 189

ties are one of the major reasons for seeking patients recline on the analysts couch, al-
treatment, dynamic clinicians conceptualize though some prefer to sit for parts of the
defenses as embedded in relatedness. Vail- treatment when visual contact with the ana-
lant and Vaillant (1999) emphasized that the lyst is seen as necessary. The patient is asked
symptoms of patients with personality disor- to say whatever comes to mind in an effort to
der often are designed to cope with unbear- facilitate the process of free association. Pa-
able relationships or unbearable people, tients generally have difficulty saying what
whether in the present or in the past. comes to mind because of anxieties about
Psychoanalysts and psychodynamic what the analyst will think as well as shame
therapists view the specific constellation of about certain aspects of themselves that they
defenses that work in concert with character- find unacceptable.
istic patterns of object relations as having The difficulties encountered in lying on a
enormous importance for the diagnostic un- couch and saying whatever comes to mind
derstanding of the patient. For example, four or five times a week inevitably lead to
someone with OCPD would use defensive the development of transference and resis-
operations such as reaction formation, intel- tance, two of the major foci of psychoanalytic
lectualization, undoing, and isolation of af- treatment. Resistance is not simply the reluc-
fect (Gabbard 2000c). These defenses tone tance to say what comes to mind. It is also a
down powerful affect states so that the pa- manifestation of the patients unique defense
tient is not in danger of losing control. These mechanisms as they enter into the treatment
patients may be responsible, dutiful, and un- process. In other words, defenses are intra-
failingly courteous toward the analyst to be psychic mechanisms, but they become inter-
sure that no trace of aggression is revealed in personalized as resistances in the relationship
their clinical interactions. with the analyst (Gabbard 2000c). Resistance
The fifth component of charactercogni- also reveals significant internal object rela-
tive styleis intimately related to the pa- tionships. A contemporary view of resis-
tients characteristic defenses. Persons with tances would include the fact that they are
OCPD, for example, will come across as lack- forces that oppose the optimal state of con-
ing flexibility and spontaneity in their sciousness sought in analysis. In an ideal an-
thought processes because they are directed alytic process, patients develop a dual con-
toward the control of all affect states. They sciousness in which they relive certain
will also address every detail of a situation in experiences from their past in the transfer-
their pursuit of a perfect solution. Histrionic ence to the analyst while also reflecting on
personality disorder patients, on the other those experiences and being curious about
hand, have a cognitive style that is directly their meanings and origins (Friedman 1991).
linked to their excessive emotionality. Hence Psychoanalysts view resistances not sim-
they avoid detail and will give impressionis- ply as obstacles to be avoided but as a major
tic and global responses to questions that re- source of significant information about pa-
flect the feel of a situation. These cognitive tients characteristic defensive operations
styles appear to be reasonably consistent and their deeply ingrained personality traits.
across personality types (Shapiro 1965). Analysts no longer spend most of their time
in an archeological search for buried relics
from the patients past. Contemporary ana-
MAJOR PRINCIPLES OF TECHNIQUE lysts focus more on the relationship between
analyst and patient as a privileged view of
Psychoanalysis is traditionally conducted in how the patients past has created certain
45- to 50-minute sessions four or five times a patterns of conflict and problematic object re-
week and may last for several years. Most lations in the present. In his 1914 paper, Re-
190 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

me mbe ring , Re pe atin g, an d Work ing - patients usual mode of object relatedness
Through, Freud (1914/1958) noted that outside the treatment situation. When the
what the patient cannot remember will be re- role being evoked by the patient is unfamil-
peated in action in the patients here-and- iar and distressing to the analyst, such as the
now behavior with the analyst, the original role of an abusive parent, some analysts may
meaning of the term acting-out. Hence the pa- feel that an alien force has taken them over,
tients characterological pattern of internal and their subjective experience may be some-
object relations and the conflicts about those thing along the lines of Im not behaving
relationships unfold in front of the analyst like myself (Gabbard 2001). If the wished-
without necessarily digging into childhood for interaction being actualized by the pa-
traumas to unlock hidden secrets. tient is that of an idealized parent who is nur-
While transference reflects the patients turing and understanding toward a needy
past experiences with similar figures, it also child, the analyst may feel quite comfortable
incorporates the real aspects of the analyst. in the role and be unaware of its counter-
Hence transference is now considered to be a transference origins. One of the key compo-
mixture of old relationships from the past nents of technique, then, is for the analyst to
and the new and real relationship with the clarify the nature of these unconscious rela-
analyst in the present. In other words, if an tional patterns, acknowledging that they are
analyst chooses to be aloof, silent, and emo- jointly created, and then make them under-
tionally remote, the patient may well de- standable to the patient. Transference-coun-
velop a transference to the analyst as a cold tertransference enactments may need to re-
and unfeeling figure. peat themselves a number of times before
As the patient repeats long-standing pat- they are apparent and can be interpreted to
terns of relatedness during the sessions, the the patient.
analyst is gradually drawn into a dance. Analysts listen to the development of
Through the process of projective identifica- themes in the associations of the patient. They
tion described earlier, the analyst is trans- carefully track patterns in the transference re-
formed into a transference object. Not all an- lationship that also emerge in narrative ac-
alysts will react in the same way, and the counts of the patients life in the present and
specific features of the analysts intrapsychic in the past. As recurrent themes emerge, they
world will work in concert with what is be- begin to make these unconscious patterns
ing projected by the patient to shape the more available to the patients conscious
unique form of the analysts countertransfer- awareness. A primary intervention is inter-
ence. Some analysts may ignore the role be- pretation, which seeks to make connections
ing thrust on them or reject it. Others may or linkages for the patient that are largely out-
defend against the role by assuming an op- side the patients awareness.
posite stance. Some analysts who are being Psychodynamic psychotherapy is based
pressured to take on the characteristics of a on the same principles of technique but is
projected abusive object, for example, may less focused on interpretation of transference
become overly kind and empathic as a reac- as the central intervention. Moreover, the fre-
tion formation to their growing feelings of quency of sessions (usually from one to three
sadism or anger. per week) is less intense and the patient sits
In any case, a key aspect of the psychoan- in a chair instead of reclining on a couch with
alytic treatment of patients with personality the analyst out of view. Depending on the na-
disorders is to maintain a free-floating re- ture of the psychopathology, some patients
sponsiveness (Sandler 1981) to what is being require more support, more visual contact
evoked by the patient and to use this re-cre- with the therapist, and less regression than
ated dance as a way of understanding the what occurs typically in analysis. Patients
Psychoanalysis and Psychodynamic Psychotherapy 191

