Vous êtes sur la page 1sur 788

Handbook of Autism and

Pervasive Developmental Disorders

Third Edition

Volume 1: Diagnosis, Development, Neurobiology, and Behavior

Edited by
Fred R. Volkmar
Rhea Paul
Ami Klin
Donald Cohen


Handbook of Autism and
Pervasive Developmental Disorders
Third Edition

Volume 1: Diagnosis, Development, Neurobiology, and Behavior

Edited by
Fred R. Volkmar
Rhea Paul
Ami Klin
Donald Cohen


This book is printed on acid-free paper.

Copyright © 2005 by John Wiley & Sons, Inc. All rights reserved.

Published by John Wiley & Sons, Inc., Hoboken, New Jersey.

Published simultaneously in Canada.

MCMI-III is a trademark of DICANDRIEN, Inc.

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any
means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section
107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher,
or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222
Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 646-8600, or on the web at www.copyright.com.
Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons,
Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008.

Limit of Liability/ Disclaimer of Warranty: While the publisher and author have used their best efforts in
preparing this book, they make no representations or warranties with respect to the accuracy or completeness of
the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a
particular purpose. No warranty may be created or extended by sales representatives or written sales materials.
The advice and strategies contained herein may not be suitable for your situation. You should consult with a
professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other
commercial damages, including but not limited to special, incidental, consequential, or other damages.

This publication is designed to provide accurate and authoritative information in regard to the subject matter
covered. It is sold with the understanding that the publisher is not engaged in rendering professional services. If
legal, accounting, medical, psychological or any other expert assistance is required, the services of a competent
professional person should be sought.

Designations used by companies to distinguish their products are often claimed as trademarks. In all instances
where John Wiley & Sons, Inc. is aware of a claim, the product names appear in initial capital or all capital
letters. Readers, however, should contact the appropriate companies for more complete information regarding
trademarks and registration.

For general information on our other products and services please contact our Customer Care Department
within the United States at (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002.

Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be
available in electronic books. For more information about Wiley products, visit our web site at www.wiley.com.

Library of Congress Cataloging-in-Publication Data:

Handbook of autism and pervasive developmental disorders / edited by Fred R. Volkmar . . .

[et al.].—3rd ed.
p. cm.
Includes bibliographical references and index.
Contents: V. 1. Diagnosis, development, neurobiology, and behavior—v. 2. Assessment,
interventions, and policy.
ISBN 0-471-71696-0 (cloth : v. 1)—ISBN 0-471-71697-9 (cloth : v. 2)—ISBN
0-471-71698-7 (set)
1. Autism in children. 2. Developmental disabilities. 3. Autistic children—Services for. 4.
Developmentally disabled children—Services for. I. Volkmar, Fred R.
RJ506.A9H26 2005

Printed in the United States of America.

10 9 8 7 6 5 4 3 2 1
To the Memory of Donald Cohen

At the time of his death, Donald Cohen was actively involved in the planning of this
edition of the Handbook. His untimely passing made it impossible for him to see the
final product. We are deeply grateful to him for his thoughtful counsel and mentorship
as well as the truly impressive example he presented as a clinician-researcher. We hope
that this Handbook is a testament to his vision and a fitting tribute to his memory.

Photo: Michael Marsland, Yale University



University of the Republic Harvard Medical School
School of Medicine Massachusetts Hospital
Cavia Boston, Massachusetts
Montevideo, Uruguay
GEORGE M. ANDERSON, PHD Autism Research Centre
Child Study Center University of Cambridge
Yale University School of Medicine Cambridge, England
New Haven, Connecticut
JOEL R. ARICK, PHD Department of Psychiatry
Special Education University of North Carolina at Chapel Hill
Portland State University Chapel Hill, North Carolina
Portland, Oregon
CHRIS ASHWIN, PHD Center for Autism
Autism Research Centre North Shore Long Island Jewish Health
University of Cambridge System
Departments of Experimental Psychology Bethpage, New York
and Psychiatry
Cambridge, England COURTNEY BURNETTE, MS
Department of Psychology
GRACE T. BARANEK, PHD, OTR/L University of Miami
Division of Occupational Science Coral Gables, Florida
Department of Allied Health Sciences
University of North Carolina at Chapel Hill ALICE S. CARTER, PHD
Chapel Hill, North Carolina Department of Psychology
University of Massachusetts Boston
SIMON BARON-COHEN, PHD Boston, Massachusetts
Autism Research Centre
University of Cambridge BHISMADEV CHAKRABARTI, BA, BSC
Departments of Experimental Psychology Autism Research Centre
and Psychiatry University of Cambridge
Cambridge, England Cambridge, England

viii Contributors


Child Study Center GUATENA
Yale University School of Medicine San Sebastian, Spain
New Haven, Connecticut


Division of Child and Adolescent Psychiatry Special Education
Seoul National University Hospital Portland State University
Seoul, Korea Portland, Oregon

Department of Psychiatry and Behavioral
Saint Mary’s College of California
David Geffen School of Medicine at UCLA
Moraga, California
Los Angeles, California

Department of Psychology and Human PETER F. GERHARDT, EDD
Development Gerhardt Autism /Aspergers Consultation
Vanderbilt University Group, LLC
Nashville, Tennessee Baltimore, Maryland


Autism and Communication Disorders Center TEMPLE GRANDIN, PHD
University of Michigan Department of Animal Science
Ann Arbor, Michigan Colorado State University
Fort Collins, Colorado
Boston University School of Medicine
Boston, Massachusetts RICHARD GRIFFIN, BA
Autism Research Centre
RUTH FALCO, PHD University of Cambridge
Special Education Cambridge, England
Portland State University
Portland, Oregon
PIERRE FERRARI, MD Douglas Developmental Disabilities Center
Centre Hospitalier Public De Psychiatrie Rutgers, The State University of New Jersey
De L’Enfant Et De L’Adolescent New Brunswick, New Jersey
Foundation Vallee
Gentilly Cedex, France
Department of Pediatrics and Neurology Social, Genetic and Developmental
University of California Psychiatry Centre
Irvine College of Medicine Institute of Psychiatry
Orange, California King’s College, London

McGill University SANDRA L. HARRIS, PHD
Department of Psychiatry at the Montreal Douglas Developmental Disabilities Center
Children’s Hospital Rutgers, The State University of New Jersey
Montreal, Quebec, Canada New Brunswick, New Jersey
Contributors ix


The Tavistock Clinic Department of Psychology
Adult Department Illinois Wesleyan University
London, United Kingdom Bloomington, Illinois


Lifespan Services, LLC Communication Crossroads
Princeton, New Jersey North Kingstown, Rhode Island
Department of Neuropsychiatry JOHN LAWSON, PHD
Hikarigaoka Autism Research Centre
Fukushima-shi, Japan University of Cambridge
Cambridge, England
St. George’s Hospital Medical School GABRIEL LEVI, MD
Cranmer Terrace Departimento di Scienze Neurologische e
London, United Kingdom Psichiatriche dell’eta Evolutina
Rome, Italy
Oregon Institute on Disability and
Development Child Development and JENNIFER A. LONCOLA, PHD
Rehabilitation DePaul University
Center Oregon Health and Science University School of Education
Portland, Oregon Chicago, Illinois


Douglas Developmental Disabilities Center Autism Specialist
Rutgers, The State University of New Jersey Oregon Department of Education
New Brunswick, New Jersey Salem, Oregon


Child Study Center UMACC
Yale University School of Medicine University of Michigan
New Haven, Connecticut Ann Arbor, Michigan


Child Study Center
University of Texas Health Sciences Center
Yale University School of Medicine
at Houston
New Haven, Connecticut
Department of Psychiatry and Behavioral
Houston, Texas
Child Study Center
Yale University School of Medicine
New Haven, Connecticut MYRNA R. MANDLAWITZ, BA, MED, JD
MRM Associates
Norwalk, Connecticut
DAVID A. KRUG, PHD Child Study Center
Special Education Yale University School of Medicine
Portland State University Private Practice
Portland, Oregon New Haven, Connecticut
x Contributors


Division TEACCH Department of Psychology
Department of Psychiatry University of Miami
University of North Carolina School Coral Gables, Florida
of Medicine
Chapel Hill, North Carolina J. GREGORY OLLEY, PHD
Clinical Center for the Study of Development
ANDREŚ MARTIN, MD and Learning
Child Study Center University of North Carolina at Chapel Hill
Yale University School of Medicine Chapel Hill, North Carolina
New Haven, Connecticut
Douglas Developmental Disabilities Center U.C. Davis Medical Center
Rutgers, The State University of New Jersey Sacramento, California
New Brunswick, New Jersey
GAIL G. MCGEE, PHD Medical Psychopedagogical Center of
Emory University School of Medicine North Greece
Department of Psychiatry and Behavioral Greece
Department of Occupational Science and
JAMES MCPARTLAND, MS Occupational Therapy
Child Study Center University of Southern California
Yale University School of Medicine Los Angeles, California
New Haven, Connecticut
ADRIENNE MERYL, BA Department of Communication Disorders
M.I.N.D. Institute Southern Connecticut State University
U.C. Davis Medical Center New Haven, Connecticut
Sacramento, California
GARY B. MESIBOV, PHD Center for Children with Special Needs
Division TEACCH Glastonbury, Connecticut
University of North Carolina at Chapel Hill and
Chapel Hill, North Carolina Child Study Center
Yale University School of Medicine
RICHARD MILLS, CQSW, RMPA, MA, FRSA New Haven, Connecticut
NAS Southern Region Office
Church House, Church Road BARRY M. PRIZANT, PHD
Filton, United Kingdom Childhood Communication Services
Cranston, Rhode Island
Western Psychiatric Institute and Clinic Center for the Study of Human Development
Pittsburgh, Pennsylvania Brown University
Providence, Rhode Island
Emory University School of Medicine SHERRI PROVENCAL, PHD
Department of Psychiatry and Behavioral Department of Psychology
Sciences University of Utah
Atlanta, Georgia Salt Lake City, Utah
Contributors xi


