Concepts and Controversiesin Obsessive-Compulsive Disorder

SERIES IN ANXIETY AND RELATED DISORDERSSeries Editor: Martin M. Antony, Anxiety Treatment and Research Centre, St Joseph’s Hospital,Hamilton, Ontario, Canada

CONCEPTS AND CONTROVERSIESIN OBSESSIVE-COMPULSIVE DISORDEREdited by Jonathan S. Abramowitz and Arthur C. Houts

SOCIAL ANXIETY AND SOCIAL PHOBIA IN YOUTH:Characteristics, Assessment, and Psychological TreatmentChristopher A. Kearney

A continuation Order Plan is available for this series. A continuation order will bring delivery of eachnew volume immediately upon publication. Volumes are billed only upon actual shipment. For furtherinformation please contact the publisher.

Concepts and Controversiesin Obsessive-Compulsive

Disorder

Edited by

Jonathan S. AbramowitzMayo Clinic

Rochester, MN

Arthur C. HoutsThe West ClinicMemphis, TN

Jonathan S. Abramowitz Arthur C. HoutsThe Mayo Clinic The West ClinicDepartment of Psychiarty and Psychology 100 Humphreys Blvd. North200 First St. SW Memphis, TN 38120Rochester, MN 55905

Library of Congress Control Number: 2005927379

ISBN-10: 0-387-23280-X e-ISBN 0-387-23370-9 Printed on acid-free paper.ISBN-13: 978-0387-23280-X

C© 2005 Springer Science+Business Media, Inc.All rights reserved. This work may not be translated or copied in whole or in part without the writtenpermission of the publisher (Springer Science+Business Media, Inc., 233 Spring Street, New York,NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use inconnection with any form of information storage and retrieval, electronic adaptation, computer software,or by similar or dissimilar methodology now know or hereafter developed is forbidden.

The use in this publication of trade names, trademarks, service marks and similar terms, even if the arenot identified as such, is not to be taken as an expression of opinion as to whether or not they are subjectto proprietary rights.

Printed in the United States of America. (MVA/TB)

9 8 7 6 5 4 3 2 1

springeronline.com

To Stacy, Emily, and Miriam, and in loving memory ofMorris “Grandpop” Abramowitz—J.S.A.

To Mary and Margaret—A.C.H.

ABOUT THE EDITORS

Jonathan S. Abramowitz, PhD, ABPP received his PhD from the University ofMemphis in 1998 and is currently Associate Professor and Director of theOCD/Anxiety Disorders Program in the Department of Psychiatry and Psychologyat Mayo Clinic in Rochester, Minnesota. He is a member of the Obsessive Compul-sive Foundation Scientific Advisory Board and the Anxiety Disorders Associationof America Clinical Advisory Board. He also served on the Anxiety Disorders WorkGroup for the DSM-IV-TR. Dr Abramowitz conducts research on the psychopathol-ogy and treatment of OCD and other anxiety disorders and has published widelyon these topics. He has received numerous grants to support his research and wasawarded the Outstanding Contributions to Research Award in 2003 from the MayoClinic Department of Psychiatry and Psychology and the David Shakow Early CareerAward for Outstanding Contributions to Clinical Psychology in 2004 from Division12 of the American Psychological Association. Dr Abramowitz also maintains an ac-tive consultation and clinical practice involving the supervision and mentorship oftrainees in clinical psychology and psychiatry.

Arthur C. Houts, PhD received his PhD from Stony Brook University in 1981 andtaught at the University of Memphis from 1981 to 2003. He was Professor andDirector of Clinical Training. He retired in 2003 to work full time at The West Clinic, alarge community oncology center in Memphis, Tennessee. He is currently developinga model for delivering empirically supported behavior therapy for cancer patients inthe community setting. He is also developing technology to provide better quality oflife assessment in cancer care and building a research network of community oncologypractices for clinical trials. He has published broadly in adult and child clinical psy-chology as well as the interdisciplinary field of science studies. His interest in animalmodels of OCD goes back to his graduate training where he studied learning modelsof psychopathology with the early founders of behavior therapy, many of whom hadprimary interests in animal learning.

vii

CONTRIBUTORS

Jonathan S. Abramowitz, PhD, OCD/Anxiety Disorders Program, Departmentof Psychiatry and Psychology, Mayo Clinic, Rochester, Minnesota, USACheryl N. Carmin, PhD, Stress and Anxiety Disorders Center, Departmentof Psychiatry, University of Illinois at Chicago, Chicago, Illinois, USAMeredith E. Coles, PhD, Department of Psychology, Binghamton University,Binghamton, New York, USAMaria Conceicao do Rosario-Campos, MD, MSc, Yale University Schoolof Medicine, New Haven, Connecticut, USA and University of Sao Paulo MedicalSchool, Sao Paulo, BrazilBrett J. Deacon, PhD, Department of Psychology, University of Wyoming, Laramie,Wyoming, USANicholas H Dodman, BVMS, DACVB, Tufts University School of VeterinaryMedicine, Boston, Massachusetts, USARonald Faber, PhD, University of Minnesota, Minneapolis, Minnesota, USAJeanne Fama, MA, Department of Psychology, Harvard University, Charlestown,Massachusetts, USAAndrea Fougere, BA, Department of Psychiatry and Behavioral Neurosciences,Wayne State University School of Medicine, Detroit, Michigan, USAMartin E. Franklin, PhD, Center for the Treatment and Study of Anxiety,Department of Psychiatry, University of Pennsylvania School of Medicine,Philadelphia, Pennsylvania, USAJennifer P. Friedberg, MA, Ferkauf Graduate Psychology/Albert Einstein Collegeof Medicine, Yeshiva University, Bronx, New York, USAScott Hannan, PhD, Anxiety Disorders Center, The Institute of Living, Hartford,Connecticut, USASian Hemmings, MSc, Department of Psychiatry, University of Stellenbosch, CapeTown, South AfricaEric Hollander, MD, Compulsive, Impulsive, and Anxiety Disorders Program,Department of Psychiatry, Mount Sinai School of Medicine, New York, New York,USAArthur C. Houts, PhD, The West Clinic, Memphis, Tennessee, USA

ix

x CONTRIBUTORS

Rupa Iyengar, Compulsive, Impulsive, and Anxiety Disorders Program,Department of Psychiatry, Mount Sinai School of Medicine, New York, New York,USASony Khemlani-Patel, PhD, Biobehavioral Institute, Great Neck, New York, USACraig Kinnear, MSc, Department of Psychiatry, University of Stellenbosch, CapeTown, South AfricaMichael J. Kozak, PhD, National Institute of Mental Health, Bethesda, Maryland,USAJames F. Leckman, MD, Child Study Center, Yale University School of Medicine,New Haven, Connecticut, USAMichael R. Liebowitz, MD, New York State Psychiatric Institute and ColumbiaUniversity, New York, New York, USAChristine Lochner, MA, Department of Psychiatry, University of Stellenbosch, CapeTown, South AfricaDavid Mataix-Cols, PhD, Institute of Psychiatry, London, UKJames E. Mitchell, MD, Neuropsychiatric Research Institute, University of NorthDakota School of Medicine and Health Sciences, Fargo, North Dakota, USAFugen Neziroglu, PhD, Biobehavioral Institute, Great Neck, New York, USAKieron O’Connor, PhD, Fernand Seguin Research Center, Louis-H LafontaineHospital, University of Montreal, Montreal, CanadaDavid R. Rosenberg, MD, Department of Psychiatry and Behavioral Neurosciences,Wayne State University School of Medicine, Detroit, Michigan, USAAileen Russell, BA, Department of Psychiatry and Behavioral Neurosciences,Wayne State University School of Medicine, Detroit, Michigan, USAStefanie A. Schwartz, PhD, Department of Psychiatry and Psychology, Mayo Clinic,Rochester, Minnesota, USARoz Shafran, PhD, Department of Psychiatry, Oxford University, Oxford, UKLouis Shuster, PhD, Tufts University School of Veterinary Medicine, Boston,Massachusetts, USAH. Blair Simpson, MD, PhD, New York State Psychiatric Institute and ColumbiaUniversity, New York, New York, USAAnn Speckens, MD, Institute of Psychiatry, University of London, London, UKDan J. Stein, MD, PhD, Department of Psychiatry, University of Stellenbosch, CapeTown, South AfricaLorraine A. Swan-Kremeier, PsyD, Neuropsychiatric Research Institute, Universityof North Dakota School of Medicine and Health Sciences, Fargo, North Dakota,USASteven Taylor, PhD, Department of Psychiatry, University of British Columbia,Vancouver, British Columbia, CanadaDavid F. Tolin, PhD, Anxiety Disorders Center, The Institute of Living, Hartford,Connecticut, USA and University of Connecticut School of Medicine, Farmington,Connecticut, USA