with deficit-based pathology have often had patterns of relatedness in action and help the
extensive childhood deprivation or trauma. patient become aware of these patterns. A
They may hear interpretations as attacks and therapist may observe, for example, that the
feel shamed by them. In those situations, the patient looks ashamed whenever talking
therapist may need to use affirmative inter- about his mother and will thus point that out
ventions (Killingmo 1989) that confirm ex- to the patient. Fonagy (1999) stresses that a
actly the way the patient is feeling and em- crucial avenue for therapeutic change may
pathically validate the patients right to feel lie in a patients increasing capacity to find
that way. Many patients with this type of himself in the therapists mind. By consis-
background may ultimately be able to use in- tently observing and commenting on the pa-
terpretations if the way is paved for them by tients feeling states and nonverbal commu-
validating and affirming interventions (Gab- nications, the patient may begin to assemble
bard et al. 1994). a portrait of himself or herself based on the
The goals of the treatment vary according analysts observation and thereby develop
to the patients presenting complaints, the na- increased capacity for mentalization. Consis-
ture and severity of the psychopathology, and tent observations about the characterological
the clinician and analysts theoretical model. patterns of the patient also make ego-syn-
Among those goals that are commonly estab- tonic character traits more ego-dystonic as
lished are resolution of conflict (Brenner the patient recognizes the problematic as-
1976); a search for an authentic or true self, as pects of the traits as well as the interpersonal
Winnicott (1962) suggested; improved rela- impact that the traits have on others.
tionships as a result of a gain in understand- In all forms of psychoanalysis and psy-
ing about ones internal object relationships chodynamic psychotherapy, another source
(Gabbard 1996); an improved capacity to seek of therapeutic action is internalization of the
out appropriate selfobjects (Kohut 1984); the analyst or therapist and of the treatment re-
generation of new meanings within the ther- lationship. Internalizing does not necessarily
apeutic dialogue (Mitchell 1997); and an im- require the use of a conscious, declarative
proved capacity for mentalization (Fonagy representation. The clinical relationship itself
and Target 1996). (Selfobjects and mentalization is accompanied by unconscious affective
are defined and discussed later in the chap- connections that have been referred to by Ly-
ter.) Regardless of the diverse goals, all psy- ons-Ruth et al. (1998) as implicit relational
choanalytically oriented treatment probably knowing. This phenomenon refers to mo-
works through several modes of therapeutic ments of meeting between clinician and pa-
action, of which one is the provision of insight tient that are not symbolically represented or
through interpretation. dynamically unconscious in the ordinary
Another mode of therapeutic action is sense. Some change occurs in the realm of
simply making observations from an outside procedural knowledge involving how to act,
perspective on what one sees in the patient feel, and think in a particular relational con-
(Gabbard 1997b). Patients cannot know how text. As patients internalize the therapists
they come across to others because they are accepting and tolerant attitude, their super-
inside themselves. The therapist has the per- ego is also modified so they are less self-crit-
spective of an object and therefore can help ical and more accepting of their humanness.
them see things that they do not see. More- From a cognitive neuroscience perspec-
over, the how to of relatedness that is inter- tive, the internalization of the therapeutic
nalized in the earliest childhood relation- relationship gradually builds a new neural
ships is embedded in implicit procedural network with a different type of object repre-
memory (Gabbard and Westen 2003). Clini- sentation and a corresponding self-represen-
cians can see the automatic and unconscious tation. The old networks are not completely
192 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