Albert Einstein College of Medicine Division TEACCH
Bronx, New York Department of Psychiatry
University of North Carolina School
of Medicine
DIANA L. ROBINS, PHD Chapel Hill, North Carolina
Department of Psychology
Georgia State University
Atlanta, Georgia
Department of Psychology
University of California, San Diego
SALLY J. ROGERS, PHD La Jolla, California
M.I.N.D. Institute
U.C. Davis Medical Center ROBERT T. SCHULTZ, PHD
Sacramento, California Child Study Center
Yale University School of Medicine
EMILY RUBIN, MS, CCC/SLP New Haven, Connecticut
Communication Crossroads
Carmel, California VICTORIA SHEA, PHD
Division TEACCH
The University of North Carolina at
FRCPSYCH, FRS Chapel Hill, North Carolina
Social, Genetic and Developmental
Psychiatry Centre
Institute of Psychiatry MIKLE SOUTH, MS
DeCrespigny Park Department of Psychology
Denmark Hill University of Utah
King’s College, London Salt Lake City, Utah


Child Study Center
Karolinska Institute
Yale University School of Medicine
Department of Woman and Child Health
New Haven, Connecticut
Child and Adolescent Psychiatry Unit
St. Goran’s Children’s Hospital
Stockholm, Sweden MATTHEW STATE, MD, PHD
Child Study Center
Yale University School of Medicine
CELINE SAULNIER, PHD New Haven, Connecticut
Child Study Center
Yale University School of Medicine
New Haven, Connecticut
Vanderbilt Children’s Hospital
Nashville, Tennessee
Yale University School of Medicine Autism Services Center
New Haven, Connecticut Huntington, West Virginia


Kinder Jundenpsychiatrische Klinik Department of Speech Therapy
Zentralinstitute fur Seelische Genundheit Canterbury University
Mannheim, Germany Christchurch, New Zealand
xii Contributors


University of Pittsburgh College of Education
Western Psychiatric Institute and Clinic University of Illinois at Chicago
Pittsburgh, Pennsylvania Chicago, Illinois


McMaster University College of Education
Department of Psychiatry University of Illinois at Chicago
Faculty Health Sciences Chicago, Illinois
Hamilton, Ontario, Canada
Department of Anatomy and Neurobiology Yale University
Boston University School of Medicine New Haven, Connecticut
Boston, Massachusetts
KUO-TAI TAO, MD Divisie Psychiatrie Kinder en
Division of Nanging Jeugdpsychiatrie
Child Mental Health Research Center Utrecht, The Netherlands
Nanging, China
BRUCE TONGE, MD Center for Human Development and Disability
Centre for Developmental Psychiatry Autism Center Psychophysiology Laboratories
Monash Medical Center University of Washington
Australia Seattle, Washington


Mood and Anxiety Disorders Program Department of Communication Disorders
National Institute of Mental Health Executive Director, Center for Autism and
Bethesda, Maryland Related Disorders
Florida State University
KATHERINE D. TSATSANIS, PHD Tallahassee, Florida
Child Study Center
Yale University School of Medicine SALLY WHEELWRIGHT, MA
New Haven, Connecticut Autism Research Centre
University of Cambridge
Center for Human Development Research
University of Texas Health Sciences Center LORNA WING, MD
at Houston National Autistic Society Centre for Social
Houston, Texas and Communication Disorders
Bromley, Kent, United Kingdom
The Geha Psychiatric Hospital DIANNE ZAGER, PHD
The Beilinson Medical Center Pace University
Tel Aviv University Medical School New York, New York
Tel Aviv, Israel
Editorial Board


Child Study Center Department of Child and Adolescent
Yale University School of Medicine Psychiatry
New Haven, Connecticut Göteborg University
Göteborg, Sweden
Department of Psychology LYNN KERN KOEGEL, PHD
University of Massachusetts Boston Autism Research and Training Center
Boston, Massachusetts University of California
Santa Barbara, California
Department of Psychiatry
University of Chicago
Mental Science Institute
Chicago, Illinois
Department of Psychiatry and Behavioral
PETER DOEHRING, PHD University of Texas Health Sciences Center
Delaware Autism Program Houston, Texas
Newark, Deleware
Division TEACCH
Kennedy Center for Research on Human
University of North Carolina
Chapel Hill, North Carolina
Vanderbilt University
Nashville, Tennesse
B. J. FREEMAN, PHD The Center for Children with Special Needs
Department of Psychiatry Glastonbury, Connecticut and
University of California Yale Child Study Center
Los Angeles, California New Haven, Connecticut


Department of Psychiatry Department of Communication Sciences
McGill University University of Vermont
Montreal, Canada Burlington, Vermont

xiv Editorial Board


The M.I.N.D. Institute Center for Autism Research and Treatment
University of California Davis Medical Center UCLA School of Medicine
Sacramento, California Los Angeles, California


Department of Psychology Child Study Center
University of California Yale University School of Medicine
San Diego, California New Haven, Connecticut


Child Study Center Department of Psychiatry
Yale University School of Medicine McMaster University
New Haven, Connecticut Hamilton, Ontario


Director, Autism Clinic Benhaven
Children’s Center at Langley Porter East Haven, Connecticut
University of California
San Francisco, California

A comprehensive Handbook devoted to autism that allow a child to become a family member
and pervasive developmental disorders testifies and social being.
to the volume of research, services, theory, This Handbook is guided by a developmen-
and advocacy related to children and adults tal psychopathological orientation (Cicchetti
with the most severe disorders of development. & Cohen, 1995). Within this framework, prin-
Indeed, the third edition of this work is now ciples and findings about normal development
literally two books. The expansion in size and are used to illuminate how development may
sophistication reflects substantial advances in become derailed and lead to pathological con-
knowledge during the one decade that sepa- ditions, and, conversely, studies of disorders
rates it from its predecessor published in 1997. such as autism are used to cast light on normal
Autism has attracted remarkable interest developmental processes. Autism and similar
and concern of clinicians and researchers from developmental disorders may serve as “experi-
the time of its first scientific description over ments of nature.” Their underlying biology and
60 years ago by Leo Kanner (1943). As a disor- psychology, as well as the types of adaptations
der that afflicts the core of socialization, it that individuals can use to compensate for
has posed scientific challenges to theories of their difficulties, may reveal mechanisms and
developmental psychology and neurobiology as processes that are otherwise concealed from
well as therapy and education. Virtually every awareness or scientific scrutiny.
type of theory relating to child development— As a serious, generally lifelong condition,
cognitive, social, behavioral, affective, neuro- autism has generated important challenges to
biological—has been applied to understanding the systems that relate to individuals with dis-
the enigmatic impairments and competencies abilities, including educational, vocational,
of autistic individuals. And the results of em- medical, and psychiatric systems, as well as to
pirical studies inspired by these diverse theo- social policy, legislation, and the legal systems.
retical perspectives have enriched not only the Because of its multifaceted impact on develop-
field of autism but also the broad field of de- ment, autism also has focused the attention of
velopmental psychopathology. Indeed, autism all the professions concerned with children and
has served as a paradigmatic disorder for adults with difficulties, including psychology,
theory testing and research on the essential education, psychiatry, physical rehabilitation,
preconditions for normal social-cognitive mat- recreational therapy, speech and language,
uration—expression and recognition of emo- nursing, pediatrics, neurology, occupational
tions, intersubjectivity, sharing a focus of therapy, genetics, social work, law, neuroradi-
interest with other people, the meaning and ology, pharmacology—indeed, virtually every
uses of language, forming first attachments caring profession. By drawing these disci-
and falling in love, empathy, the nuanced un- plines together in the clinic and laboratory,
derstanding of the minds of others—indeed, autism has helped forge the multidisciplinary
the whole set of competencies and motivations approach to developmental disabilities. One

xvi Preface

goal of this Handbook is to provide an orienta- family is experienced by parents, siblings, and
tion of shared concepts and knowledge to fa- extended family as profoundly painful. There
cilitate the future collaboration among the can, of course, be consolations in dealing well
disciplines and professionals who work with with adversity; yet, however well a family and
autistic individuals and their families. individual cope, a lifetime with autism brings
Nothing strikes more at the core of a fam- with it more than a fair share of disappoint-
ily’s functioning than the birth of a child with ment, sadness, and emotional scarring for all
a serious disability. Kanner recognized the involved. Only with scientific advances that
central involvement of families in his first re- will prevent, greatly ameliorate, or even cure
ports when he described the peculiarities of these conditions will this pain be fully eased.
social relations in families who came for his Clinicians and researchers have been drawn to
consultation and care. In his first accounts, he autism in the hope of achieving this result, and
misread the data presented to him and postu- their remarkable commitments are also re-
lated an etiologic role of parental behavior in flected in this Handbook and in services
the pathogenesis of autism. This mistake throughout the world.
haunted the field and pained families for many At times, however, therapeutic zeal has ex-
years; it still may arise in certain places, as ceeded the knowledge available. The Hand-
ghosts tend to do. However, Kanner soon book aims at providing authentic knowledge,
righted his theory and emphasized the central broadly accepted by experts. Yet, we recog-
message of his initial report that autism is es- nize that there are sometimes sharp differ-
sentially a reflection of an inborn dysfunction ences of opinion and theoretical perspective
underlying affective engagement. Because so- and that today’s wisdom may be tomorrow’s
cial interaction is a two-way street, parents delusion. Thus, it is important to foster diver-
and others who spend time with an autistic sity while encouraging everyone to pursue rig-
child will no doubt relate differently than with orous, empirical research that will improve
his or her socially engaged, ebullient, linguis- future treatments. Scientific progress oddly
tically gifted siblings. Of interest, more recent leads to many divergent ideas and findings for
genetic information about autism and As- a long time before a deeper level of clarity is
perger syndrome, discussed in the Handbook, achieved.
returns us to Kanner’s observations about so- While we encourage tolerance of differing
cial variations and impairments running scientific views, we do not think that “any-
within families. New findings of aggregation thing goes.” Virtually every month or two,
of autism, cognitive problems, and social dif- parents and others who care for autistic chil-
ficulties within families suggest that an under- dren and adults are likely to hear announce-
lying vulnerability may be transmitted from ments of new, miraculous treatments. They
one generation to the next. If so, explicating may be confused by the options and feel guilty
the interaction between genetic and environ- for not making the sacrifices necessary to try
mental factors in the course of these disorders still another approach. Today, within a stone’s
will bring us back to questions not too far from throw of our own university, parents are en-
where Kanner started his speculations. gaged in a medley of divergent treatments. As
The impact of autistic individuals on family the recent review by the National Research
life has changed with the creation of more ade- Council (2001) has shown, a variety of treat-
quate services. Burdens on families have been ments have now been shown to be effective for
eased by early identification, initiation of edu- individuals with autism. The efficacy of a host
cational and other treatments during the first of other treatments, commonly referred to as
years of life, suitable family guidance and sup- complementary or alternative treatments, re-
port, high-quality educational and other pro- mains to be scientifically well established.
grams, respite care, supportive living and other Often, such treatments compete with more tra-
arrangements for adults with autism, effective ditional ones. Parents, and sometimes profes-
pharmacological treatments, and knowledge sionals, may feel at a loss in terms of
that can guide lifetime planning. Yet, with per- evaluating such treatments and making sound,
haps rare exception, an autistic child in the empirically based decisions about which treat-
Preface xvii