CONTRIBUTORS xi

Stacey Wasserman, MD, Compulsive, Impulsive, and Anxiety Disorders Program,Department of Psychiatry, Mount Sinai School of Medicine, New York, New York,USAPamela S. Wiegartz, PhD, Stress and Anxiety Disorders Center, Department ofPsychiatry, University of Illinois at Chicago, Chicago, Illinois, USASabine Wilhelm, PhD, OCD Clinic, Massachusetts General Hospital-East andHarvard Medical School, Charlestown, Massachusetts, USAKevin D. Wu, PhD, Stress and Anxiety Disorders Center, Department of Psychiatry,University of Illinois at Chicago, Chicago, Illinois, USAChin-Chin Yeh, BA, Compulsive, Impulsive, and Anxiety Disorders Program,Department of Psychiatry, Mount Sinai School of Medicine, New York, New York,USA

PREFACE

Few syndromes in psychopathology generate as much popular curiosity and clinicalexploration as does obsessive-compulsive disorder (OCD). Since the 1970s, researchon OCD has increased exponentially. Specific advances include an improved grasp ofthe heterogeneity of the disorder, identification of putative subtyping schemes, andthe development of increasingly sophisticated theoretical models of the etiology andmaintenance. Perhaps most importantly, research has led to advances in treatment;and whereas the first line therapies (cognitive-behavior therapy and serotonergic med-ication) are not entirely effective for every sufferer, they have transformed OCD froman unmanageable lifetime affliction into a treatable problem that need not reducequality of life.

Despite the aforementioned advances, there have emerged a number of sharpdisagreements concerning OCD. Differences have surfaced over phenomenologicalissues, etiological models, and approaches to treatment, and often occur (but notexclusively) along disciplinary lines between biologically oriented and cognitive-behaviorally oriented authorities. For example, medical approaches posit that abnor-mal biological processes cause OCD, whereas psychosocial formulations emphasizethe role of learning and dysfunctional cognitions. Yet because theoretical conjectureand empirical findings from within each tradition are typically addressed toward dis-tinct and narrow audiences, clinicians, researchers, and students with broad interestsare hindered from gaining a clear grasp of the diverse (and sometimes polarized)perspectives.

In our view, scholarly debate and empirical scrutiny of divergent viewpointsis a healthy method by which our understanding of OCD can be enhanced. How dobiologically oriented researchers reconcile seemingly parsimonious accounts of obses-sions and compulsions that do not appeal to diseased neuroanatomic processes? Howdo cognitive-behaviorally oriented theorists deal with proposals for animal models ofOCD based on shared response to serotonin medication? Unfortunately, owing to therelative insularity of the different scientific communities that contribute to researchon OCD, such discussions rarely occur; at least not in published form. Therefore, ouraim for this edited book is to subject differing viewpoints on a variety of key con-ceptual, etiological, and therapeutic issues in the field of OCD to mutual debate. It isour hope that by bringing under one cover this vast literature, the volume will be aunique resource for clinicians, researchers, and students, regardless of theoretical andprofessional allegiances.

xiii

xiv PREFACE

We have chosen to focus on seven topics that represent sources of disagreementamong OCD experts. These topics are organized into three sections. The first section,Phenomenology, covers the issues of OCD symptom subtypes, animal models, and ex-plores the possibility of a spectrum of obsessive-compulsive disorders. In the secondsection, Etiology, neuropsychiatric and cognitive-behavioral models of OCD are eachpresented and critiqued. The third section, Treatment, includes chapters on the use ofcognitive therapy versus exposure and response prevention, the importance of ther-apist involvement in exposure-based treatment, and whether combining medicationand cognitive-behavioral treatment is preferable to monotherapy.

The book was composed as follows: First, noted authorities were asked to producean empirically grounded position paper on their particular area of expertise. Afterreceiving both chapters on a particular issue, the manuscripts were given to the authorof the opposing viewpoint for a brief response. We purposely limited the scope of ourediting in order to uphold the authors’ intended points; and authors were not allowedto amend their original manuscripts.

We are very pleased with this book and what it represents. The world’s expertson the nature and treatment of OCD debate and provide a clear and firm statementof their positions. Readers will become aware of the finer points of many argumentsand enjoy a revealing look into each author’s reaction to an opposing point of view. Insome cases, the authors have discovered overlaps in ostensibly diverse positions; withsome opening a critical eye toward their own point of view. In other cases, authorshave used their rejoinder as an opportunity to reiterate differences and further refinetheir own arguments.

We are fortunate to be standing at the dawn of a new century when we can lookback and forward with hope. In looking back, we can hope that the days are gone forgood when individuals with OCD endured years of psychoanalytic treatment withlittle improvement and provided endless intellectual fodder for speculative theoriesthat were ill-formed and unhinged from both behavior and physiology. Today, wecan see that there are treatments that are useful and oftentimes highly effective. Wecan also see that there is energetic disagreement among experts in the OCD researchcommunities, and it is the sort of disagreement that can lead to productive competition,fruitful debate, and more refined care of patients. In looking forward, we can hopethat investigators with differing backgrounds and traditions of research will stayengaged with one another as they pursue their own lights so that at the close of thenext century we will have an understanding of OCD that integrates the best methodsof neurochemistry and neurophysiology with the best methods of behavioral science.To be able to explain in a causal way the development and the maintenance andextinction of the vexing and self defeating behaviors of OCD remains the goal, andthat is a prize worthy of many a lifetime of work.

Jonathan S. Abramowitz, Rochester, Minnesota, USAArthur C. Houts, Memphis, Tennessee, USA

CONTENTS

PART I. PHENOMENOLOGY 1

1. Symptom Dimensions in OCD: Developmental andEvolutionary Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3James F. Leckman, David Mataix-Cols, and Maria Conceicao doRosario-Campos

2. Dimensional and Subtype Models of OCD . . . . . . . . . . . . . . . . 27Steven Taylor

Reply to Taylor. Combined Dimensional and CategoricalPerspectives as an Integrative Approach to OCD . . . . . . . . . . . . 43James F. Leckman, David Mataix-Cols, and Maria Conceicao doRosario-Campos

Reply to Leckman et al. Putting the Symptom Dimension Modelto the Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49Steven Taylor

3. Animal Models of Obsessive Compulsive Disorder:A Neurobiological and Ethological Perspective . . . . . . . . . . . . . 53Nicholas H. Dodman and Louis Shuster

4. Behavioral and Functional Models of OCD . . . . . . . . . . . . . . . . 73Arthur C. Houts

Reply to Houts. A Dysfunctional Animal Model of OCD . . . . . . 87Nicholas H. Dodman

Reply to Dodman and Shuster. Animal Models and TwoTraditions in OCD Research . . . . . . . . . . . . . . . . . . . . . . . . . . . 91Arthur C. Houts

5. The Case for the OCD Spectrum . . . . . . . . . . . . . . . . . . . . . . . . 95Eric Hollander, Jennifer P. Friedberg, Stacey Wasserman, Chin-Chin Yeh,and Rupa Iyengar

6. Obsessive-Compulsive Disorder: Essential Phenomenology andOverlap with Other Anxiety Disorders . . . . . . . . . . . . . . . . . . . 119Jonathan S. Abramowitz and Brett J. Deacon

xv

xvi CONTENTS

Reply to Abramowitz and Deacon. Beyond Anxiety: Etiologicaland Functional Overlaps Between OCD and OC SpectrumDisorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137Eric Hollander and Chin-Chin Yeh