obliterated by the treatment, but they are rel- Another feature that is necessary for
atively weakened or deactivated while the analysis or dynamic therapy is the capacity
new networks based on the treatment rela- to regress in the service of the ego. Patients
tionship are strengthened (Gabbard and must be able to let down their guard, relax
Westen 2003). their defenses, and get in touch with primi-
All of these techniques and modes of tive and unpleasant emotional states to learn
therapeutic action are adapted to the individ- about what drives them to behave or think in
ual patient and the type of personality disor- the way they do. They must also have high
der that most closely fits that patient. There is tolerance for frustration, intact reality test-
inevitably a trial-and-error component to ing, reasonably good impulse control, and
this approach as one finds out which types of enough suffering to motivate them for the
interventions are most suited to the patients treatment.
capacity to use psychoanalytically oriented Other signs of ego strength that make a
treatment. person a reasonable candidate for psycho-
analysis are the ability to sustain a job over a
long period of time despite encountering dif-
INDICATIONS AND ficulties and the capacity for enduring mean-
CONTRAINDICATIONS ingful relationships with others. Finally, the
presence of the capacity for mentalization
In determining for whom psychoanalysis based on secure attachment also is a positive
and/or psychodynamic psychotherapy is in- sign for analyzability. Mentalization refers to
dicated, two separate but related perspec- ones ability to differentiate inner from outer
tives are necessary: 1) suitability according states and to recognize that ones perceptions
to the psychological characteristics of the pa- are only representations rather than accurate
tient, and 2) suitability according to diag- replicas of external reality. In other words,
noses. Regardless of whether a patient meets the patient is aware that his or her own and
diagnostic criteria for a specific personality other peoples ways of viewing the world are
disorder, the patients psychological features influenced by inner beliefs, feelings, and past
may contraindicate the use of psychoanaly- experiences. This capacity may be present to
sis or dynamic therapy. a greater or lesser extent and exists on a con-
The indications and contraindications for tinuum that is influenced by the nature of the
psychoanalysis and psychodynamic psycho- relationship and the patients early experi-
therapy according to personality disorder ence. Some patients with impaired mental-
are summarized in Table 101. ization may nevertheless be amenable to
Foremost among the psychological char- psychoanalysis and psychodynamic psycho-
acteristics necessary to recommend psycho- therapy (Fonagy 2001).
analytically oriented treatment is psycholog- Patients in a severe life crisis are rarely
ical mindedness. Although there are various suited for psychoanalytically oriented treat-
components to this construct, the key com- ment, although it may ultimately be appropri-
ponents are the capacity to see meaningful ate after the crisis is over. Other contraindica-
connections between ones difficulties and tions are poor reality testing, poor impulse
ones inner world. The capacity to think in control, lack of psychological mindedness, lit-
terms of analogy and metaphor is also crucial tle capacity for self-observation, cognitive im-
to psychological mindedness. In addition, pairment based on neurological dysfunction,
there must be a curiosity about the origins of extreme concreteness, and poor frustration
ones suffering and a strong motivation to tolerance (Gabbard 2004). These guidelines
endure anxiety and discomfort in the process are helpful in assessing analyzability, but ana-
of learning more about oneself. lysts recognize that they must retain a certain
Psychoanalysis and Psychodynamic Psychotherapy 193

Table 101. Indications and contraindications for psychoanalysis and psychodynamic

psychotherapy according to personality disorder diagnosis

Diagnosis Suitability for psychoanalytically oriented treatment

Paranoid personality disorder Rarely indicated

Schizoid personality disorder May be indicated in exceptional circumstances
Schizotypal personality Contraindicated
Borderline personality Psychodynamic psychotherapy, particularly transference-
disorder focused psychotherapy and mentalization-based therapy,
are efficacious
Narcissistic personality Strong indication for psychoanalysis or psychodynamic
disorder psychotherapy
Antisocial personality disorder Contraindicated
Histrionic/hysterical Strongly indicated for hysterical personality disorders and
personality disorders for most patients with histrionic personality disorder
Obsessive-compulsive Strong indication for psychoanalysis or psychodynamic
personality disorder psychotherapy
Avoidant personality disorder Indicated for cases that do not respond to brief cognitive-
behavioral or behavior therapy treatments
Dependent personality Likely to do well in psychoanalysis if motivation is sufficient
disorder and usually responds well to psychodynamic therapy
Masochistic or self-defeating Strong indication for psychoanalysis or dynamic therapy
personality disorder

degree of humility because prediction of how sonality disorders, clinicians must rely on
a particular patient will do in the psychoana- clinical wisdom and the psychological charac-
lytic process is less than perfect. Kantrowitz teristics favoring analyzability described
(1987), in a study of 22 patients in analysis, above. Moreover, the presence of various co-
concluded that even with highly sophisti- morbid conditions on Axis I, such as affective
cated psychological testing, clinicians cannot disorder, anxiety disorder, eating disorder, or
reliably predict who will do well in psycho- substance abuse, may complicate psychoana-
analysis. The following sections describe the lytic treatment even if the personality disor-
indications and contraindications for psycho- der itself is likely to be amenable to this ap-
analysis in patients with specific personality proach. As a general principle, the Cluster C
disorders. personality disorders appear to be amenable
Determining the appropriateness of psy- to dynamic therapy or psychoanalysis. Only a
choanalysis or psychodynamic therapy is also very small subgroup of patients with Cluster
influenced by ones diagnostic understanding A personality disorders are likely to respond
of the personality disorder, but this method, well to psychoanalytic efforts. Those patients
too, is imperfect in predicting outcomes. In with Cluster B personality disorders respond
the absence of randomized controlled trials of variably, depending on the diagnosis and the
psychoanalytic treatment for the various per- psychological characteristics.
194 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