ment(s) should be pursued with respect to an a major impact on the care and treatment of in-
individual child. Occasionally, differences be- dividuals with autism, as well. Far more than
tween advocates and skeptics in relation to most experts believed possible 20 or even 10
treatment ethics and efficacy arouse passions, years ago, many individuals with autism have
including legal proceedings and splits between not only the right but also the capacities to par-
professionals or within the family. How are ticipate within their communities—to study,
parents and professionals best able to make in- work, live, recreate, and share in family life.
formed decisions? The Handbook reflects this important educa-
Like other areas of science, the field of tional and cultural evolution in which a philos-
autism will advance when we adopt, whenever ophy of despair has given way to one of hope.
possible, the rigorous standards of scientific We also appreciate that there are enormous
research. Indeed, our own work as clinician- differences among individuals with autism
researchers has led us to the conclusion that and related conditions in their abilities and
we should offer no less. Thus, in the Handbook needs, among families in their strengths and
we have attempted to provide a comprehensive resources, and among communities and na-
account of current, scientific thinking and tions in their own viewpoints and histories.
findings and to mark out speculation and the- These differences should be respected, and
ory for what these are. We also have eschewed policy and discussion should recognize that
accounts of ideas and treatments, however fas- “autistic people” do not form a homogeneous
cinating they might be, that are too far from class. Clinicians and practitioners generally
the mainstream of scientific research and em- are able to keep the individual at the focus of
pirically guided practice. Such decisions are concern, as we do when we think together with
our responsibility and may leave some advo- families about their unique child or with an
cates feeling shortchanged or even angry; they adult with autism about his or her special life
retain their right to free speech and, who situation. At such times, broader issues of so-
knows, may yet be vindicated. cial policy recede into the background as the
In underlining the importance of data in fullness of the individual’s needs and interests
guiding decisions about treatment, we also rec- are paramount. In shaping social policy and
ognize that clinical care always occurs within planning regional and national systems, how-
a social context and is shaped by beliefs, val- ever, there is a clear consensus for the ap-
ues, and other historical and cultural values. proach to treatment and lifetime planning
Prevailing views about the rights of individu- captured by the ideology of autonomy and
als with disabilities and their role in society community-based living and working. We hope
have changed dramatically over the past that this orientation is conveyed by this Hand-
decades. Embodied in legislation and judicial book. At the same time, there is no single, right
decision, the emergent viewpoints about rights formula for every child or adult with autism: A
to education, services, access, job opportuni- community and nation should strive to have
ties—to basic human respect—have shaped available a spectrum of services to satisfy the
services and improved the quality of the lives varied and changing needs and values of indi-
of individuals who would only decades ago viduals with autism and their families.
have been subject to abuses of various types Clearly defined concepts are essential for
that limited freedom, stigmatized, or dehu- communication among scientists, especially
manized. We have been delighted to see this for interdisciplinary and international collabo-
view gaining increasingly wide acceptance ration. In the field of autism and other behav-
around the world. ioral disorders, there has been substantial
Parents and individuals with disabilities progress in nosology and diagnosis. This prog-
have been effective advocates. Communities ress has enhanced discussion, research, and
and professionals have been sensitized to the cross-disciplinary exchange. It had the merit of
subtle ways in which individuals with disabili- underlining the concept of developmental dis-
ties may be deprived of autonomy and are order and the breadth of dysfunctions in social,
made to be more handicapped by lack of provi- cognitive, language, and other domains. Simi-
sion for their special needs. This trend has had larly, the introduction of multiaxial diagnosis
xviii Preface

underscored the need for patients to be seen the Handbook began in 2000 with an expansion
from varied points of view and the need to of the number of editors in light of the increas-
supplement “categorical disorders” (e.g., ingly diverse and sophisticated body of re-
autism) with knowledge about other aspects of search that was becoming available.
functioning, including medical status and In this edition, we have retained the best
adaptive abilities. As we discuss in the first features of the second edition with expanded
section of this Handbook, advances in classifi- coverage in selected areas. In many instances,
cation have led new knowledge and increas- authors have kindly revised earlier contribu-
ingly focused and refined research. The tions in light of current research; in other
consensus exemplified in Diagnostic and Sta- cases, we have solicited new contributors and
tistical Manual of Mental Disorders, fourth chapters. As a result of the expanded coverage,
edition (DSM-IV; American Psychiatric Asso- the book has expanded into two volumes with a
ciation, 1994), and International Classifica- total of nine sections. This more extensive
tion of Diseases, 10th edition, (ICD-10; World coverage reflects the increasing depth and
Health Organization, 1992), has stimulated a breadth of work within the field.
tremendous increase in research over the past In creating this Handbook, we invited chap-
decade. Today the two internationally recog- ters from recognized scholars. The responses
nized systems provide a consistent approach to to the invitations were gratifying. Each com-
the diagnosis of the most severe disorders of pleted chapter was reviewed by the editors and
early onset. While there are still some regional by two members of a distinguished editorial
or national diagnostic alternatives, the trend committee. The use of peer review is not typi-
is, fortunately, toward consensus. At the same cal for volumes such as this, and we are grate-
time, the universal acceptance of a standard ful that all authors of chapters welcomed this
meter and of Greenwich time does not ensure process. The reviewers wrote careful critiques,
great science or lack of debate and much work sometimes many pages in length; these reviews
remains to be done, but the current approach were provided to the authors for their consider-
has helped provide a solid framework on which ation during revision. The interactive process
future refinements can sensibly be made. of revising chapters has helped ensure that the
The thousands of publications—scientific contributions are as good as the field allows.
papers, monographs, chapters, books—about The past several years have seen a major in-
autism and pervasive developmental disorder crease in the funding of research on autism.
are evidence of its intrinsic interest to re- While we are gratified by this increased sup-
searchers and clinicians and to the human im- port, we hope for even more because only
portance of these disorders for those who through research will we be able to change in-
suffer from them and their families. The grow- cidence and alter the natural history of autistic
ing body of books and resources specifically and other pervasive disorders. The cost of car-
designed for parents and family members has ing for one autistic individual over a lifetime
been a noteworthy achievement of the past sev- may be more than any single investigator will
eral years. At the same time, you could reason- ever have to spend during a career of research.
ably ask why a revision of the Handbook is Many hundreds of millions of dollars are spent
needed now. internationally on direct services; only a tiny
This third edition of the Handbook of percentage of this expenditure is devoted to
Autism and Pervasive Developmental Disorders any type of formal research. It is as if the
is the second revision of a book that first ap- United States committed all of its funding to
peared in 1987. This edition quickly became building iron lungs and considered virology to
established as an important scholarly resource. be a secondary concern in relation to polio. To
Within a decade much had changed, and the fully exploit the many new methods for study-
second edition of this volume appeared. The ing brain development and brain-behavior rela-
rapid pace of scientific progress was reflected tions and to attempt to translate biological and
in the second edition, which was expanded to behavioral research findings into treatments
increase coverage of new research and treat- will require substantial investment of research
ment methods. Preparations for this version of funds. The recent network of federal centers
Preface xix

through the Collaborative Program of Excel- A Handbook portrays what is known and re-
lence in Autism (CPEA) and the Studies to veals what is poorly understood. Although
Advance Autism Research and Treatment many studies have been conducted and areas
(STAART) as well as through the Research explored, there is no hard biological or behav-
Units on Psychopharmacology (RUPP) and the ioral finding that can serve as a reliable com-
Centers for Disease Control (CDC) have al- pass point to guide research; in spite of great
ready had major benefits. These benefits will efforts and decades of commitment by re-
eventually include not only a reduction in suf- searchers and clinicians, the fate of many
fering and in costs for those with autism, but autistic individuals remains cloudy; and even
also important knowledge that will benefit a with new knowledge, there are still too many
far larger group of children and adults with areas of controversy. That investigators and
other serious neuropsychiatric and develop- clinicians, working alongside families and ad-
mental disorders. We hope that one contribu- vocates, have learned so much, often with very
tion of the Handbook will be to underscore the tight resources, speaks to their commitment to
gains from systematic research and the impor- understanding and caring for autistic children
tance of sustained support for multidiscipli- and adults. The goal of this Handbook is to
nary clinical research groups. document their achievements and inspire their
We wish to recognize the support that has future efforts.
been provided over the decades to our own clin- FRED R. VOLKMAR, MD
ical and research program by the National Insti- AMI KLIN, PHD
tute of Child Health and Human Development, RHEA PAUL, PHD
National Institute of Deafness and Communica- Yale Child Study Center
tion Disorders, and the National Institute of New Haven, Connecticut
Mental Health, as well as by the Korczak Foun- November, 2004
dation, the W. T. Grant Foundation, the Doris
Duke Foundation, the Simon’s Foundation, REFERENCES
Cure Autism Now, the National Alliance for American Psychiatric Association. (1980). Diag-
Autism Research, and private donors. nostic and statistical manual of mental disor-
We thank the members of our editorial ders (3rd ed.). Washington, DC: Author.
board for their excellent contributions to this American Psychiatric Association. (1994). Diag-
process and Lori Klein, who helped us coor- nostic and statistical manual of mental disor-
dinate this effort, as well as the wonderful ders (4th ed.). Washington, DC: Author.
editorial staff at Wiley, who have consis- Cohen, D. J., & Donnellan, A. M. (1987). Hand-
tently sought to help us deliver the best possi- book of Autism and Pervasive Developmental
ble work. We have been very fortunate in Disorders. New York: Wiley.
Cicchetti D., & Cohen D. J. (1995). Developmental
being able to work within the scholarly envi-
Psychopathology. (Vols. 1–2). New York: Wiley.
ronment provided by the Yale School of Medi-
Kanner, L. (1943). Autistic disturbances of affec-
cine and the Child Study Center. The unique tive contact. Nervous Child 2, 217–250.
qualities of the Child Study Center reflect the Volkmar, F., Klin, A., Siegel, B., et al. (1994). Field
contributions of generations of faculty who trial for autistic disorder in DSM-IV. American
have committed themselves to clinical schol- Journal of Psychiatry, 151, 1361–1367.
arship, teaching, and service. We particularly World Health Organization. (1977). Manual of the
wish to acknowledge the guidance and sup- international statistical classification of dis-
port of senior mentors—Albert J. Solnit, eases, injuries and causes of death (9th ed.,
Sally Provence, Sam Ritvo, Sara Sparrow, Vol. 1). Geneva, Switzerland: Author.
and Edward Zigler—as well as many col- World Health Organization. (1992). The ICD-10
classification of mental and behavioral disor-
leagues and collaborators in this work, in-
ders. Clinical descriptions and diagnostic
cluding Robert Schultz, Cheryl Klaiman, guidelines. Geneva, Switzerland: Author.
Larry Scahill, Matt State, Elenga Grigorenko, World Health Organization. (1993). The ICD-10
George Anderson, James Leckman, Kasia classification of mental and behavioral disor-
Chawarska, Katherine Tsatsanis, Wendy ders. Diagnostic criteria for research. Geneva,
Marans, and Emily Rubin. Switzerland: Author.