Reply to Hollander et al. The OC Spectrum: A Closer Look at theArguments and the Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141Jonathan S. Abramowitz and Brett J. Deacon

7. Trichotillomania: An Obsessive-Compulsive Spectrum Disorder 151Dan J. Stein, Christine Lochner, Sian Hemmings, and Craig Kinnear

8. Overlap of Body Dysmorphic Disorder and Hypochondriasiswith OCD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163Fugen Neziroglu and SonyKhemlani-Patel

9. Contrasting Nonparaphilic Sexual Addictions and OCD . . . . . . 177Stefanie A. Schwartz and Jonathan S. Abramowitz

10. Compulsive Buying: A Disorder of Compulsivity or Impulsivity 185Lorraine A. Swan-Kremeir, James E. Mitchell, and Ronald J. Faber

11. Contrasting Tourette’s Syndrome and Tic Disorders with OCD . 191Kieron O’Connor

PART II. ETIOLOGY 207

12. Neuropsychiatric Models of OCD . . . . . . . . . . . . . . . . . . . . . . . 209David R. Rosenberg, Aileen Russell, and Andrea Fougere

13. Cognitive-Behavioral Models of OCD . . . . . . . . . . . . . . . . . . . . 229Roz Shafran

Reply to Shafran. Biological and Cognitive Models of OCD:Seeking Similarities and Achieving Progress Together . . . . . . . . 253David R. Rosenberg, Aileen Russell, and Andrea Fougere

Reply to Rosenberg et al. Biological Versus PsychologicalApproaches to OCD: War or Peace? . . . . . . . . . . . . . . . . . . . . . . 255Roz Shafran and Anne Speckens

PART III. TREATMENT 261

14. Formal Cognitive Therapy: A New Treatment for OCD . . . . . . . 263Jeanne Fama and Sabine Wilhelm

15. Treatment for OCD: Unleashing the Power of Exposure . . . . . . . 283Michael J. Kozak and Meredith E. Coles

Reply to Kozak and Coles. Expanding the Conceptualization ofCognitive Therapy and its Therapeutic Potential . . . . . . . . . . . . 305Jeanne Fama and Sabine Wilhelm

CONTENTS xvii

Reply to Fama and Wilhelm. Cognitive Therapy and ExposureTreatment for OCD: Contrast and Rapprochement . . . . . . . . . . . 311Michael J. Kozak and Meredith E. Coles

16. The Role of the Therapist in Behavior Therapy for OCD . . . . . . 317David F. Tolin and Scott Hannan

17. Self-Directed Exposure in the Treatment of OCD . . . . . . . . . . . . 333Cheryl N. Carmin, Pamela S. Wiegartz, and Kevin D. Wu

Reply to Carmin et al. What’s in a Name? The DistinctionBetween Self-Directed and Self-Conducted Treatment . . . . . . . . 347David F. Tolin and Scott Hannan

Reply to Tolin and Hannan. Self-Directed Versus Therapist-Directed Treatment: Additional Considerations . . . . . . . . . . . . . 353Cheryl N. Carmin, Pamela S. Wiegartz, and Kevin D. Wu

18. Combining Pharmacotherapy and Cognitive-BehavioralTherapy in the Treatment of OCD . . . . . . . . . . . . . . . . . . . . . . . 359H. Blair Simpson and Michael R. Liebowitz

19. Combining Serotonin Medication with Cognitive-BehaviorTherapy: Is it Necessary for all OCD Patients? . . . . . . . . . . . . . . 377Martin E. Franklin

Reply to Franklin. Using Combination Treatments for OCD . . . . 391H. Blair Simpson and Michael R. Liebowitz

Reply to Simpson and Liebowitz. Meeting in the Middle, thenMoving Forward Together . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395Martin E. Franklin

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401

Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423

Part I

PHENOMENOLOGY

Chapter 1

SYMPTOM DIMENSIONS IN OCD:DEVELOPMENTAL AND

EVOLUTIONARY PERSPECTIVES

James F. Leckman, David Mataix-Cols, andMaria Conceicao do Rosario-Campos

The extraordinary intricacy of all the factors to be taken into accountleaves only one way to presenting them open to us. We must first selectone and then another point of view, and follow it up through the materialas long as the application of it seems to yield results.

—Sigmund Freud, 1915The idea of a disease-entity is not an objective to be reached, but our mostfruitful point of orientation.

—Karl Jaspers, 1923

At the present time, in the absence of definitive etiological markers of vulnerability forobsessive-compulsive disorder (OCD), obsessive-compulsive (OC) symptom dimen-sions appear to offer a fruitful point of orientation. The complex clinical presentationof OCD can be summarized using a few consistent and temporally stable symptomdimensions. These can be understood as a spectrum of potentially overlapping vul-nerabilities that are likely to be continuous with “normal” worries and extend beyondthe traditional nosological boundaries of OCD. Although the dimensional structureof OC symptoms is still imperfect, this quantitative approach to phenotypic traits hasthe potential to advance our understanding of OCD and may aid in the identificationof more robust endophenotypes. Preliminary data suggest that these dimensionalphenotypes may be useful in studies of the natural history, genetics, neurobiology,and treatment outcome of OCD. A dimensional approach also appears to be con-gruent with evolutionary and developmental perspectives on OCD. This view pointposits that each symptom dimension reflects the dysregulation of highly conservedcomplex and partially overlapping neural systems that serve to detect, appraise, andrespond to potential threats. A dimensional approach is also not mutually exclusiveof other methods to parse the larger spectrum of disorders related to OCD. Indeed, thecombined use of categorical subtypes and dimensional assessments are likely to offer

3

4 LECKMAN ET AL.

the greatest promise. Thus far, age-of-onset of OC symptoms and the individual’s“tic-related” status appear to be particularly useful categorical distinctions. Finally,existing assessment methods are inadequate and new dimensional scales are neededto take full advantage of a dimensional approach in clinical and population-basedstudies.

INTRODUCTION

Obsessive-compulsive disorder is a chronic and potentially disabling conditionaffecting from 1% to 3% of the general population. Patients with OCD describe thesudden intrusion into consciousness of unwanted thoughts or unpleasant images. Fre-quently these obsessions are accompanied by a profound sense of dread and the urge tocomplete specific compulsions. Compulsions are repetitive acts, typically performeda certain number of times or according to certain private rules, that the individual isdriven to complete, even though these acts are perceived as excessive.

The DSM-IV (American Psychiatric Association [APA], 1994) and other standarddiagnostic classifications such as ICD-10 (World Health Organization [WHO], 1992)regard OCD as a unitary nosological entity. While this parsimony has a certain formalappeal, it is misleading. The symptoms used to define OCD are heterogeneous andinclude various intrusive thoughts and preoccupations, rituals, and compulsions.Two individuals with OCD may have totally different and non-overlapping symptompatterns.

From as far back as the earliest descriptions of OCD, investigators have at-tempted to dissect the phenotype into homogeneous subtypes. For example, Falretmade the distinction between “Folie du doute” (madness of doubt) and “Delire dutoucher” (delusion of touch) in 1869 (Hantouche & Lancrenon, 1996). Most com-monly investigators have distinguished “washers” from “checkers” (Horesh, Dolberg,Kirschenbaum-Aviner, & Kotler, 1997; Khanna & Mukherjee, 1992; Matsunaga et al.,2001; Rachman & Hodgson, 1980). With a few notable exceptions, these attempts hadlimited success in relating the identified subtypes to biological markers, genetic fac-tors or treatment response in part because pure subtypes of patients are rare, andthe recruitment of sufficient sample sizes of each subtype is difficult and highlyimpractical.

The following review considers an alternative approach to OC symptoms thataims to identify valid quantitative dimensions for use in genetic, neurobiological, andtreatment outcome studies. The review then proceeds to examine the potential valueof a dimensional approach from developmental and evolutionary perspectives. Thechapter closes with a call for the development of state-of-the-art assessment methods.