Paranoid Personality Disorder borderline personality disorder. Problems of

impulsivity and difficulties in maintaining a
In a study of 100 patients who applied for therapeutic alliance make psychoanalytic
psychoanalysis at the Columbia Psychoana- treatment extremely challenging for patients
lytic Center, Oldham and Skodol (1994) in this category. There are reports in the liter-
noted that 12 met research criteria for the di- ature (Abend et al. 1983; Boyer 1977; Fonagy
agnosis of paranoid personality disorder. Of and Target 1996; Gabbard 1991) of patients
those, only four were selected for analysis, with borderline pathology who were ana-
and two of the four did not finish the analytic lyzed using the couch with somewhat modi-
treatment. Their data suggested that most fied forms of psychoanalytic technique. Bor-
patients with paranoid personality disorder derline patients who can use an analytic
are not suited, but for a very small number process represent a very small subgroup and
with exceptional characteristics analysis may are probably closer to the Kernberg construct
be worth a try. In general, paranoid patients of borderline personality organization
do not have sufficient trust to allow for the (Kernberg 1975) than a DSM-IV-TR border-
development of an analytic process. line personality disorder. These patients are
often conceptualized as part of the widen-
Schizoid and Schizotypal ing scope of psychoanalysis and generally
Personality Disorders require supportive interventions to make in-
terpretation acceptable to the patient (Hor-
Schizotypal personality disorder is thought
witz et al. 1996). Both mentalization-based
to be genetically linked with schizophrenia,
therapy (MBT) (Bateman and Fonagy 1999,
and patients with schizotypal personality
2001, 2008) and transference-focused psy-
disorder are rarely, if ever, suitable for psy-
chotherapy (TFP) (Clarkin et al. 2007) have
choanalysis or psychodynamic therapy. Pa-
been demonstrated efficacious in random-
tients with schizoid personality disorder, on
ized controlled trials.
the other hand, may in some cases be appro-
priate for analytically oriented treatment but
rarely seek it. In the study by Oldham and Narcissistic Personality Disorder
Skodol (1994), only one person applying for
Most experts in the treatment of narcissistic
psychoanalysis was diagnosed with schizoid
personality disorder agree that psychoanaly-
personality disorder. The British psychoana-
sis is the treatment of choice if the patient has
lytic literature suggests that there may be a
the psychological and financial resources to
small number of patients who can be reached
undertake a commitment to psychoanalysis.
by psychoanalytic approaches, and a num-
Although the treatments are long and ardu-
ber of British analysts have reported on work
ous, sometimes nothing short of this in-
with these patients (Balint 1979; Fairbairn
depth approach will touch a patient. In prac-
1954; Winnicott 1963/1965).
tice, many narcissistic patients are treated
with psychodynamic psychotherapy, and
Borderline Personality Disorder
clinical experience suggests that they may re-
After surveying the entire literature on the spond favorably if motivated. A frequent
treatment of borderline personality disorder, challenge in working with patients with nar-
the American Psychiatric Association prac- cissistic personality disorder is the common
tice guideline concluded that psychother- pattern of exaggerated self-importance, com-
apy, rather than psychoanalysis, in concert bined with devaluation of others, frequently
with medication, is probably the treatment of leading to counter-transference reactions in
choice for the great majority of patients with the therapist.
Psychoanalysis and Psychodynamic Psychotherapy 195