Chapter 1. Issues in the Classification of Autism and Related Conditions 5

Fred R. Volkmar and Ami Klin

Chapter 2. Epidemiological Studies of Pervasive Developmental Disorders 42

Eric Fombonne

Chapter 3. Childhood Disintegrative Disorder 70

Fred R. Volkmar, Kathy Koenig, and Matthew State

Chapter 4. Asperger Syndrome 88

Ami Klin, James McPartland, and Fred R. Volkmar

Chapter 5. Rett Syndrome: A Pervasive Developmental Disorder 126

Richard Van Acker, Jennifer A. Loncola, and Eryn Y. Van Acker

Chapter 6. Pervasive Developmental Disorder Not Otherwise Specified 165

Kenneth E. Towbin

Chapter 7. Outcomes in Autism Spectrum Disorders 201

Patricia Howlin


Chapter 8. Autism in Infancy and Early Childhood 223

Katarzyna Chawarska and Fred R. Volkmar

Chapter 9. The School-Age Child with an Autistic Spectrum Disorder 247

Katherine A. Loveland and Belgin Tunali-Kotoski

xxii Contents

Chapter 10. Adolescents and Adults with Autism 288

Victoria Shea and Gary B. Mesibov

Chapter 11. Social Development in Autism 312

Alice S. Carter, Naomi Ornstein Davis, Ami Klin, and Fred R. Volkmar

Chapter 12. Language and Communication in Autism 335

Helen Tager-Flusberg, Rhea Paul, and Catherine Lord

Chapter 13. Neuropsychological Characteristics in Autism and Related Conditions 365

Katherine D. Tsatsanis

Chapter 14. Imitation and Play in Autism 382

Sally J. Rogers, Ian Cook, and Adrienne Meryl

Chapter 15. Autism and Emotion 406

Peter Hobson


Chapter 16. Genetic Influences and Autism 425

Michael Rutter

Chapter 17. Neurochemical Studies of Autism 453

George M. Anderson and Yoshihiko Hoshino

Chapter 18. Neurologic Aspects of Autism 473

Nancy J. Minshew, John A. Sweeney, Margaret L . Bauman,
and Sara Jane Webb

Chapter 19. Functional Neuroimaging Studies of Autism Spectrum Disorders 515

Robert T. Schultz and Diana L . Robins

Chapter 20. Medical Aspects of Autism 534

Pauline A. Filipek


Chapter 21. Problems of Categorical Classification Systems 583

Lorna Wing

Chapter 22. Executive Functions 606

Sally Ozonof f, Mikle South, and Sherri Provencal

Chapter 23. Empathizing and Systemizing in Autism Spectrum Conditions 628

Simon Baron-Cohen, Sally Wheelwright, John Lawson, Richard Grif fin,
Chris Ashwin, Jac Billington, and Bhismadev Chakrabarti
Contents xxiii

Chapter 24. The Weak Central Coherence Account of Autism 640

Francesca Happé

Chapter 25. Joint Attention and Neurodevelopmental Models of Autism 650

Peter Mundy and Courtney Burnette

Chapter 26. The Enactive Mind—From Actions to Cognition: Lessons from Autism 682
Ami Klin, Warren Jones, Robert T. Schultz, and Fred R. Volkmar

Author Index I•1

Subject Index I•39




Chapter 27. Screening for Autism in Young Children 707

Elaine E. Coonrod and Wendy L . Stone

Chapter 28. Diagnostic Instruments in Autistic Spectrum Disorders 730

Catherine Lord and Christina Corsello

Chapter 29. Clinical Evaluation in Autism Spectrum Disorders: Psychological

Assessment within a Transdisciplinary Framework 772
Ami Klin, Celine Saulnier, Katherine Tsatsanis, and Fred R. Volkmar

Chapter 30. Assessing Communication in Autism Spectrum Disorders 799

Rhea Paul

Chapter 31. Behavioral Assessment of Individuals with Autism: A Functional

Ecological Approach 817
Michael D. Powers

Chapter 32. Sensory and Motor Features in Autism: Assessment and Intervention 831
Grace T. Baranek, L . Diane Parham, and James W. Bodfish


Chapter 33. Curriculum and Classroom Structure 863

J. Gregory Olley

Chapter 34. Behavioral Interventions to Promote Learning in Individuals with Autism 882
Laura Schreibman and Brooke Ingersoll

Chapter 35. Behavioral Interventions 897

Joel D. Bregman, Dianne Zager, and John Gerdtz
xxiv Contents

Chapter 36. Critical Issues in Enhancing Communication Abilities for Persons with
Autism Spectrum Disorders 925
Barry M. Prizant and Amy M. Wetherby

Chapter 37. Enhancing Early Language in Children with Autism Spectrum Disorders 946
Rhea Paul and Dean Sutherland

Chapter 38. Addressing Social Communication Skills in Individuals with

High-Functioning Autism and Asperger Syndrome: Critical Priorities
in Educational Programming 977
Wendy D. Marans, Emily Rubin, and Amy Laurent

Chapter 39. School-Based Programs 1003

Joel R. Arick, David A. Krug, Ann Fullerton, Lauren Loos, and Ruth Falco

Chapter 40. Helping Children with Autism Enter the Mainstream 1029
Jan S. Handleman, Sandra L . Harris, and Megan P. Martins

Chapter 41. Models of Educational Intervention for Students with Autism: Home, Center,
and School-Based Programming 1043
Sandra L . Harris, Jan S. Handleman, and Heather K. Jennett

Chapter 42. Working with Families 1055

Lee M. Marcus, Linda J. Kunce, and Eric Schopler

Chapter 43. Employment: Options and Issues for Adolescents and Adults with
Autism Spectrum Disorders 1087
Peter F. Gerhardt and David L . Holmes

Chapter 44. Psychopharmacology 1102

Lawrence Scahill and Andrés Martin


Chapter 45. Preparation of Autism Specialists 1123

Gail G. McGee and Michael J. Morrier

Chapter 46. Educating Children with Autism: Current Legal Issues 1161
Myrna R. Mandlawitz

Chapter 47. Cross-Cultural Program Priorities and Reclassification of Outcome

Research Methods 1174
Eric Schopler


Chapter 48. International Perspectives 1193

Fred R. Volkmar
Contents xxv


Chapter 49. Community-Integrated Residential Services for Adults with Autism:

A Working Model (Based on a Mother’s Odyssey) 1255
Ruth Christ Sullivan

Chapter 50. A Sibling’s Perspective on Autism 1265

Jason B. Konidaris

Chapter 51. A Personal Perspective of Autism 1276

Temple Grandin

Chapter 52. A Teacher’s Perspective: Adult Outcomes 1287

Virginia Walker Sperry

Chapter 53. Autism: Where We Have Been, Where We Are Going 1304
Isabelle Rapin

Author Index I•1

Subject Index I•39



The paired processes of diagnosis and classifi- unique signs and symptoms are provided a
cation are fundamental to research and inter- context. They are given a more general mean-
vention. The diagnostic process includes all of ing. For example, the clinician will assign the
the activities in which a clinician engages in patient’s coughing and fever to the category
trying to understand the nature of an individ- pneumonia. This categorical diagnosis is
ual’s difficulty. The result of this process is placed within the narrative of the patient’s life
often a narrative account—a portrait of the in- and current problems. It may be related to the
dividual’s past, the current problems, and the patient’s family or genetic background, expe-
ways in which these problems can be related to riences, exposures, vulnerabilities, and the
each other and to possible, underlying causes. like, and it will be used to explain why the pa-
A useful diagnostic process also suggests tient has come for help and what type of treat-
methods for being helpful, including specific ment may be useful.
treatments. In the course of the diagnostic pro- The diagnostic process is based on current
cess, a clinician will learn about the patient’s knowledge, technologies, and skills; it can
history, talk to others about the patient, ob- sometimes be quite brief (as in the diagnostic
serve the patient, engage in specialized exami- processes for an earache) or remarkably exten-
nations, and use laboratory and other methods sive (as in the diagnostic process for autism).
for helping define patients’ problems and their Diagnostic classifications, also, are based on
causes. The clinician will integrate the find- available knowledge and laboratory methods;
ings from these activities, based on special- they also embody conventions, the consensus
ized, scientific knowledge. Often, a patient among clinicians and experts about a useful
will have several types of problems; the diag- way for sorting illnesses and troubles.
nostic process may lead to a narrative that New knowledge and methodologies change
links these to an underlying, common cause or the diagnostic process as well as the classifica-
may separate the problems on the basis of their tion system. The advent of methods such as mo-
differing causes or treatments. Often, more lecular genetic testing, magnetic resonance
than one clinician may be involved in the diag- imaging of the brain, and structured, formal as-
nostic process; then, the final clinical, diag- sessment of cognitive processes have changed
nostic formulation will integrate the pooled the diagnostic process and classification and
information into a coherent and consensual will continue to do so in the future.
narrative that reflects the varied information. The skillful diagnostic process, and the re-
One component of the diagnostic process is sultant account about the patient and his ill-
the assignment of the patient’s difficulties—his ness, often is broad-based, nuanced, and
or her signs, symptoms, pains, troubles, worries, individualized. The clinical formulation, the
dysfunctions, abnormal tests—to a specific full statement of findings, may capture the
class or category of illness or disorder. Through many dimensions of a person’s life, including
classification, the patient’s individualized, his or her competencies as well as specific