CHALLENGES TO THE CONCEPT OF OCD ASA UNITARY DIAGNOSTIC ENTITY

In addition to the clinical diversity seen in OCD, genetic and treatment studiesalso support the view that OCD is a heterogeneous disorder. The influence of geneticfactors has been suggested from the earliest descriptive accounts of OCD. Data from

SYMPTOM DIMENSIONS 5

twin and family aggregation studies provide limited support for the vertical trans-mission of genetic vulnerability factors within some families (Pauls & Alsobrook,1999). However, inspection of the available pedigrees suggests that OCD is likelygenetically complex and heterogeneous. Some cases are familial and related to ticdisorders, some cases are familial and unrelated to tics, while in other cases there isno family history of either OCD or tics. Recent segregation analyses also support theview that OCD is genetically heterogeneous (Cavallini, Pasquale, Bellodi, & Smeraldi,1999).

With regard to treatment, evidence from the past 20 years indicates that OCDis preferentially responsive to pharmacotherapy with potent serotonin reuptake in-hibitors (SRIs) and to specific forms of cognitive-behavior therapy. Despite these ad-vances, a substantial number of patients remain symptomatic or show no improve-ment with these treatments. Between 40% and 60% of patients are nonresponders toSRIs (Greist, Jefferson, Kobak, Katzelnick, & Serlin, 1995). Even among responders toSRIs, the magnitude of response is variable with few patients becoming asymptomatic.Similarly, while cognitive-behavioral treatment with exposure and response preven-tion frequently leads to significant and lasting improvement in OCD symptoms, atleast 25% of patients fail to respond and many more patients refuse to participate (Foa& Kozak, 1996).

The factors associated with an unfavorable treatment response remain largelyunknown. For example, pretreatment symptom severity appears not to be a use-ful predictor of response (Ackerman, Greenland, Bystritsky, Morgenstern, & Katz,1994; Steketee, 1993). Preliminary evidence, however, indicates that patients with acomorbid personality disorder appear to be less responsive to SRIs as well as toexposure and response prevention (Cavedini, Erzegovesi, Ronchi, & Bellodi, 1997;Ravizza, Barzega, Bellino, Bogetto, & Maina, 1995). Likewise, comorbid tic disor-ders are also associated with a poor response to SRI monotherapy (McDougle et al.,1994).

While subtyping OCD cases based on the presence or absence of tics or some otherpatient characteristic, such as age of onset, may lead to increased biological homo-geneity, other quantitative approaches may prove to be of greater value in identifyingthe relevant genetic vulnerability factors. Such quantitative approaches might also aidin the identification of treatment responders. Ideally, such a quantitative vulnerabil-ity factor would be readily measurable and could serve as an “endophenotype,” onethat is functionally intermediate between a specific vulnerability gene or pathway fortreatment response and specific phenotypic features.

In the absence of such neuropsychological, neurophysiological, or neurochem-ical measures, another potentially valuable approach concerns the identification ofcomponent aspects of the clinical phenotype itself. A similar approach has provenuseful in the study of dyslexia in which component features of this learning disabil-ity appear to be associated with specific genetic loci that segregate independentlyof one another (Grigorenko, Wood, Meyer, & Pauls, 2000). If OCD is like dyslexia,then it may be useful to consider the possibility that the symptoms of OCD can bedecomposed into several dimensions that are themselves continuous within the pop-ulation. If true, the use of these quantitative phenotypes may provide superior powerand efficiency of parameter estimation within linkage analyses as well as potentiallyimportant predictors of treatment response.

6 LECKMAN ET AL.

A WORD ABOUT AVAILABLE RATINGINSTRUMENTS

In this review, we have only included studies that used comprehensive and non-biased instruments to ascertain OC symptoms, such as the Yale-Brown Obsessive-Compulsive Scale Symptom Checklist (Y-BOCS-SC; Goodman et al., 1989) and theObsessive-Compulsive Inventory (OCI; Foa et al., 2002). The Y-BOCS-SC lists morethan 50 examples of obsessions and compulsions, organized under 13 major categoriesplus two categories of miscellaneous obsessions and compulsions, covering the vastmajority of OC symptoms (Goodman et al., 1989). Other frequently used instrumentssuch as the Maudsley Obsessive-Compulsive Inventory (Rachman & Hodgson, 1977)and the Padua Inventory (Sanavio & Vidotto, 1988) were excluded because their itemsare heavily biased toward specific symptoms (eg, checking, cleaning) or that omit keysymptoms (eg, hoarding, symmetry). More problematic is the use of composite sever-ity ratings based on all of the patient’s obsessions and compulsions (Fals-Stewart, 1992;Kim, Dysken, Pheley, & Hoover, 1992; McKay, Danyko, Neziroglu, & Yaryura-Tobias,1995). We also take note of the limitations of the Y-BOCS-SC, particularly the fact thatsome of the symptom categories are too broadly conceived, and as such are heteroge-neous and inherently ambiguous. For example, any given checking compulsion couldbe related to any one of a number of symptom dimensions, yet any checking-relatedcompulsion can only be rated within a single symptom domain. Similar critiques canonly be made for the wide range of mental rituals and avoidance-related behaviors.

QUANTITATIVE OCD PHENOTYPES:INITIAL STUDIES

The first study to factor-analyze the Y-BOCS-SC was that of Baer (1994). Hefactor-analyzed the 13 major categories of the Y-BOCS-SC in a sample of 107 patientsand identified three factors, accounting for 48% of the variance; these were named“symmetry/hoarding,” “contamination/cleaning,” and “pure obsessions.” Follow-ing Baer’s seminal work, Leckman et al. (1997) evaluated the 13 a priori Y-BOCS-SCcategories in two large groups of OCD patients totaling over 300 cases (Leckman,Walker, Goodman, Pauls, & Cohen, 1994; Pauls, Alsobrook, Goodman, Rasmussen, &Leckman, 1995). A principal components factor analysis was performed with a Vari-max rotation separately using the symptom information from each group of OCDpatients. In an effort to identify valid “traits,” they included any OCD symptoms thatpatients “ever” experienced over the course of their lifetimes, as opposed to limitingthese analyses to current symptoms. Remarkably, both data sets yielded nearly identi-cal results. Four factors were identified that in total accounted for more than 60% of thevariance in each data set (Leckman et al., 1997). The first factor included obsessionsabout aggressive behavior toward self or others, sexual obsessions and obsessionsrelated to moral rightness or religion, as well as related checking compulsions. Thisfactor accounted for 30.1% of the variance. A second factor accounted for 13.8 %of the variance and included obsessions concerning a need for symmetry or exact-ness, repeating rituals, counting compulsions, and ordering/arranging compulsions.A third factor that accounted for 10.2% of the variance was composed of contamina-tion obsessions and cleaning and washing compulsions. The fourth factor included

SYMPTOM DIMENSIONS 7

hoarding and collecting obsessions and compulsions and accounted for 8.5% of thevariance.

Subsequently, Summerfeldt, Richter, Antony, and Swinson (1999) evaluated exist-ing models of OCD symptom structure in a sample of 203 individuals. Using confirma-tory factor analyses, they examined four models: a single-factor (ie, OCD as a single di-mension), a two-factor (ie, obsessions and compulsions), the three-factor model of Baer(1994), and our four-factor model. Adequate fit was found solely for the four-factormodel described in Leckman et al. (1997), specifying aggressive, sexual, and religiousobsessions and checking compulsions; symmetry/ordering; contamination/cleaning;and hoarding. However, parameter estimates showed within-factor heterogeneity, aswell as overlap between factors, most notably between factors 1 and 3.

At present, there have been at least 11 studies corresponding to 10 large OCDdatasets and involving over 2000 patients (Table 1.1; Mataix-Cols, Fullana, Alonso,Menchon, & Vallejo, 2004). While the factorial studies available to date have been fairlyconsistent, the number of factors has ranged from 3 to 6. Some of the symptom dimen-sions have been consistently replicated across studies (eg, contamination/washing,symmetry/ordering, hoarding) but the aggressive/checking and sexual/religious di-mensions need further study, as it is unclear whether they form a unique factor (Cav-allini, Di Bella, Siliprandi, Malchiodi, & Bellodi, 2002; Leckman et al., 1997, 2003;Summerfeldt et al., 1999), or can be broken down into two separate dimensions(Baer, 1994; Foa et al., 2002; Hantouche & Lancrenon, 1996; Mataix-Cols, Marks, Greist,Kobak, & Baer, 2002a; Mataix-Cols, Rauch, Manzo, Jenike, & Baer, 1999; Summerfeldt,Richter, Anthony, & Swinson, 1997; Tek & Ulug, 2001). Similarly, it is unclear how toregard somatic obsessions, as they loaded on the contamination/washing factor intwo studies (Baer, 1994; Hantouche & Lancrenon, 1996), on the obsessions/checkingfactor in three other studies (Cavallini et al., 2001; Leckman et al., 1997, 2003), and justwith sexual obsessions in two other studies (Feinstein, Fallon, Petkova, & Liebowitz,2003; Mataix-Cols et al., 2002a).