Case Example ject relations, and an intact superego. There-

fore, they are considered good candidates for
A patient with narcissistic personality dis- psychoanalysis or psychodynamic psycho-
order had been in psychodynamic psy- therapy. Those patients on the other end of
chotherapy for 5 weeks with a young ther-
the spectrum with histrionic personality dis-
apist in training. The patient began one
session by complaining about the thera- order may require modified versions of psy-
pists relative youth. He asserted that the choanalytic psychotherapy because they of-
therapist seemed to be a beginner who ten cannot tolerate the intense affective states
probably did not know what he was do- that are brought about in analytic treatment.
ing. Am I your first therapy case? he Some histrionic patients, however, appear to
taunted. The candidate replied, What
makes you ask that? The patient laughed
be able to make use of psychoanalysis and
and said, Ill bet your supervisor told you dynamic therapy.
not to answer my questions, right? The
candidate, feeling he was being treated
with contempt, became defensive and
said, No, Im able to think for myself. Personality Disorder
The patient responded with further con-
tempt, I dont see much evidence of Patients with OCPD, who must be differenti-
that. Without considering the potential ated from those with obsessive-compulsive
consequences of his comment, the thera- disorder, are generally good candidates for
pist blurted out, Maybe youre too busy
psychoanalytically oriented treatments.
insulting me to notice. As soon as the
words came out of his mouth, the thera- While there are no data from rigorously con-
pist felt a deep sense of shame. He had al- trolled trials on the efficacy of psychoanaly-
lowed himself to be nudged into sarcasm sis, psychodynamic therapy has been shown
by the patients contempt. to be useful in Cluster C personality disor-
ders, including OCPD (Svartberg et al. 2004).
Antisocial Personality Disorder
Because of lack of motivation, insufficient su- Avoidant and Dependent
perego development, general dishonesty, Personality Disorder
and impulsivity, patients with antisocial per-
sonality disorder represent a contraindica- Some patients with avoidant personality dis-
tion to psychoanalysis or psychodynamic order appear to respond well to behavioral
psychotherapy. or cognitive-behavioral techniques (Alden
1989; Brown et al. 1995; Stravynski et al.
Histrionic and Hysterical 1982). However, when patients do not re-
spond to brief behav ioral or cognitive-
Personality Disorders
behavioral treatments for avoidant personal-
The DSM-IV-TR construct of histrionic per- ity disorder, they may do well in analysis or
sonality disorder represents a particular type psychodynamic psychotherapy, particularly
of patient with personality disorder who is if they are motivated to understand the ori-
very close to the borderline diagnosis. In gins of their anxieties about intimate rela-
clinical practice, one frequently encounters a tionships with others.
higher level of histrionic personality disor- Patients with dependent personality dis-
der that has traditionally been referred to as order may do well in either psychoanalysis
hysterical personality disorder (Gabbard or psychodynamic psychotherapy depend-
2000c). These patients represent a neuroti- ing on their psychological mindedness, the
cally organized individual with reasonably extent of their suffering, and their motivation
good impulse control, mature triangular ob- to change and understand themselves.
196 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

Masochistic or Self-Defeating our understanding of patients with schizoid

Personality Disorder personality disorder, viewed the schizoid
retreat from object relations as a defense
Although not in the official DSM-IV-TR no- against a conflict between a wish to relate to
menclature, patients with predominantly others and a fear that ones neediness would
self-defeating or masochistic symptom pat- harm others. The child who initially per-
terns are widely seen in clinical practice. For ceives its mother as rejecting may withdraw
most of these patients, psychoanalysis or dy- from the world; however, the infants greed
namic therapy is the treatment of choice, pro- and neediness grow until they are experi-
vided they have the psychological character- enced as insatiable. The child then fears that
istics necessary for the treatment. its greed will devour the mother, resulting in
her disappearance. Relationships are experi-
enced as dangerous and to be avoided.
PSYCHOANALYTIC APPROACHES TO The clinician who endeavors to treat pa-
SPECIFIC PERSONALITY DISORDERS tients with schizoid personality disorder
must recognize that their fear of relation-
As noted above, the psychoanalytically ori- ships will manifest itself in the transference
ented therapists technique needs to be tai- and allow the patient the opportunity to re-
lored to the specific type of personality dis- treat without making demands for more ac-
order. Although most personality disorders tive participation. Winnicott (1963/1965) be-
are mixed, in that they have traits of several lieved that the isolation of the patients with
different personality disorders, here we con- schizoid personality disorder preserves an
sider each of the personality disorders ame- important authenticity that is absolutely sa-
nable to psychoanalysis and psychodynamic cred to the evolving self of the patient: There
psychotherapy in its pure form for the sake is an intermediate stage in healthy develop-
of clarity. In actual practice, several technical ment in which the patients most important
approaches may need to be combined for pa- experience in relation to the good or poten-
tients who have different personality fea- tially satisfying object is the refusal of it
tures in mixed personality disorder. The dis- (p. 182). Hence the therapist must recognize
cussion here is confined to those personality that the schizoid withdrawal is a way to com-
disorders that are likely to benefit from psy- municate with the true self within the pa-
choanalysis or psychodynamic psychother- tient instead of sacrificing that authenticity
apy approaches at least some of the time. to artificial interactions with others, includ-
ing the analyst, that would lead to a false
self adaptation.
Schizoid Personality Disorder
Although patients with schizoid person-
Much of our understanding of the inner ality disorder do not commonly seek help
world of a patient with schizoid personality from a psychoanalytically oriented therapist,
disorder derives from the work of the British when they do, the therapist must provide the
object relations theorists. Balint (1979) patient with a holding environment in which
viewed these patients as having a fundamen- their frozen internal object relations will
tal deficit in their ability to relatea basic eventually thaw through the provision of a
fault caused by significant inadequacies in new experience of relatedness. The goal of
the mothering they received as infants. He the psychoanalytic approach, then, is more to
believed that the patients difficulty in relat- provide a new relationship for internaliza-
ing to others stems from this basic incapacity tion than to interpret unconscious conflict.
rather than from neurotic conflict. Fairbairn Extraordinary patience is required for the
(1954), perhaps the foremost contributor to treatment to take effect, and the patient must
Psychoanalysis and Psychodynamic Psychotherapy 197