2 Diagnosis and Classification

impairments and difficulties. However, a diag- criteria were provided for PDD, but the clini-
nostic categorization—a label or classification cal description conveyed a sense of the contour
of specific troubles and their designation as a of its clinical territory. To be a citizen of this
syndrome, disorder or disease—-is delimited. territory, a child had to exhibit difficulties
Providing the label of a specific disease delim- from the first several years of life involving
its individuality for the sake of being able to several domains (social, language, emotional,
utilize general knowledge gained from scien- cognitive) and with significant impairment of
tific study and experience with others with functioning. In 1980, and again when DSM-III
similar problems. In this important respect, it was revised in 1987 (DSM-III-R), the only ex-
is useful to think that individuals are engaged ample of a specifically defined example of
in the process of diagnosis and symptoms and PDD was autism. Indeed, autism remains the
signs are classified and labeled. A diagnostic paradigm or model form of PDD. From 1980 to
label is not able or meant to capture the full- 1994, other children whose difficulties were
ness of an individual. Diagnostic classification captured by the sense of PDD, but who were
systems and specific assignment to a disease not diagnosed as having autism, were de-
or disorder category are tools, which when scribed as having “pervasive developmental
combined with other tools should lead to help- disorder that is not otherwise specified”
ful understanding and treatment. (PDD-NOS). Although not an official diagnos-
The newer methods of classification of de- tic term, the phrase autism spectrum disorder
velopmental, psychiatric, behavioral, or mental (ASD) is now in widespread use and is synony-
disorders respect the distinction between diag- mous with the term PDD.
nosing an individual and classifying his or her The 1994 edition of the Manual of Mental
problems. They are also multidimensional and Disorders (DSM-IV), based on new evidence
elicit information about other domains of the and international field testing, refined the di-
patient’s life, in addition to areas of leading agnostic criteria for autism and formalized
impairment. This approach shapes and has three new classes or types of pervasive devel-
been shaped by the two international systems opmental disorders: childhood disintegrative
of classification in which autism and perva- disorder, Asperger’s disorder, and Rett’s dis-
sive developmental disorders are included: the order. Also, a consensus was reached between
Diagnostic and Statistical Manual of Mental the two major systems, DSM and ICD, for the
Disorders of the American Psychiatric Associ- system of classification and specific diagnos-
ation and the International Statistical Classifi- tic criteria. Thus, for the first time, there is
cation of Diseases and Related Health Problems happily an internationally accepted, field-
of the World Health Organization (WHO). The tested, diagnostic system for the most severe
introductions to the recent editions of these disorders of development. The DSM-IV and
two systems (DSM-IV, American Psychiatric ICD-10 systems form the epistemological
Association, 1994; and ICD-10, WHO, 1992) backbone of this Handbook.
provide helpful overviews of the goals of clas- The chapters in this section of the Handbook
sification and the roles of diagnostic cate- describe current frameworks for classification,
gories in clinical understanding. the four forms of pervasive developmental dis-
A new diagnostic term was introduced in the orders for which specific criteria are provided
DSM-III in 1980: the concept of pervasive de- in DSM-IV, and the kinds of disturbances that
velopmental disorder (PDD). The umbrella remain within the territory of pervasive devel-
term PDD gained broad popularity among pro- opmental disorders that are not further classi-
fessionals from various disciplines as well as fied. This section also provides a review of
with parents and advocates. Without a previous studies of natural history and outcome.
history in psychiatry, psychology, or neurology, It is our expectation that advances in under-
the novel term PDD had the advantage of not standing the pathogenesis of pervasive develop-
carrying excessive theoretical baggage or con- mental disorders will continue to have a major
troversy. It also had a broad inter-disciplinary impact on the diagnostic and classification
appeal and a nice emphasis on development and processes. Thus, in any discussion about diag-
disorders of development. No specific diagnostic nosis and nosology, it is important to recognize
Diagnosis and Classification 3

their provisional nature. Advances in knowl- American Psychiatric Association. (1987). Diag-
edge may lead to changes in diagnostic ap- nostic and statistical manual of mental disor-
proaches. It is also critical to remember the ders (3rd ed., rev.). Washington, DC: Author.
importance of balancing categorical ap- American Psychiatric Association. (1994). Diag-
proaches to diagnosis with a fuller understand- nostic and statistical manual of mental disor-
ders (4th ed.). Washington, DC: Author.
ing of the many dimensions of individual
World Health Organization. (1992). International
children and adults, that is, as whole people. classification of diseases (10th ed.). Geneva,
Switzerland: Author.

American Psychiatric Association. (1980). Diag-

nostic and statistical manual of mental disor-
ders (3rd ed.). Washington, DC: Author.

Issues in the Classification of Autism and

Related Conditions


Clinicians and researchers have achieved con- et al., 2000; Piven, Palmer, Jacobi, Childress,
sensus on the validity of autism as a diagnostic & Arndt, 1997; Volkmar, Lord, Bailey,
category and the many features central to its Schultz, & Klin, 2004).
definition (Rutter, 1996). This has made pos- Today, autism is probably the complex psy-
sible the convergence of the two major diag- chiatric or developmental disorder with the
nostic systems: the fourth edition of the best empirically based, cross-national diag-
American Psychiatric Association’s Diagnos- nostic criteria. Data from a number of re-
tic and Statistical Manual of Mental Disorders search groups from around the world have
(DSM-IV, 1994) and the 10th edition of the In- confirmed the usefulness of current diagnostic
ternational Classification of Diseases (ICD- approaches, and, even more importantly, the
10; World Health Organization [WHO], 1992). availability of a shared clinical concept and
Although some differences remain, these language for differential diagnosis is a great
major diagnostic systems have become much asset for clear communication among clini-
more alike than different; this has facilitated cians, researchers, and advocates alike (Buite-
the development of diagnostic assessments laar, Van der Gaag, Klin, & Volkmar, 1999;
“ keyed” to broadly accepted, internationally Magnusson & Saemundsen, 2001; Sponheim,
recognized guidelines (Rutter, Le Couteur, & 1996; Sponheim & Skjeldal, 1998). In the fu-
Lord, 2003; see Chapter 28, this Handbook, ture, the discovery of biological correlates,
Volume 2). It is somewhat surprising that, as causes, and pathogenic pathways will, no
greater consensus has been achieved on the doubt, change the ways in which autism is
definition of strictly defined autism, an inter- diagnosed and may well lead to new nosologi-
esting and helpful discussion on issues of cal approaches that, in turn, will facilitate fur-
“ broader phenotype” or potential variants of ther scientific progress (Rutter, 2000).
autism has begun (Bailey, Palferman, Heavey, Simultaneously, considerable progress has
& Le Couteur, 1998; Dawson et al., 2002; been made on understanding the broader range
Pickles, Starr, Kazak, Bolton, Papanikolaou, of difficulties included within the autism

The authors acknowledge the support of the National Institute of Child Health and Human Development
(CPEA program project grant 1PO1HD3548201, grant 5-P01-HD03008, and grant R01-HD042127-02), the
National Institute of Mental Health (STAART grant U54-MH066494), the Yale Children’s Clinical Re-
search Center, and of the National Alliance of Autism Research, Cure Autism Now, and the Doris Duke
Foundation as well as the Simons Foundation. We also gratefully acknowledge the helpful comments of Pro-
fessor Michael Rutter on an earlier version of this manuscript.

6 Diagnosis and Classification

spectrum; that is, as our knowledge of autism DEVELOPMENT OF AUTISM AS A

has advanced, so has our understanding of DIAGNOSTIC CONCEPT
a broader range of conditions with some
similarities to it. Table 1.1 lists categories of Although children with what we now would
pervasive developmental disorders (PDDs) as describe as autism had probably been de-
classified by ICD-10 and DSM-IV. scribed much earlier as so called wild or feral
In addition to the international and cross- children (Candland, 1993; Simon, 1978) it was
disciplinary agreement about diagnostic crite- Leo Kanner who first elaborated what today
ria for autism, a consensus has emerged about would be termed the syndrome of childhood
other issues that were once debated. Today, autism.
there is broad agreement that autism is a
developmental disorder, that autism and asso- Kanner’s Description—Early
ciated disorders represent the behavioral Controversies
manifestations of underlying dysfunctions in
the functioning of the central nervous system, Kanner’s (1943) seminal clinical description
and that sustained educational and behavioral of 11 children with “autistic disturbances of
interventions are useful and constitute the affective contact ” has endured in many ways.
core of treatment (National Research Coun- His description of the children was grounded
cil, 2001). in data and theory of child development, par-
In this chapter, we summarize the develop- ticularly the work of Gesell, who demon-
ment of current diagnostic concepts with a strated that normal infants exhibit marked
particular focus on autism and on the empiri- interest in social interaction from early in life.
cal basis for its current official definition. We Kanner suggested that early infantile autism
address the rationale for inclusion of other was an inborn, constitutional disorder in which
nonautistic PDDs/autism spectrum disorders children were born lacking the typical motiva-
(ASDs), which are discussed in detail in other tion for social interaction and affective com-
chapters in this section. We also note areas in ments. Using the model of inborn errors of
which knowledge is lacking, such as the rela- metabolism, Kanner felt that individuals with
tionships of autism to other comorbid condi- autism were born without the biological pre-
tions and the ongoing efforts to provide conditions for psychologically metabolizing
alternative approaches to subtyping these the social world. He used the word autism to
conditions. convey this self-contained quality. The term