TEMPORAL STABILITY OF OCD SYMPTOMDIMENSIONS

Preliminary data are available that support the temporal stability of the OCsymptom dimensions, at least in adult patients. Rettew, Swedo, Leonard, Lenane, andRapoport (1992) assessed the longitudinal course of OC symptoms in 76 children andadolescents with OCD followed over a period of 2–7 years, using the categories of theY-BOCS-SC. They found that none of the patients maintained the same constellationof symptoms from baseline to follow-up. Nevertheless, these authors acknowledgedthat these changes could have occurred within (rather than between) symptom di-mensions; and this hypothesis was not specifically addressed. In the Leckman et al.(1997) report, two sets of OCD patients were studied. Fourteen subjects participatedin both studies. The mean (SD) test–retest interval was 51.2 (11.7) months with a rangeof 17–61 months. In each case, the family-genetic study data were collected prior to thedata for the phenomenological studies. Pearson correlations were in the moderate tostrong range for each symptom dimension, ranging from .51 to .75. These preliminarytest–retest findings indicate that these symptom dimensions may be relatively stableindividual characteristics over time. In a more recent study (Mataix-Cols et al., 2002b),

TAB

LE

1.1.

The

stru

ctur

eof

obse

ssio

nsan

dco

mpu

lsio

nsas

repo

rted

in11

avai

labl

est

udie

s

Con

tam

inat

ion/

Agg

ress

ive/

Sexu

al/

Sym

met

ry/

%E

xpla

ined

was

hing

chec

king

relig

ious

Hoa

rdin

gor

der

ing

vari

ance

Stud

yB

aer

(199

4)X

2X

3X

3X

1X

148

Inst

rum

ent

Y-B

OC

S-SC

Ana

lysi

ste

chni

que

PCA

,cur

rent

sym

ptom

sN

107

OC

D

Stud

yH

anto

uche

and

Lan

cren

on(1

996)

X3

X2

X2

X1

X1

Inst

rum

ent

Y-B

OC

S-SC

Ana

lysi

ste

chni

que

PCA

,cur

rent

sym

ptom

sN

615

OC

D

Stud

yL

eckm

anet

al.(

1997

)X

3X

1X

1X

4X

262

.6In

stru

men

tY-

BO

CS-

SCA

naly

sis

tech

niqu

ePC

A,l

ifet

ime

sym

ptom

sN

292

OC

D

Stud

yL

eckm

anet

al.(

1997

)X

3X

1X

1X

4X

2In

stru

men

tY-

BO

CS-

SCA

naly

sis

tech

niqu

ePC

A,l

ifet

ime

sym

ptom

sN

292

OC

D

Stud

ySu

mm

erfe

ldte

tal.

(199

9)X

3X

1X

1X

4X

2In

stru

men

tY-

BO

CS-

SCA

naly

sis

tech

niqu

eC

FA,c

urre

ntsy

mpt

oms

N20

3O

CD

Stud

yM

atai

x-C

ols

etal

.(19

99)

X3

X4

X5

X2

X1

65.6

Inst

rum

ent

Y-B

OC

S-SC

Ana

lysi

ste

chni

que

PCA

,cur

rent

sym

ptom

sN

354

OC

D

Stud

yTe

kan

dU

lug

(200

1)X

1X

2X

4X

2X

1,2,

365

.6In

stru

men

tY-

BO

CS-

SCA

naly

sis

tech

niqu

ePC

A,c

urre

ntsy

mpt

oms

N45

OC

D

Stud

yC

aval

linie

tal.

(200

1)X

1X

3X

3X

2X

4,5

58.8

Inst

rum

ent

Y-B

OC

S-SC

Ana

lysi

ste

chni

que

PCA

,lif

etim

esy

mpt

oms

N18

0O

CD

Stud

yM

atai

x-C

ols

etal

.(20

02a)

X2

X1

X1

X4

X1,

363

.5In

stru

men

tY-

BO

CS-

SCA

naly

sis

tech

niqu

ePC

A,c

urre

ntsy

mpt

oms

N15

3O

CD

Stud

yL

eckm

anet

al.(

2003

)X

3X

1X

1X

4X

2,3

Inst

rum

ent

Y-B

OC

S-SC

73.1

Ana

lysi

ste

chni

que

PCA

,lif

etim

esy

mpt

oms

N23

6T

S+

FDFM

Stud

yFo

aet

al.(

2002

)X

1X

2X

2X

3X

4,5,

680

.8In

stru

men

tO

CI

Ana

lysi

ste

chni

que

PCA

,CFA

,cur

rent

sym

ptom

sN

215

OC

D

Stud

yFe

inst

ein

etal

.(20

03)

X2

X2

X4

X3

X1

54.2

Inst

rum

ent

Y-B

OC

S-SC

Ana

lysi

ste

chni

que

PCA

,cur

rent

sym

ptom

sat

item

leve

l,ex

clud

ing

mis

cella

neou

s16

0O

CD

N Not

e:T

S=

Tour

ette

’sSy

ndro

me;

FDFM

=fir

std

egre

efa

mily

mem

bers

;Y-B

OC

S-SC

=Ya

le-B

row

nob

sess

ive

com

puls

ive

scal

esy

mpt

omch

eckl

ist;

OC

I=ob

sess

ive

com

puls

ive

inve

ntor

y;PC

A=

prin

cipa

lcom

pone

nts

anal

ysis

;CFA

=co

nfirm

ator

yfa

ctor

anal

ysis

.Num

eric

subs

crip

tsre

fer

tose

para

tefa

ctor

sin

des

cend

ing

ord

erba

sed

onth

eam

ount

ofva

rian

ceex

plai

ned

.Sy

mpt

omca

tego

ries

that

shar

eth

esa

me

num

eric

subs

crip

tw

here

iden

tifie

das

asi

ngle

dim

ensi

onin

the

stud

y.C

ateg

orie

sw

ith

mul

tipl

esu

bscr

ipts

wer

eid

enti

fied

asco

ntai

ning

mor

eth

anon

ed

imen

sion

.

10 LECKMAN ET AL.

a large sample of adult patients were repeatedly administered the Y-BOCS-SC over aperiod of 2 years. For the most part, patients maintained their symptoms across thisinterval and the most robust predictor of having a particular symptom was havinghad that symptom in the past. For those symptoms that changed across time, changestypically occurred within rather than between previously identified symptom dimen-sions, suggesting that the symptoms of adult OCD patients are more stable than itis often assumed. Longitudinal studies following patients from childhood to adult-hood are needed to gain a more complete understanding of the natural history of OCsymptoms.

HOW USEFUL ARE OC SYMPTOMDIMENSIONS IN SORTING OUT COMORBID

DISORDERS AND OC SPECTRUMDISORDERS?

Obsessive-compulsive disorder shares a relatively high comorbidity with a num-ber of other psychiatric conditions, including other anxiety disorders, panic disorder,specific phobias, affective disorders, and tic disorders. Separation anxiety, while anindex of healthy attachment in typically developing infants and toddlers, is muchmore common in children and adolescents with OCD—at a point in developmentwhen significant separation anxiety is indicative of disorder rather than positive ad-justment (Carter, Pauls, & Leckman, 1995). Some studies suggest that as much as 65%of individuals with OCD also meet diagnostic criteria for depression (Pauls, Leck-man, & Cohen, 1994). Obsessive-compulsive symptomatology is also typically part ofthe clinical presentation of individuals with Tourette syndrome (TS) or other chronictic disorders (Pauls & Leckman, 1986; Pauls, Raymond, Stevenson, & Leckman, 1991).