be allowed to go at his or her own pace. Ther- from being locked into the reality of one ex-
apists must recognize that their own agenda clusive view. In the MBT technique the focus
for change may get in the way of the patients is mainly on the patients current state of
timetable for venturing out into a world of mind. Transference interpretations that em-
potentially dangerous relationships. phasize genetic reconstructions are generally
not used. This approach stresses mentaliz-
Borderline Personality Disorder ing the transference. In other words, trans-
ference provides an opportunity to explore
Borderline personality disorder has a greater how the mind works with the therapists as-
research base than other personality disor- sistance. The therapist encourages the pa-
ders regarding the efficacy of psychodynamic tient to think about the present relationship
psychotherapy. As noted earlier, both MBT with the therapist in the same way that other
and TFP have demonstrated efficacy in ran- relationships are considered. Another mind
domized controlled trials. is in the room with the patient, so that the pa-
Mentalizing is defined as an understand- tients view can be contrasted with the ther-
ing that other peoples perceptions and be- apists view. The transference is used to dem-
havior are related to their thoughts, wishes, onstrate how the same behavior may be
feelings, and desires. In other words, they experienced differently and thought about
have a mind that interprets reality based on differently by two different minds.
who they areperceptions are not absolute Bateman and Fonagy (1999, 2001, 2008)
truths, but impressions. This capacity varies tested MBT in a randomized controlled trial
considerably from person to person. Bate- in which 38 patients with borderline person-
man and Fonagy (2004) base their MBT tech- ality disorder were randomly assigned to a
nique on the observation that patients with psychoanalytically oriented partial hospital
borderline personality disorder have a frag- program or to a control group. The partial
ile mentalizing capacity that is vulenerable to hospitalization condition (treatment group)
interpersonal and social interactions. This consisted of once-weekly individual psycho-
difficulty relates to the patients insecure ca- analytic psychotherapy, thrice-weekly group
pacity for attachment, which in turn can be psychotherapy, once-weekly expressive ther-
linked to the insecure attachment to the care- apy informed by psychodrama techniques, a
giver during childhood. As a result, the at- weekly community meeting, regular meet-
tachment context of psychotherapy is essen- ings with the case coordinator, and medica-
tial to improve ones capacity for mentalizing. tion review by a resident psychiatrist. The
As one builds a solid therapeutic alliance with control treatment consisted of regular psy-
the therapist, one is also building a secure at- chiatric review an average of two times per
tachment base. month with a senior psychiatrist, inpatient
The MBT therapist spends a good deal of admission as appropriate, outpatient and
time during psychotherapy sessions on community follow-up, no psychotherapy,
drawing the patients attention to the pa- and medication similar to that received by
tients mental states. By systematically ex- the treatment group.
ploring what it is the patient feels from mo- At 18-month follow-up, the investigators
ment to moment, the therapist helps build a found a clear reduction in the proportion of
sense of self in the patient. The dyadic nature the treatment groups patients who had at-
of psychotherapy provides the opportunity tempted suicide in the previous 6 months.
to generate multiple perspectives on situa- The average length of hospitalization in-
tions that may seem absolute. The therapist creased dramatically in the control group
repeatedly re-presents the patients subjec- durin g th e last 6 mont hs of th e stu dy,
tive experience in a way that frees the patient whereas it remained stable in the treatment
198 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

group at around 4 days per 6 months. Self- domains to an approximately equivalent ex-
reported state and trait anxiety both de- tent. In certain areas, though, TFP seemed to
creased substantially in the treatment group do better than the alternative treatments. It
but remained unchanged in the control was associated with significant improvement
group. Depression scores also decreased sig- in 10 of the 12 variables across the six symp-
nificantly in the treatment group. Five years tom domains, compared with improvement
after discharge from MBT, the treatment in six variables with SP and five with DBT.
group continued to show clinical and statis- Only TFP resulted in significant changes in
tical superiority to the control group. impulsivity, irritability, verbal assault, and di-
TFP was developed by Otto Kernberg, rect assault. Both TFP and DBT did better
John Clarkin, and their colleagues at New than SP in reducing suicidality. TFP also
York Presbyterian HospitalCornell Medi- brought about a move from an insecure at-
cal Center, Westchester Division (Clarkin et tachment classification to a secure one (Levy
al. 2007). It is based on an object relations et al. 2006), a change not seen with the other
model of borde rline psy chopathology. two treatments. In addition, the TFP patients
Hence TFP emphasizes representations of showed significantly greater changes in men-
self and others and how they are activated talizing capacity and in narrative coherence
with the therapist and with others in signifi- compared with those in other groups.
cant relationships outside the therapy. Affect The research on borderline personality
states connected with internal representa- disorder suggests that more than one ap-
tions of self and others are also a major focus proach may be effective. Hence the clinician
of the treatment. Controlling negative affect must tailor the therapeutic approach to the
states by understanding their origins is cen- particular patient (Gunderson 2001; Gunder-
tral to the treatment. In addition, the thera- son and Links 2008). Some patients, for exam-
pist hopes to awaken positive affect, particu- ple, may find transference work disturbing
larly through the protection of positive and will prefer a focus on outside relation-
attachments to others. TFP establishes a con- ships or the acquisition of coping skills.
tract regarding the expectations of the pa-
tient, the therapist, and the treatment. The Narcissistic Personality Disorder
therapist thus sets a frame designed to con-
trol destructive behavior and systematically The technical approaches to narcissistic per-
examines patterns of relationships to others sonality disorder tend to diverge along a di-
in the therapeutic relationship itself and in viding line marked by whether the clinician
outside relationships. In contrast to MBT, views the pathology as based on deficit, ac-
TFP therapists interpret aggression in the cording to the self psychology of Kohut
transference relationship and link it to its ge- (1971), or as related to conflict, as described
netic and intrapsychic origins. by Kernberg (1974a, 1974b, 1984). The differ-
TFP has been tested in a randomized con- ences between the two approaches are sum-
trolled trial involving 90 patients with bor- marized in Table 102.
derline personality disorder who were ran- Kohuts approach to the psychoanalysis
domly assigned to one of three treatment of narcissistic personality disorders used em-
groups: 1) TFP, 2) dialectical behavior therapy pathy as the cornerstone of the technique
(DBT), or 3) supportive psychotherapy (SP). (Ornstein 1974, 1998). Kohut stressed the
Six domains of outcome measures were as- need to empathize with the patients experi-
sessed at 4-month intervals over a 12-month ences of having numerous empathic failures
period by raters who were blind to treatment at the hands of parents. In this regard, he
group. All three treatments appeared to have advised analysts to accept the patients com-
brought about positive change in multiple ments in a straight manner, just as the pa-
Psychoanalysis and Psychodynamic Psychotherapy 199