TABLE 1.1 Conditions Currently Classified as Pervasive Developmental Disorders Correspondence

of ICD-10 and DSM-IV Categories
Childhood autism Autistic disorder
Atypical autism Pevasive developmental disorder not otherwise specified
Rett syndrome Rett’s disorder
Other childhood disintegrative disorder Childhood disintegrative disorder
Overactive disorder with mental retardation No corresponding category with stereotyped movements
Asperger syndrome Asperger’s disorder
Other pervasive developmental disorder PDD-NOS
Pervasive developmental disorder, unspecified PDD-NOS
Sources: Diagnostic and Statistical Manual of Mental Disorders, fourth edition, by American Psychiatric Associa-
tion, 1994, Washington, DC: Author; and International Classification of Diseases: Diagnostic Criteria for Research,
tenth edition, by the World Health Organization, 1992, Geneva, Switzerland: Author.
Issues in the Classification of Autism and Related Conditions 7

was borrowed from Bleuler (1911/1950), who children were born long after the theory was
used autism to describe idiosyncratic, self- dead; unfortunately, this notion still prevails in
centered thinking. Autism for Kanner was in- some countries.
tended to suggest that autistic children, too, Two types of information went against the
live in their own world. Yet, the autism of indi- psychogenic theories. It is now known that
viduals with autism is distinct from that of children with autism are found in families
schizophrenia: It represents a failure of devel- from all social classes if studies control for
opment, not a regression, and fantasy is impov- possible factors that might bias case ascertain-
erished if present at all. The sharing of the ment (e.g., Wing, 1980); while additional data
term increased early confusion about the rela- on this topic are needed, more recent and rig-
tionship of the conditions. orous research has failed to demonstrate asso-
In addition to the remarkable social failure ciations with social class (see Chapter 2, this
of autistic individuals, Kanner observed other Handbook, this volume, for a review). A more
unusual features in the clinical histories of the central issue relevant to psychogenic etiology
children. Kanner described the profound dis- concerns the unusual patterns of interaction
turbances in communication. In the original that children with autism and related condi-
cohort, three of the children were mute. The tions have with their parents (and other people
language of the others was marked by as well). The interactional problems of autistic
echolalia and literalness, as well as a fascinat- individuals clearly can be seen to arise from
ing difficulty with acquiring the use of the the side of the child and not the parents
first person, personal pronoun (“I”), and refer- (Mundy, Sigman, Ungerer, & Sherman, 1986)
ral to self in the third person (“ he” or by first although parents may be at risk for various
name). Another intriguing feature was the problems (see Chapter 15, this Handbook, this
children’s unusual responses to the inanimate volume). Probably most important, data sup-
environment; for example, a child might be un- port the role of dysfunction in basic brain sys-
responsive to parents, yet overly sensitive to tems in the pathogenesis of the disorder (see
sounds or to small changes in daily routine. Volkmar et al., 2004). Today, the data appear
While Kanner’s brilliant clinical accounts to support the concept that biological factors,
of the unusual social isolation, resistance to particularly genetic ones, convey a vulnerabil-
change, and dysfunction in communication ity to autism; as Rutter (1999) has noted, the
have stood the test of time, other aspects of the issue of interaction between genetic and envi-
original report have been refined or refuted by ronmental vulnerabilities of all types remains
further research. an important one relevant to a host of disorders
A contentious issue early in the history of in addition to autism.
autism research concerned the role of parents Kanner speculated that autism was not re-
in pathogenesis. Kanner observed that parents lated to other medical conditions. Subsequent
of the initial cases were often remarkably suc- research has shown that various medical con-
cessful educationally or professionally; he also ditions can be associated with autism (see
appreciated that there were major problems in Chapter 2, this Handbook, this volume) and,
the relations between these parents and their most importantly, that approximately 25% of
child. In his initial paper, he indicated that he persons with autism develop a seizure disorder
believed autism to be congenital, but the issue (Rutter, 1970; Volkmar & Nelson, 1990; see
of potential psychological factors in causing also Chapters 18 & 20, this Handbook, this
autism was taken up by a number of individu- volume). With the recognition of the preva-
als; this issue plagued the history of the field lence of medical problems, some investigators
for many years. From the 1960s, however, it proposed a distinction between “primary” and
has been recognized that parental behavior as “secondary” autism depending on whether as-
such played no role in pathogenesis. Yet, the sociated medical conditions, for example, con-
pain of parents having been blamed for a genital rubella (Chess, Fernandez, & Korn,
child’s devastating disorder tended to linger in 1978), could be demonstrated. As time went
the memories of families, even those whose on, it became apparent that, in some basic
8 Diagnosis and Classification

sense, all cases were “organic,” and designa- Other Diagnostic Concepts
tions such as primary and secondary autism
are no longer generally made. In contrast to autism, the definition of autistic-
Kanner also misconstrued the relation be- like conditions remains in need of more clarifi-
tween autism and intellectual disability. His cation (Rutter, 1996; Szatmari, 2000; Szat-
first cases were attractive youngsters without mari, Volkmar, & Walther, 1995). Although
unusual physical features, who performed well the available research is less extensive than
on some parts of IQ tests (particularly those that on autism, several of these autistic-like
that test rote memory and copying, such as conditions were well enough studied, broadly
block design, rather than comprehension of ab- recognized, and clinically important enough to
stract, verbal concepts). Kanner felt that autis- be included in DSM-IV and ICD-10. We antic-
tic children were not mentally retarded, and he, ipate that further studies will improve the def-
and many psychologists after him, invoked mo- inition of these conditions and that new
tivational factors to explain poor performance. disorders may well be delineated within the
Autistic individuals were called “ functionally broad and heterogeneous class of PDD.
retarded.” Decades of research have now shown Diagnostic concepts with similarities to
that when developmentally appropriate tests are autism were proposed before and after Kan-
given in their entirety, full-scale intelligence ner’s clinical research. Shortly after the turn
and developmental scores (IQ and DQ scores) of the century, Heller, a special educator in
are in the mentally retarded range for the ma- Vienna, described an unusual condition in
jority of individuals with autism (Rutter, Bai- which children appeared normal for a few
ley, Bolton, & Le Couter, 1994) and maintain years and then suffered a profound regression
stability over time (Lockyer & Rutter, 1969, in their functioning and a derailment of future
1970). Kanner’s impression of potentially nor- development (Heller, 1908). This condition
mal intelligence, even in the face of apparent was originally known as dementia infantilis or
retardation, was based on what has proven to disintegrative psychosis; it now has official
be a consistent finding on psychological test- status in DSM-IV as childhood disintegrative
ing. Children with autism often have unusu- disorder (see Chapter 3, this Handbook, this
ally scattered abilities, with nonverbal skills volume). Similarly, the year after Kanner’s
often significantly advanced over more ver- original paper, Hans Asperger, a young physi-
bally mediated ones (see Chapter 29, this cian in Vienna, proposed the concept of autis-
Handbook, Volume 2); at the same time, chil- tic psychopathy or, as it is now known,
dren with autism differ in their pattern of Asperger’s disorder (Asperger, 1944; see
behavior and cognitive development from chil- Chapter 4, this Handbook, this volume). Al-
dren with severe language disorders (Bartak, though Asperger apparently was not aware of
Rutter, & Cox, 1977). On the other hand, Kanner’s paper or his use of the word autism,
when the focus shifts from autism, strictly de- Asperger used this same term in his descrip-
fined, to the broader autistic spectrum, a tion of the marked social problems in a group
much broader range of IQ scores is observed of boys he had worked with. Asperger’s con-
(Bailey et al., 1998). cept was not widely recognized for many
The severity of the autistic syndrome led years, but it has recently received much
some clinicians in the 1950s to speculate that greater attention and is now included in both
autism was the earliest form of schizophrenia DSM-IV and ICD-10. Another clinician, An-
(Bender, 1946). Clinicians during the first dreas Rett, observed an unusual developmental
decades of the study of autism tended to at- disorder in girls (Rett, 1966) characterized by
tribute complex mental phenomena such as hal- a short period of normal development and then
lucinations and delusions to children who were, a multifaceted form of intellectual and motor
and remained, entirely mute (Volkmar & Cohen, deterioration. Rett’s disorder is also now offi-
1991a). In the 1970s, research findings began to cially included in the PDD class (see Chapter 5,
show that these two conditions are quite dis- this Handbook, this volume).
parate in terms of onset patterns, course, and The descriptions proposed by some other
family genetics (Kolvin, 1971; Rutter, 1972). clinicians have not fared as well. For example,
Issues in the Classification of Autism and Related Conditions 9

Mahler, a child psychoanalyst, proposed the studies, to share knowledge among investiga-
concept of symbiotic psychosis (Mahler, 1952) tors, and to encourage the development of a
for children who seemed to fail in the task of body of knowledge. For clinicians and educa-
separating their psychological selves from the tors, classification helps guide selection of
hypothesized early fusion with their mothers. treatments for an individual and the evaluation
This concept now has only historical interest, of the benefits of an intervention for groups of
as does her view of a “normal autistic phase” individuals with shared problems (Cantwell,
of development. In contrast, Rank (1949), also 1996). For the legal system, government regu-
working from the framework of psychoanaly- lation, insurance programs, and advocates,
sis, suggested that there is a spectrum of dys- classification systems define individuals with
functions in early development that affects special entitlements. If a diagnostic classifica-
children’s social relations and their modula- tion system is to be effective in these varied
tion of anxiety. Her detailed descriptions of domains, the system must be clear, broadly ac-
atypical personality development are of con- cepted, and relatively easy to use. Diagnostic
tinuing interest in relation to the large number stability is an important goal; difficulties
of children with serious, early-onset distur- arise if diagnostic systems are changed too
bances in development who are not autistic. rapidly, for example, interpretation of previous
These ideas were developed by Provence in her research becomes a problem. A classification
studies of young children with atypical devel- system should provide descriptions that allow
opment (Provence & Dahl, 1987; see also disorders to be differentiated from one another
Chapter 6, this Handbook, this volume). in significant ways, for example, in course or
In the first (1952) and second (1968) edi- associated features (Rutter, 1996). Official
tions of the American Psychiatric Associa- classification systems must be applicable to
tion’s Diagnostic and Statistical Manuals only conditions that afflict individuals of both
the term childhood schizophrenia was offi- sexes and of different ages; at different devel-
cially available to describe autistic children. opmental levels; and from different ethnic, so-
Much of the early work on autism and related cial, and geographical backgrounds. Finally, a
conditions is, therefore, difficult to interpret system must be logically consistent and com-
because it is unclear exactly what was being prehensive (Rutter & Gould, 1985). Achieving
studied. As information on life course and these divergent goals is not always easy (Volk-
family history became available (Kolvin, mar & Schwab-Stone, 1996).
1971; Rutter, 1970), it became clear that The clinical provision of a diagnosis or mul-
autism could not simply be considered an early tiple diagnoses is only one part of the diagnos-
form of schizophrenia, that most autistic indi- tic process (Cohen, 1976). The diagnostic
viduals were retarded, that the final behavioral process provides a richer description of a child
expression of the autistic syndrome was poten- or adult as a full person; it includes a historical
tially the result of several factors, and that the account of the origins of the difficulties and
disorder was not the result of deviant parent- changes over time, along with other relevant
child interaction (Cantwell, Baker, & Rutter, information about the individual’s develop-
1979; DeMyer, Hingtgen, & Jackson, 1981). ment, life course, and social situation. The
These findings greatly influenced the inclu- diagnostic process highlights areas of compe-
sion of autism in the third edition of DSM tence, as well as difficulties and symptoms; it
(American Psychiatric Association, 1980), to notes the ways the individual has adapted; it
which we return later. describes previous treatments, available re-
sources, and other information that will allow
ISSUES IN CLASSIFICATION a fuller understanding of the individual and his
or her problems. Also, the diagnostic process
Systems for classification exist for many dif- may suggest or delineate biological, psycholog-
ferent reasons, but a fundamental purpose is to ical, and social factors that may have placed
enhance communication (Rutter, 2002). For the individual at risk, led to the disorder,
researchers, this is essential to achieve relia- changed its severity, or modified the symp-
bility and validity of findings from research toms and course. The result of the diagnostic
10 Diagnosis and Classification