In recent years, there has been a growing popularity of the concept of a spectrumof disorders related to OCD (Hollander, 1993; Rasmussen, 1994). While there has beensome debate about the breadth of membership and the criteria used for inclusion,certain disorders including chronic tic disorders, body dysmorphic disorder, eatingdisorders, and trichotillomania are routinely considered to be part of this spectrum(Jenike & Wilhelm, 1998).

How useful are OC symptom dimensions in sorting out comorbid disorders andOC spectrum disorders? Baer (1994) reported that patients with high scores on hissymmetry/hoarding factor were more likely to have a comorbid diagnosis of chronictics and OC personality disorder. Similarly, Leckman et al. (1997) found that patientswith high scores on the obsessions/checking and symmetry/ordering factors weremore likely to present with tics. Mataix-Cols et al. (1999) found that male but notfemale OCD patients with chronic tics scored higher than patients without tics on thesymmetry/ordering dimension. There also appears to be a clear overlap with someeating disorders. For example, Halmi et al. (2003) reported that approximately 70% ofpatients with anorexia nervosa had lifetime OC symptoms, especially symmetry andsomatic obsessions and ordering and hoarding compulsions.

Mataix-Cols, Baer, Rauch, and Jenike (2000) examined the presence of all DSM-III-R Axis II diagnoses and their relation to OC symptom dimensions in a sample of75 OCD patients. They found that hoarding symptoms were strongly related to thepresence and number of all personality disorders, especially from the anxious-fearful

SYMPTOM DIMENSIONS 11

cluster. Similarly, Frost, Steketee, Williams, and Warren (2000) found that hoarding wasassociated with higher levels of comorbidity (ie, anxiety, depression, and personalitydisorders), as well as work and social disability, compared to nonhoarding OCDand other anxiety disorders. In another study (Samuels et al., 2002), the presenceof hoarding was associated with male gender, earlier age of onset, comorbid socialphobia, personality disorders, and pathological grooming conditions (skin picking,nail biting, and trichotillomania). While Samuels et al. (2002) found that hoarding wasassociated with greater overall illness severity (total Y-BOCS scores) another studydid not (Saxena et al., 2002).

Taken together, these studies suggest that a symptom-based dimensional ap-proach may be useful in efforts to integrate previous classification attempts based onage of onset, gender, or presence of comorbid and OC spectrum conditions. A dimen-sional approach has the advantage of allowing each patient to have scores in one ormore symptom dimension. It also permits studies that cut across traditional diagnos-tic boundaries. This approach may be particularly valuable for genetic studies whereit increasingly appears that some vulnerability genes may be shared by more than asingle disorder and subthreshold cases are likely to be found in family members.

INITIAL VALIDATION FROM FAMILYGENETIC STUDIES

Like many other psychiatric disorders, twin and family studies suggest that ge-netic factors play a role in the expression of OCD (Alsobrook & Pauls, 1998). Recentadvances in molecular genetics have greatly increased the capacity to localize dis-ease genes on the human genome. These methods are now being applied to complexdisorders, including OCD. Although earlier studies have indicated that the verticaltransmission of OCD in families is consistent with the effects of a single major auto-somal gene (Cavallini et al., 1999; Nestadt et al., 2000; Nicolini, Kuthy, Hernandez, &Velazquez, 1991), it is more likely that there are a number of vulnerability genes in-volved. One of the major difficulties in the application of these approaches is the likelyetiologic heterogeneity of OCD and related phenotypes. Heterogeneity reduces thepower of gene-localization methods, such as linkage analysis (Alcais & Abel, 1999; Gu,Province, Todorov, & Rao, 1998; Zhang & Risch, 1996). Etiologic heterogeneity maybe reflected in phenotypic variability, thus it would be highly desirable to dissect thesyndrome, at the level of the phenotype, into valid quantitative heritable components.

Alsobrook, Leckman, Goodman, Rasmussen, and Pauls (1999) were the first touse OC symptom dimensions in a genetic study. They found that the relatives of OCDprobands who had high scores on the obsessions/checking and symmetry/orderingfactors were at greater risk for OCD than were relatives of probands who had lowscores on those factors.

Subsequently, using data collected by the Tourette Syndrome Association Inter-national Consortium for Genetics (TSAICG) Affected Sibling Pair Study, Leckmanet al. (2003) selected all available affected TS pairs and their parents for which theseOC symptom dimensions (factor scores) could be generated using the four factor al-gorithm first presented by Leckman et al. (1997). Remarkably, over 50% of the siblingswith TS were found to have comorbid OCD and greater than 30% of mothers and 10%of fathers also had a diagnosis of OCD. The factor scores for aggressive, sexual, and

12 LECKMAN ET AL.

religious obsessions and checking compulsions and symmetry and ordering obses-sions and compulsions scores were significantly correlated in sibling pairs concordantfor TS. In addition, the mother–child correlations, but not father–child correlations,were significant for these two factors. On the basis of the results of the complex seg-regation analyses, significant evidence for genetic transmission was obtained for allfactors. More recently, a genome scan of the hoarding dimension was completed usingthe same TSAICG data set (Zhang, Leckman, Tsai, Kidd, & Rosario-Campos, 2002). Theanalyses were conducted for hoarding as both a dichotomous trait and a quantitativetrait. Not all sibling pairs in the sample were concordant for hoarding. Standard link-age analyses were performed using GENEHUNTER and Haseman-Elston methods.In addition, novel analyses with a recursive-partitioning technique were employed.Significant allele sharing was observed for both the dichotomous and the quantita-tive hoarding phenotypes for markers at 4q34, 5q35.2, and 17q25. The 4q site is inproximity to D4S1625, which was identified by the TSAICG as a region linked to theTS phenotype. A recursive-partitioning analytic technique also examined multiplemarkers simultaneously. Results suggest joint effects of specific loci on 5q and 4q.

In summary, the use of quantitative traits that are familial may provide a power-ful approach to detect the genetic susceptibility loci that contribute to OCD presen-tations. Thus far, this approach has provided especially promising leads with regardto the hoarding OC phenotype. Next steps include, first, the use of these symptomdimensions in large multigenerational families in order to refine the initial geneticlinkage results for the hoarding phenotype. Obviously, if specific loci are identified,this will provide compelling evidence for the validity of this multidimensional ap-proach to OCD. Second, genome scans also need to be conducted using the remainingOC symptom dimensions. Families segregating for TS or early onset OCD may beespecially valuable in this enterprise. Given the high mother–child correlations in theLeckman et al. (2003) study, it may also be valuable to examine the linkage results foralleles that are identical by descent from the mother. Third, twin and cross-fosteringstudies are needed to further evaluate the heritability of these symptom dimensionswithin the general population. Finally, future genetic studies will also need to exam-ine the relationship between these dimensions and other closely related phenotypesincluding various eating disorders (Halmi et al., 2003) and body dysmorphic disorder.

HINTS AT VALIDATION FROM IN VIVONEUROIMAGING STUDIES

Functional neuroimaging studies have the potential to increase our understand-ing of the neural mechanisms underlying OCD. Taken as a whole, these studiesstrongly link OC symptoms with activation of the orbitofrontal cortex, with less con-sistent involvement of anterior cingulate gyrus, lateral frontal and temporal cortices,caudate nucleus, thalamus, amygdala, and insula (Saxena & Rauch, 2000). We wouldpredict that if a dimensional approach is useful, then a significant portion of the in-dividual variation seen in these studies might be accounted for by the unique mix ofsymptom dimensions seen in any given patient. Initial studies generally support thisconclusion. Rauch et al. (1998) completed a study with seven adult OCD subjects whoreceived 15O water during a classic continuous performance task as part of a positronemission tomography study. They found that scores on the obsessions/checking

SYMPTOM DIMENSIONS 13

dimension correlated with increased blood flow to the striatum during the contin-uous performance task.