Table 102. Techniques for treating narcissistic personality disorders: Kohut versus

Kohut Kernberg

Views mirror and idealizing transferences Views mirror and idealizing as aspects of
as two different poles of bipolar (Kohut transference related to projection and
1977) or tripolar (Kohut 1984) self reintrojection of patients grandiose self
Accepts idealization of patient as normal Interprets idealization as a defense
developmental need
Empathizes with patients feeling as an Helps patient see his or her own contribution
understandable reaction to failures of to problems in relationships
parents and others
Accepts patients comments at face value, Confronts and interprets resistances as
viewing resistances as healthy psychic defensive maneuvers
activities that safeguard the self
Looks at the positive side of patients Examines both positive and negative aspects
experience of patients experience (if only positive
experiences are emphasized, the patient
may develop an increased fear of internal
envy and rage)
Calls attention to patients progress Focuses on envy and how it prevents patient
from acknowledging and receiving help
Has treatment goal of helping patient Has treatment goal of helping patient to
acquire ability to identify and seek out develop guilt and concern and to integrate
appropriate selfobjects idealization and trust with rage and

Source. Based on Gabbard 2000c.

tient experiences them, rather than to inter- transferences typical of narcissistic personal-
pret hidden meanings (Miller 1985). Kohut ity disorders. The analysis of those transfer-
would accept this idealizing transference as a ences is the major part of the technical ap-
normal developmental need rather than in- proach. Kohut also strongly endorsed calling
terpreting what might lie beneath it. He attention to the positive aspects of the pa-
viewed the mirror and idealizing transfer- tients experience to avoid shaming or criti-
ences as the major developments in the anal- cizing the patient and repeating the traumas
ysis of narcissistic personality disorder. The of childhood. He would not hesitate to call
mirror transference is an effort on the patients attention to the patients progress. The goal
part to capture the gleam in mothers eye of the analysis was to help the patient acquire
by trying to impress the analyst. Sensitive to more appropriate and mature selfobjects
the patients potential for self-fragmentation, with the full recognition that the patient
Kohut would empathize with the patients could not possibly outgrow the need for self-
need to be affirmed and validated. object responses, such as idealization, affir-
Kohut (1984) later postulated a third mation, and validation.
transference, the twinship transference. All Kernbergs (1974a, 1974b) approach dif-
three of these were regarded as the selfobject fers in that he sees the patients grandiose
200 E ss e n t i a l s o f P e r s o n a l i ty D i s o r d e r s