process should be a rich formulation—an ac- may be tied to specific diagnostic categories
count that will be elaborated with new knowl- (Rutter & Schopler, 1992). Such an approach
edge, including the response of the individual tends, unfortunately, to emphasize the diag-
to intervention. It cannot be overemphasized nostic label, rather than the diagnostic process.
that while the diagnostic label or labels pro- On the other hand, if a governmental body
vide important and helpful information, they adopts a broad diagnostic concept, the avail-
do not substitute for a full and rich under- able resources may be diluted and individuals
standing of the individual’s strengths and most in need of intensive treatment may be de-
weaknesses and life circumstances. Thus, pro- prived while those with less clearly definable
grams should be designed around individuals service requirements are included in programs
rather than labels. (Rutter & Schopler, 1992).
A diagnostic formulation, based on an ex- There are many misconceptions about diag-
tended diagnostic process, is provisional and nosis and classification (see Rutter, 1996;
subject to change with new information and Volkmar & Schwab-Stone, 1996; Volkmar,
experience. In this sense, it is a continuing ac- Schwab-Stone, & First, 2002). For example,
tivity involving the individual, family, clini- DSM-IV and similar systems of classification
cians, and educators. The diagnostic process, are organized around dichotomous categories;
as a clinical activity, depends on a body of sci- in these systems, an individual either has or
entific knowledge and is enriched when there does not have a disorder. Yet, classification
is a common diagnostic language used for clin- can also be dimensional, in which an individ-
ical and research purposes. Information pro- ual has a problem, group of problems, or dys-
vided by this process is useful at the level of function to a certain degree. Dimensional
the individual case but also has important pub- approaches offer many advantages, as exempli-
lic health and social policy implications, for fied by the use of standard tests of intelli-
example, in formulating intervention strate- gence, adaptive behavior, or communication;
gies and allocating resources. in many ways, such approaches have domi-
Diagnostic systems lose value if they are nated in other branches of medicine and fre-
either overly broad or overly narrow. The clas- quently coexist with categorical ones (see
sification system must provide sufficient Rutter, 2002, for a review). Not only can the
detail to be used consistently and reliably by disease process (e.g., hypertension) be dimen-
clinicians and researchers across settings. sional but also various risk factors may be
When they achieve “official” status, as is the dimensional, and a dimensional focus has im-
case for ICD and DSM, classification schemes portant advantages for advancing knowledge in
have important regulatory and policy implica- this regard. On the other hand, at some point
tions. Sometimes, there may be conflicts be- qualitative and dimensional changes (as in
tween scientific and clinical needs, on one blood pressure) may lead either to functional
hand, and the impact of definitions on policy, impairment or specific symptoms (e.g., a high
on the other. For example, there may be good blood pressure can lead to angina), and the
scientific reasons for a narrowly defined cate- categorical approach is needed to address this
gorical diagnosis that includes only individuals important implication of what is basically a di-
who definitely and clearly have a specifically mensional phenomenon. Depression is a rele-
defined condition and excludes individuals vant example from psychiatry; for example, all
where there is less certainty. From the point of of us have the experience of mood fluctuations
view of service provision, however, broader di- during the course of our daily lives, but when
agnostic concepts may be most appropriate. depression becomes so significant that it be-
Unfortunately, there has often been a failure to gins to interfere with functioning or causes
recognize the validity of these two tensions impairment in other ways, we can consider use
around aspects of diagnosis. of specific treatments for depression.
Classification schemes of an “official” na- Dimensional and categorical classification
ture may have unintended, but important, im- systems are not incompatible. It is possible to
plications, for example, in terms of legal set a boundary point along a dimension that
mandates for services; this is particularly true can be used to define when a disorder is diag-
in the United States where federal regulations nosed. This boundary can be determined by
Issues in the Classification of Autism and Related Conditions 11

empirical studies that indicate that an important most important thing to convey about a phe-
threshold has been crossed that will in- nomenon or set of observations. Such notions
fluence functional status or impairment; or provide us with a sense of orderliness or narra-
the boundary can be defined by convention tive coherence. However, there is no truly
reached by clinicians, researchers, those who naive form of description or a naive descrip-
establish policy, or some combination of fac- tion of what clinicians and researchers mean
tors. For example, disorders such as depression by symptoms of a disorder. Even the decision
are readily amenable to dimensional defini- about what to consider a disorder of an individ-
tions. To some extent, all of us have experience ual presupposes a theory of what should be
of the symptoms of depression, yet, for the considered a disorder or dysfunction.
clinical syndrome of depression, a threshold The boundaries of the nosology for DSM-
must be surpassed: There must be a sufficient IV and ICD reflect a history of the profes-
number and range of symptoms that cause suf- sions of neurology, psychiatry, and general
fering, interfere with daily functioning, and medicine as well as preconceptions of where
persist (see Rutter, 2002; Chapter 28, this the current lines should be drawn. For exam-
Handbook, Volume 2). ple, the inclusion of Rett’s disorder in DSM-
For studies of autism and associated condi- IV raised the question of why a disorder with
tions, various dimensional approaches have such clear neurological aspects should be
been employed. Some instruments used for classified within the PDDs (Gillberg, 1994).
purposes of screening or diagnostic assessment However, neurological factors play a strong
focus on behaviors or historical features (or role in many disorders (including autism), but
both) that may be highly suggestive of a diag- that does not mean that they are only neuro-
nosis of autism. Such approaches have not logical. Much of the issue of where disorders
(with some notable exceptions—see Chapter such as autism or Rett’s are placed has to do
28, this Handbook, Volume 2) typically tried to with a practical issue of usage (see Rutter,
relate in a straightforward way with categori- 1994, for a discussion). A similar argument
cal approaches. Given the issues of focusing on could be had about Alzheimer’s disease,
highly unusual behaviors, other problems are which clearly falls within the professional
posed in the development and standardization purview of both psychiatrists and neurolo-
of such instruments. At the same time, such in- gists. One important effect of the decision to
struments have had a very significant role in include Rett’s disorder has been the ability to
research as well as clinical work, for example, focus specifically on this group in terms of
in screening for persons likely to have autism genetic mechanisms (see Chapter 5, this
(see Chapter 27, this Handbook, Volume 2). Handbook, this volume).
Another example of the dimensional ap- No nosology, including DSM-IV or ICD-10,
proach is embodied in the use of traditional can be totally free of theory, although there
tests of intelligence or communicative ability are good reasons for current psychiatric sys-
(see Chapters 29 & 30, this Handbook, Volume tems to aspire to be as atheoretical and de-
2). For such instruments, the provision of good scriptive as possible. This is illustrated in the
normative data is an important benefit. A earlier versions of DSM (American Psychiatric
growing body of work has focused on the di- Association, 1952, 1968) where theory was so
mensional metrification of social competence much part of definition that research work was
using the Vineland Adaptive Behavior Scales impeded. Theoretically oriented classification
(see Chapter 29, this Handbook, Volume 2). systems often are difficult to use since there
The role of theory in guiding development may be differences even among those who
of classification systems is a source of confu- share a theoretical perspective. Since 1980,
sion. Many assume that a classification system the trend in psychiatry has been toward de-
must be based on a theoretical model. To some scriptive, operational definitions that empha-
degree, all accounts of an event, process, clini- size observable behaviors and discrete clinical
cal set of findings, or disorder relate to a “ the- findings (Frances, Widiger, & Pincus, 1989);
ory” (or what more probably might be called a indeed, such an approach is represented, in
hypothesis or theory in the making). Such pro- many respects, by Kanner’s original descrip-
totheories focus on what to the viewer is the tion of autism. Such an approach to diagnosis
12 Diagnosis and Classification

is often called phenomenological although this findings during the next years may make this
term is confusing, since phenomenology is a more feasible in diagnosing and subtyping
branch of philosophy that concerns the under- autism.
lying structures of experience and the modes Like other human constructions, classifica-
of learning about mental and psychological tion systems can be misused (Hobbs, 1975).
phenomena (including the use of introspection One misuse is to confuse the person with the
and dense description). Phenomenology repre- diagnostic label. A person with a disorder is a
sents a theoretical approach to diagnosis that person first: An individual with autism is not
has an important history in psychology and psy- an “autistic.” A label does not capture the full-
chiatry. When contemporary researchers and ness of the person, nor his or her humanity.
clinicians speak of phenomenological systems, There is a risk that categorical terms may min-
they usually mean something quite different: imize the tremendous differences among per-
descriptions of the surface (signs and symp- sons who have a particular condition. The very
toms) or accounts of observable phenomena. In broad range of syndrome expression in autism
any event, DSM-IV and ICD-10 attempt to avoid requires the provision of multiple kinds of in-
all encompassing, grand theories of pathogene- formation in addition to the categorical diag-
sis and concepts that require adherence to a nosis, for example, level of communicative
particular viewpoint about the functioning of speech, intellectual abilities, interests, and ca-
the mind or the origins of psychopathology. In pacity for independent living.
this sense, they attempt to provide a relatively Another misuse of a categorical diagnosis
common language and framework that can be occurs when it is elevated to the status of
used by adherents of different theoretical being an explanation or when its use obscures
points of view. lack of knowledge. In Moliere’s plays, the
Another misunderstanding is that classifi- physician would mystify and impress the pa-
cation systems require etiologies and causes. tients with long Latin terms that were offered
Here, too, the trend within psychiatry has been as explanations but were merely redescriptions
toward systems that recognize that the causes of the patient’s symptoms. For many diag-
of most psychiatric, developmental, and emo- noses, this is still the case. For example, it is
tional disorders remain uncertain and complex helpful to parents to know that their 2-year-old
(Rutter, 1996). Also, there is a realization that child is not talking because he or she has a
many different causes may lead to the appar- disorder. However, it is different when this
ently very similar clinical condition while one disorder is deafness—which may explain the
specific cause may be associated with various muteness, at some interesting level of under-
conditions. Scientific studies will reveal new standing—than when the disorder is autism.
causes for old diseases, and there often are The diagnosis of autism clarifies some aspects
surprises as different underlying factors are of the nature of an individual child’s muteness
revealed for what has appeared to be a simple, by placing this child within a class of individu-
homogeneous clinical condition. The increas- als about whom a great deal of valuable infor-
ing knowledge and the disparity between mation about treatment and course has been
genotype (underlying cause) and phenotype learned. But the classification does not really
(clinical presentation) indicate the importance explain the language disorder any more than
of not basing a classification system only on the diagnosis of attention deficit / hyperactiv-
purported causes. However, as etiologies are ity disorder explains a child’s overactivity and
elucidated, it makes sense to consider includ- frustration intolerance. When a label is mis-
ing them within a diagnostic framework. In taken for an explanation, areas of ignorance
DSM-IV, a causal framework is most clear in may be covered over and the search for under-
the definition of posttraumatic stress disorder lying causes may end prematurely.
(PTSD), a condition in which a clear precipi- The final misuse of classification is the
tant (a traumatic experience) is related to a potential for stigmatization. Parents and
range of persistent symptoms. For autism, a advocates are anxious about the ways in
causal nosology is not yet available, although which classification may negatively skew how
genetic, neuroimaging, behavioral, or other the child or adult is seen by others or the
Issues in the Classification of Autism and Related Conditions 13