Subsequently, Phillips et al. (2000) studied seven OCD subjects with prominentcontamination preoccupations and cleaning compulsions and seven OCD subjectswith prominent checking compulsions. All patients viewed a range of normally dis-gusting and washer-relevant (rated as more disgusting by the OCD patients withwashing compulsions) stimuli. They found that unlike the nonpatient control sub-jects and checkers, washers showed a pattern of visual cortical and insular activation,with some activation by washer-relevant pictures in frontal regions including the leftinferior frontal gyrus and right fusiform gyrus not the striatum or thalamus.

Shapira et al. (2003) studied eight OCD subjects with prominent contaminationpreoccupations and cleaning compulsions who viewed a range of threat-inducingand disgust-inducing pictures from the International Affective Picture System duringfunctional magnetic resonance imaging scans. They found that in these OCD patientsthe disgust-inducing condition was strongly associated with activation in the insula,the parahippocampal region, the inferior frontal gyrus (Broadman area [BA] 47), thecaudate nucleus, and the primary sensory cortex.

More recently, Saxena et al. (2004) measured regional cerebral glucose metabolismusing [18F]-fluorodeoxyglucose-in 45 adult OCD subjects meeting DSM-IV criteria forOCD. They stratified the patients into two groups based on the prominence of theirhoarding symptoms. They found that compared to controls, patients with prominenthoarding symptoms had significantly lower glucose metabolism in regions of thecingulate gyrus and higher metabolism in the right dorsolateral prefrontal cortex; andthat OCD patients without prominent hoarding symptoms had significantly higherglucose metabolism in left orbitofrontal cortex, bilateral thalamus, left caudate, andleft dorsolateral prefrontal cortex. Perhaps most interestingly, hoarding severity wasnegatively correlated with glucose metabolism in the cingulate gyrus, and positivelycorrelated with metabolism in the right dorsolateral prefrontal cortex.

Finally, in a recent fMRI study, Mataix-Cols et al. (2003) found that when pre-sented with stimuli that could induce OC symptoms, healthy volunteers experiencedincreases in subjective anxiety. Furthermore, anxiety associated with different symp-tom dimensions was associated with different patterns of activation in ventral, dor-sal prefrontal, and limbic/paralimbic regions. Specifically, provocation of washing-related anxiety was predominantly associated with a pattern of activation withinbilateral visual regions, bilateral dorsal and dorsomedial prefrontal regions (medialand superior frontal gyri, dorsolateral prefrontal cortex, precentral gyrus, and dorsalanterior cingulate gyrus), and bilateral ventral prefrontal and limbic regions (ventro-lateral prefrontal cortex, orbitofrontal cortex, medial frontal gyrus [BA 10], ventralanterior cingulate, insula, and parahippocampal gyrus). Additional regions includedbilateral cerebellum and right putamen bilateral dorsomedial prefrontal cortices, ven-tral prefrontal cortex/limbic regions, and visual regions. In contrast, checking-relatedpictures induced activations mainly in dorsal prefrontal and visual regions, with lessactivation in ventral prefrontal or limbic regions. These findings are similar to thoseof Phillips et al. (2000) and Shapira et al. (2003). Finally, the provocation of hoarding-related anxiety was associated with activation predominantly ventral prefrontal andlimbic regions (orbitofrontal cortex, ventral anterior cingulate cortex, medial/superiorfrontal gyrus, anterior insula, ventrolateral prefrontal gyrus, parahippocampal gyrus,and amygdala) and visual regions. In contrast to the OCD patients with prominent

14 LECKMAN ET AL.

hoarding symptoms, there were few dorsal prefrontal activations (middle and supe-rior frontal gyrus, dorsolateral prefrontal cortex, and precentral gyrus). Surprisingly,there were no significant differences in activation within any regions in response tonormally aversive and hoarding-related pictures.

Taken together, these studies raise the question of whether the lack of perfectreplicability of the findings in previous functional imaging studies of OCD could beaccounted for by phenotypic variations among subjects. If these preliminary findingsare confirmed, they would suggest that discrete neural systems may mediate theexpression of different classes of OC symptoms and that these patterns of activationdiffer in degree from the activations seen among normal controls exposed to the samestimuli.

PREDICTION OF TREATMENT RESPONSE:PHARMACOTHERAPY

While controlled trials with SRIs have demonstrated a selective efficacy in OCD,up to 40–60% of patients do not have a satisfactory outcome (Hollander et al., 2002).Nonresponse to treatment in OCD is associated with serious social disability. Thesedifferences in treatment outcome emphasize the heterogeneity of OCD and the needfor identifying predictors of treatment response. While definitive studies have notbeen undertaken, recent studies have suggested that a symptom-based dimensionalapproach may prove to be valuable for identifying significant predictors of treatmentoutcome. For instance, several studies have shown that patients with high scoreson the hoarding dimension respond more poorly to SSRIs (Black et al., 1998; Erze-govesi et al., 2001; Mataix-Cols et al., 1999; Saxena et al., 2002; Winsberg, Cassic, &Koran, 1999). In another study, high scores on the sexual/religious obsessions fac-tor (Mataix-Cols et al., 1999) were associated with poorer long-term outcome withSSRIs and behavior therapy (BT) in 66 adult outpatients who were followed up from1 to 5 years (Alonso et al., 2001). Another study (Erzegovesi et al., 2001) reportedthat patients with somatic obsessions had poorer insight and responded less well toSSRIs. Other somatic treatments may also help patients with specific symptoms. Forinstance, one study found that patients with symmetry and unusual somatic obses-sions may respond well to monoamine-oxidase inhibitors (Jenike, Baer, Minichiello,Rauch, & Buttolph, 1997). In another study, the presence of symmetry obsessions, or-dering compulsions and hoarding rituals predicted better response in refractory casestreated with cingulotomy (Baer et al., 1995).

COMPLIANCE AND RESPONSE TOBEHAVIORAL INTERVENTIONS

The efficacy of BT for OCD has been demonstrated in numerous controlled andmeta-analytic studies. However, a significant number of patients still remain unim-proved or simply refuse or drop out from this treatment. Some studies have suggestedthat checking rituals may respond less well to BT (Basoglu, Lax, Kasvikis, & Marks,1988; Rachman & Hodgson, 1980) but others found no differences in outcome betweenwashers and checkers (Foa & Goldstein, 1978). Foa and Goldstein (1978), however,

SYMPTOM DIMENSIONS 15

reported that washers and checkers responded at different rates to behavioral treat-ments, with checkers being slower to respond. It is often assumed that patients with“pure” obsessions and mental rituals respond less well to classic behavioral inter-ventions, although data supporting these assumptions are sparse. In a meta-analysis,patients with primary obsessive thoughts without rituals tended to improve less withBT than those who had overt, motor rituals (Christensen, Hadzai-Pavlovic, Andrews& Mattick, 1987). In the Alonso et al. (2001) study, the presence of sexual and/or reli-gious obsessions predicted poorer long-term outcome but, because most patients hadboth SSRIs and BT, it was not clear from this study whether these symptoms predictedpoorer outcome with SRIs, BT, or both.

Patients with hoarding symptoms have been described as having poor compli-ance with and response to BT (Ball, Baer, & Otto, 1996), but little empirical evidenceis available from large patient samples. Using a dimensional approach, Mataix-Colset al. (2002a) examined 153 OCD outpatients who took part in a randomized con-trolled trial of BT. Results showed that high scorers on the hoarding dimension weremore likely to drop out prematurely from the trial and also tended to improve lessthan nonhoarding OCD patients. In addition, high scorers on the sexual/religiousdimension responded less well to BT. Interestingly, patients with mental rituals didas well as other OCD patients in this study. Therefore, it seems that BT is mostlyindicated for patients with contamination/washing, aggressive/checking, and sym-metry/ordering symptoms. Perhaps, previous anecdotal reports of unsuccessful BTin patients with hoarding symptoms are due in part to the propensity of such patientsto discontinue treatment prematurely.