self as a highly pathological and conflict- riety of narcissistic personality disorder (Gab-
based solution that is not to be regarded as bard 2000c), they feel coerced into focusing
simply an arrested development of the nor- their attention completely on the patient dur-
mal self. He regards idealization as a defense ing every moment of the session, and they
against rage, contempt, and envy, and he ad- may feel controlled into meeting the patients
vocates active interpretation of idealization. unrealistic expectations of perfect attune-
In general, Kernbergs approach is more con- ment. If they fall short, they risk an explosion
frontational than Kohuts, and he stresses the of narcissistic rage by a patient who sees a
patients greed and demandingness rather slight around every corner. Narcissistic pa-
than his or her longings for affirmation. He tients may erupt with barrages of contempt
also focuses to a greater extent on envy and and hatred when they feel they have not been
how it prevents the patient from acknowl- given the attention they are due.
edging and receiving help. Whereas Kohut
views resistances as healthy psychic activi- Hysterical/Histrionic
ties designed to safeguard the self, Kernberg Personality Disorders
confronts and interprets resistances as defen-
sive maneuvers. He would see the goal of Patients with hysterical personality disorder
psychoanalysis as much broader than Ko- (i.e., the higher-functioning patients within
huts. He would hope that the patient would the histrionic personality disorder rubric)
develop a greater sense of guilt and concern generally are ideally suited for psychoanaly-
while integrating idealization and trust with sis and psychoanalytic therapy. They have
rage and contempt. internalized conflicts about relationships
In actual practice, many psychoanalysts that can be examined in a solid therapeutic
and dynamic therapists use elements of both alliance with the analyst. They are generally
Kernberg and Kohut in their approach to committed to change and to improving their
narcissistic patients. An empathic perspec- chronic difficulties in forming intimate rela-
tive, as described by Kohut, is often extremely tionships. The analyst may need to focus ini-
helpful early in the treatment to form a thera- tially on the global and impressionistic cog-
peutic alliance with the patient so that the pa- nitive style that prevents these patients from
tient is willing to explore his or her inner reflecting in detail on what is happening in-
world. Over time, patients often are more ca- side. Their displays of emotionality can be
pable of responding to interpretation and highly appealing and lead the analyst to feel
confrontation when they feel a stable rela- a sense of obligation to rescue the patient.
tionship has been established with the ana- Nevertheless, for the analyst to be useful, the
lyst. Both Kohut and Kernberg identify valid patient first needs to provide more details of
aspects of narcissistic pathology, but differ- the situations that create distress.
ent types of narcissistic patients require dif- Superficial and shallow feelings may de-
ferent emphases. fend against more disturbing and more
Regardless of which technical strategy is deeply experienced affects. Patients may need
chosen, clinicians treating narcissistic pa- assistance in identifying their true feelings
tients encounter formidable countertransfer- and the thoughts and events linked to them.
ence difficulties. They often have a sense that They often feel buffeted by powerful feeling
they are being used as a sounding board states and by external events. The analyst can
rather than a person with a separate internal help them to identify patterns in their lives
world. They may need to tolerate long peri- that lead to particular types of emotional re-
ods in which they feel they are peripheral to sponses. Horowitz (1977) pointed out that
the narcissistic patients associations. They these patients start to develop new patterns of
may also note that with the hypervigilant va- perceiving relationships when they attend in
Psychoanalysis and Psychodynamic Psychotherapy 201

more detail to themselves and others in their these patients tend to use primitive defenses;
personal context. The analyst helps them see in addition, they have overwhelming separa-
how they play an active role in perpetuating tion anxiety and a lax superego. Some may
certain patterns of relating to others. They de- be able to tolerate the frustration of the ana-
velop a capacity to compare the actual facts lytic setting if they have high intelligence,
in an interpersonal situation with the inter- some degree of impulse control, and an abil-
nal patterns they superimpose on external ity to mentalize. However, many of these pa-
situations. tients have much in common with patients
Hysterical patients generally work well who have borderline personality disorder
within the transference and use it as a primary and lack the requisite ego strengths to toler-
vehicle for change. They may see the solution ate the process. If they are motivated to per-
to their problem as simple identification with severe in psychodynamic psychotherapy,
the analyst, and the analyst needs to confront much of what applies to the technique re-
the wish to bypass painful self-examination in quired with hysterical patients also applies
the service of identification with the analyst. to histrionic patients. However, they will re-
They also may fall in love with the analyst and quire more ego support and greater post-
feel that a loving relationship is in itself cura- ponement of interpretation to be effectively
tive. However, the transference love needs to treated.
be carefully deconstructed in terms of its mul-
tiple meanings. It functions not only as a resis- Obsessive-Compulsive
tance to a deepening of the analytic process
Personality Disorder
but also as an indirect expression of aggres-
sion, because it may create a frustrating situa- Patients with OCPD are generally well
tion for both analyst and patient. Hysterical suited for psychoanalysis or psychodynamic
patients often find themselves in triangular psychotherapy, but they must first overcome
relationships in which they constantly repeat their anxieties about being out of control.
an oedipal romance that places them in the Spontaneity and free association present
role of the excluded party. This pattern of them with a situation in which they must be
longing for an unavailable romantic partner open and drop their guard. Most obsessive-
deserves a good deal of scrutiny in the course compulsive individuals have a host of rigid
of the analysis, as it often reveals conflicts defenses to avoid spontaneity so they can
about truly separating from parents and avoid loss of control. Hence they will use re-
establishing a life with a partner outside the sistances that reflect defensive operations
family unit. Women with hysterical person- such as isolation of affect, intellectualization,
ality disorder are often Daddys girls, and reaction formation. The clinician may
whereas men with the diagnosis are often need to confront the tendency to use factual
called Mamas boys. This attachment to the information as a way of avoiding emotional
opposite-sex parent may keep them in a state expression in the analysis.
of extended adolescence that prevents them OCPD patients may also attempt to be
from coming fully into their own. The analyst perfect, saying exactly the right thing and
must systemically analyze their loyalty con- thereby pleasing the analyst or therapist.
flicts of this nature. This wish to please may be a reaction forma-
Those patients with histrionic personal- tion against hostility about the power differ-
ity disorder proper (i.e., the more dis