A DEVELOPMENTAL PERSPECTIVE

Children engage in a significant amount of ritualistic, repetitive, and compulsive-like activity that is part of their normal behavioral repertoire. Clinically, this phe-nomenon reaches a peak at about 24 months of age (Gesell & Ilg, 1943). QuotingGesell and Ilg:

Going to bed is also complicated for the two year old . . . bedtime demands haveoften grown into an elaborated and rigid structure. There is a coming upstairs ritual,brushing the teeth ritual, getting into bed, pulling down the shades, kissing, andeven a specially worded good-night ritual. . . . The ritualism so characteristic of 30-months may weigh heavily on the entire household. The child . . . is likely to knowwhere everything belongs and to insist that everything remain in its place. . . . Chairsmust be placed at specific angles and certain pictures must remain on certain tables.

Using a parent-report questionnaire, the Childhood Routines Inventory (CRI),to assess compulsive-like behavior in young children we collected data from 1492parents with children between the ages of 8 and 72 months of age (Evans et al.,1997). The CRI was found to have a strong internal consistency and a two-factorstructure. The first factor accounted for 33% of the variance and included items suchas “lines up objects in straight lines or symmetrical patterns,” “arranges objects orperforms certain behaviors until they seem ‘just right,’” and “prefers to have thingsdone in a particular order.” As such, its content closely resembles obsessions concern-ing a need for symmetry or exactness, repeating rituals, counting compulsions, and

16 LECKMAN ET AL.

ordering/arranging compulsions—the symmetry/ordering dimension of OCD.Evans et al. (1997) found the early emergence of specific behaviors that resemblethe symptom dimensions observed in OCD patients. For example, parents reportedthat their children “arranged objects or performed certain behaviors until they seemed‘just right”’ on average, beginning at age 22–25 months. They also reported that theirchildren “lined up objects in straight lines or in symmetrical patterns,” on average,beginning at age 24–25 months. Similarly, behaviors resembling those associated withthe contamination/washing dimension identified with such items as, “seemed veryconcerned with dirt or cleanliness,” were found to have their mean age of onset from22 to 24 months. Finally, parents reported that their children on average began to“collect or store objects” (resembling the hoarding dimension) from 25 to 27 monthsof age. Although direct evidence linking the emergence of these behaviors to the laterdevelopment of OCD is lacking, investigators have found that aspects of these ritual-istic and compulsive-like behaviors are correlated with children’s fears and phobias(Evans, Gray, & Leckman, 1999; Zohar & Felz, 2001). Further exploration of the factorsthat underlie the emergence and resolution of these behaviors in normally develop-ing children may provide valuable insights into the neurobiological substrates andevolutionary origins of these behaviors.

EVOLUTIONARY PERSPECTIVES

The ultimate causes for many neuropsychiatric disorders including OCD arelikely built into the genetic and neurobiological mechanisms that underlie highly con-served behavioral and cognitive repertoires (Leckman & Mayes, 1998; Marks & Nesse,1994). In the case of OCD and its composite dimensions, such an evolutionary perspec-tive seems particularly apt. Indeed, we hypothesize that each of these OC dimensionscorresponds to unconscious neural evaluation of specific threats (Table 1.2), so thatduring the evolution of our species, it is likely that if our forbears had not been acutelyattuned to potential external threats posed by other humans, by predators, by the ex-ternal manifestations of microbial disease, or periods of privation due to drought,natural disasters, or internecine conflict, our species would not have survived.

Specifically, it is possible that during our evolutionary history, there were timesof great privation such that hoarding was adaptive and likely to enhance the proba-bility of survival and reproductive success. A similar argument can be made for eachof the other dimensions. For example, compulsive checking that items in the homeenvironment were “just right,” or ensuring that food and key aspects of the homeenvironment were free of contamination, would have served families well at somepoints in what Darwin called “the struggle for life.”

The possible evolutionary origins of obsessions and compulsions related to fearsabout harm befalling a close family member are of particular interest as they mayreveal something of the normal states of heightened preoccupation that are associatedwith formation of intimate interpersonal relationships. For example, for expectantparents, the immediate perinatal period involves an altered mental state characterizedby excitement, and heightened sensitivity to environmental and emotive cues. Theinfant becomes an increasingly exclusive focus of thought and action towards theend of pregnancy and the early postpartum period. Cues from the infant before andafter birth as well as the infant’s proximity, physical appearance, and temperament

Tabl

e1.

2.T

hrea

tdom

ains

,con

serv

edbe

havi

ors,

and

dev

elop

men

tale

poch

sas

soci

ated

wit

hhe

ight

ened

sens

itiv

ity

and

obse

ssiv

e-co

mpu

lsiv

e-lik

ebe

havi

ors

Thr

eatd

omai

nFo

cus

ofco

ncer

nM

enta

lsta

teB

ehav

iora

lres

pons

eD

evel

opm

enta

lepo

chs

Har

mfr

omag

gres

sive

beha

vior

from

cons

peci

fics

Wel

l-be

ing

ofse

lfan

dcl

ose

fam

ilym

embe

rs

Intr

usiv

eim

ages

orth

ough

tsth

atco

ntai

nfe

ared

outc

omes

ofse

para

tion

orlo

ss;a

mon

gol

der

child

ren

and

adul

ts—

ahe

ight

ened

sens

eof

resp

onsi

bilit

y

Phys

ical

prox

imit

y;ch

ecki

ngto

ensu

reth

esa

fety

ofcl

ose

fam

ilym

embe

rs;a

void

ance

ofd

ange

r

Ear

lych

ildho

od—

form

atio

nof

atta

chm

entt

oca

regi

vers

;ear

lyfa

mily

life—

preg

nanc

y,d

eliv

ery,

and

care

ofyo

ung

child

ren;

thre

ats

tofa

mily

mem

bers

due

toin

jury

orot

her

exte

rnal

thre

ats

Phys

ical

secu

rity

Imm

edia

teho

me

envi

ronm

ent

Hei

ghte

ned

atte

ntio

nto

the

plac

emen

tofs

peci

ficob

ject

sin

the

envi

ronm

ent

Che

ckin

gto

ensu

reth

atth

ings

look

“jus

trig

ht”

and

are

inth

eir

expe

cted

plac

es;

arra

ngin

g/or

der

ing

obje

cts

Ear

lych

ildho

od—

init

ialp

erio

dof

expl

orat

ion

ofth

eho

me

envi

ronm

ent

byin

fant

san

dto

dd

lers

;ear

lyfa

mily

life—

preg

nanc

y,d

eliv

ery,

and

earl

ych

ildho

od;t

hrea

tsto

fam

ilym

embe

rsd

ueto

inju

ryor

othe

rex

tern

alth

reat

s

Env

iron

men

tal

clea

nlin

ess

Pers

onal

hygi

ene;

hygi

ene

offa

mily

mem

bers

;cl

eanl

ines

sof

the

hom

e

Preo

ccup

atio

nw

ith

intr

usiv

eim

ages

orth

ough

tsth

atco

ntai

nfe

ared

outc

omes

ofbe

ing

dir

tyor

caus

ing

othe

rsto

beill

;am

ong

old

erch

ildre

nan

dad

ults

—a

heig

hten

edse

nse

ofre

spon

sibi

lity

Was

hing

;che

ckin

gto

ensu

recl

eanl

ines

s;av

oid

ance

ofsh

ared

ord

isgu

stin

git

ems

Ear

lych

ildho

od—

init

ialp

erio

dof

sele

ctio

nof

item

sof

food

and

dri

nkby

tod

dle

rs;e

arly

fam

ilylif

e—pr

egna

ncy,

del

iver

y,an

dca

reof

youn

gch

ildre

n;th

reat

sto

fam

ilym

embe

rsd

ueto

inju

ryor

othe

rex

tern

alth

reat

s

Priv

atio

nE

ssen

tial

reso

urce

sPr

eocc

upat

ion

wit

hin

trus

ive

imag

esor

thou

ghts

that

cont

ain

fear

edou

tcom

esof

priv

atio

n;a

heig

hten

edse

nse

ofre

spon

sibi

lity

Col

lect

ing

item

s;ch

ecki

ngto

ensu

reth

esu

ffici

ents

uppl

ies

are

avai

labl

e

Lat

ency

—in

itia

lper

iod

ofco

llect

ing;

earl

yfa

mily

life—

preg

nanc

y,d

eliv

ery,

and

care

ofyo

ung

child

ren