limitations and adversities that may follow diagnostic process. They may also be misused.
upon being labeled. Unfortunately, this danger However, they can be helpful in clarifying the
is real. When a child has been classified as nature of an individual’s difficulties and thus
mentally retarded or intellectually disabled, suggest care and indicate course.
this has sometimes meant removal from the
mainstream of education and a lifelong reduc- THE ROLE OF RESEARCH
tion of opportunity. The diagnosis of schizo-
phrenia has had negative connotations Initial descriptions of disorders such as autism
associated with madness and danger. Autism, and related conditions were invariably made
too, has had its social disadvantages; for exam- by a clinician-investigator who noticed some
ple, at one time it may have implied a particu- seeming element(s) of commonality among
lar view of etiology in which parents were children with very complex developmental
placed at fault. A diagnostic label may exclude difficulties. Although modifications in early
individuals from programs or reduce chances descriptions of these conditions have, not sur-
in purchasing insurance. For these reasons, prisingly, often been made over time, there
parents and advocates have sometimes felt that usually has been a fundamental continuity of
inclusion of autism as a mental disorder may basic aspects of definitions with the historical
imply that autism is the result of some type of definition. Over the past several decades, em-
emotional upset within the child or family— pirical research has assumed a progressively
when it clearly is not—or that it stigmatizes greater role in refining diagnostic criteria and
the child. Dealing with these issues is a contin- categories. In this regard, even when empirical
uing process, and there have been major ad- research suggests that some feature or features
vances in destigmatization over the past years. are central to the definition, these need not,
Public education, professional awareness of necessarily, have a central etiological role.
the potential abuse of diagnostic labels, and Conversely, features less critical for purposes
legal imperatives are all important in reducing of definition may have major importance for
prejudice against individuals with handicaps intervention. In autism, the unusual pattern of
and disabilities. These issues also have had social deficit originally described by Kanner
important implications for studies of epidemi- (1943) remains the central defining core of the
ology and service planning, particularly when condition (Klin, Jones, Schultz, & Volkmar,
the available data related to labels are used for 2003); stereotyped motor mannerisms, on the
educational or intervention purposes; in such other hand, do not as clearly separate autism
contexts, parents might, for example, chose to from other conditions with severe and pro-
utilize the term autism to entitle their child to found mental handicap (Volkmar, Klin, Siegel,
additional services even if full criteria for Szatmari, Lord, et al., 1994). Similarly, un-
autism are not met or when the child might just usual sensory experiences are commonly ob-
as readily receive another label for service served in individuals with autism; they, too,
provision (a problem referred to as diagnostic may be a focus of intervention, but they are not
substitution—see Chapter 2, this Handbook, a robust, defining feature of the condition (see
this volume). Conversely, the well-intentioned Chapter 32, this Handbook, Volume 2, and
attempt to destigmatize a child by describing Rogers & Ozonoff, in press, for reviews).
his or her disability simply as a different style Other symptoms may be highly predictive of
of learning or being has the potential to reduce the presence of autism, but they are of such
entitlements and services and opportunities low frequency that they are not included in
for the gains associated with treatment (Na- usual definitions. For example, a child’s un-
tional Research Council, 2001). usual attachment to a physical object—such as
In summary, categorical diagnoses organize a string or a frying pan—is highly suggestive
professional experience and data, promote of the diagnosis of autism, but this preoccupa-
communication, and facilitate the provision of tion is not included in official diagnostic crite-
suitable treatments and interventions. They are ria because the behavior is not invariably
always open to improvement. They derive their present and even when present tends to be ob-
full meaning within the context of a continuing served only in younger individuals.
14 Diagnosis and Classification

Developmental aspects of syndrome expres- child’s vulnerability to other difficulties

sion are particularly important in autism and (Rutter, Shaffer, & Shepherd, 1975). Multiax-
related conditions. A developmental approach ial systems help to ensure that in the search
to classification views specific behaviors for a single, encompassing, categorical diag-
within the context of normative development. nosis, the rich and multifaceted diagnostic
For example, the echolalia of autistic individu- process is not undervalued.
als is similar in some respects to the repeti-
tions observed in the speech of typically APPROACHES TO CATEGORICAL
developing 2- and 3-year-olds (see Chapter 30, DEFINITIONS OF AUTISM
this Handbook, Volume 2). From this perspec-
tive, echolalia is not simply a symptom but also In contrast to many conditions in child psychi-
is seen among typical children at a particular atry, strictly defined autism does not “shade
phase of development; when an older, mute, off ” into normalcy in the usual sense (Rutter
autistic child begins to use echolalia, it may be & Garmezy, 1983) and thus represents one of
a sign of progress in language development. On the more robust disorders for purposes of cate-
the other hand, as originally noted by Kanner, gorical diagnosis; at the same time, the body of
some aspects of the functioning of individuals genetic research has raised the important issue
with autism are fundamentally not develop- of a “ broader ” phenotype, that is, of a contin-
mentally appropriate at any age (see Chapters uum of social and related vulnerabilities
28, 30, & 32, this Handbook, Volume 2). This (Volkmar et al., 2004).
is specifically true of the social dysfunction Even for strictly defined autism, there are
and lack of engagement. Even infants are en- problems in the development of explicit defini-
gaged socially. The typical aloofness of autism tions. These include the tremendous range in
and lack of reciprocity are distinctly abnormal syndrome expression and change in symptoms
at any age and appear especially so when these over the course of development. Since the per-
social disabilities are far out of proportion to son with autism may not always be able to pro-
the individual’s functioning in other domains vide a direct, verbal report, the reports of
of daily living (see Chapter 11, this Handbook, parents or caregivers must be relied on, as
this volume). with very young children, raising other poten-
Behavioral deviance, such as lack of social tial problems including reliability and validity
reciprocity or abnormal preoccupations, is of historical information. Methods have been
often the focus of the criteria used in defining proposed for diagnosis that focus on very early
a categorical diagnosis. Such deviance is also development. These methods, which some-
a focus of rating scales and other assessment times use dimensional ratings scales (see
instruments used in relation to autism. This Chapter 28, this Handbook, Volume 2), may
diagnostic approach may be combined with an be problematic in relation to providing a cate-
assessment of how the individual compares to gorical diagnosis for an adolescent or adult
typical children and adults, for example, in re- with autism. In the absence of an accepted
lation to language use. The multiaxial system measure of diagnostic pathophysiology, one
of DSM-IV is an attempt to systematically would wish to consider both the historical in-
convey the value of considering an individual formation as well as course and current func-
from multiple perspectives. This includes tioning in conferring a diagnosis of a severe
assessment of the individual’s personality, developmental or psychiatric disorder. Yet,
educational and social resources, ongoing the use of development and history raises
stresses, medical problems and diseases, and practical problems for categorical diagnostic
adaptive functioning as well as impairment systems. In general, history has been over-
(Rutter & Schopler, 1992). Multiaxial diag- looked in the current official nosologies (with
nostic approaches are especially helpful in un- the exception of noting the age of onset)—a
derstanding individuals who have disorders topic to which we return later.
that start during childhood and are persistent, There are interesting and relevant ques-
like autism, and have major impact on all tions, too, about what should be included in a
spheres of development and increase the categorical diagnostic set of criteria. Should
Issues in the Classification of Autism and Related Conditions 15

such a set emphasize only those symptoms and DSM-III

signs that most clearly differentiate one condi-
tion from another, or should the set of criteria DSM-III (1980) was a landmark in the devel-
also include important symptoms (e.g., rushes opment of psychiatric taxonomy based on
of panic and anxiety or overactivity and im- research findings and emphasizing valid, reli-
pulsiveness) that are also found among other able descriptions of complex clinical phenom-
conditions? Should the criteria capture the ena. Autism was included along with several
largest number of children who may have the other disorders in a newly designated class
condition or be more selective? What about of childhood onset disorders, Pervasive Devel-
symptoms that may be infrequent but of great opmental Disorders (PDD). Other disorders
clinical importance when they occur, such as included residual infantile autism, child-
self-injurious behavior? To what degree should hood onset pervasive developmental disorder
diagnostic criteria also be fuller descriptions (COPDD), and residual COPDD. A subthresh-
of the condition? old condition was included as well, atypical
Investigators began to propose more ex- PDD. The class name pervasive developmental
plicit categorical definitions of autism in the disorder was newly coined and was meant to
1970s as a consensus on the validity of autism convey that individuals with these conditions
emerged. This was parallel to attempts in adult suffered from impairment in the development
psychiatry to provide better definitions of and unfolding of multiple areas of function-
psychiatric disorders for research purposes ing. The term also was meant to avoid a theo-
(Spitzer, Endicott, & Robins, 1978). The im- retical presupposition about etiology, and it
portance of a multiaxial or multidimensional quickly achieved broad acceptance. Subse-
approach to diagnosis became increasingly ap- quently, the choice of the term PDD has been
preciated (Rutter et al., 1975). Rutter (1978) debated (see Gillberg, 1991; Volkmar &
synthesized Kanner’s original report and Cohen, 1991b), and other terms, for example,
subsequent research in a highly influential def- autism spectrum disorder (ASD), have also
inition of autism as having four essential fea- come into common usage; the two terms are
tures: (1) early onset by age 22 years, (2) used synonymously here.
impaired social development, (3) impaired The DSM-III system was a major advance.
communication, and (4) unusual behaviors It extended official recognition to autism, dis-
consistent in many ways with Kanner’s con- carded the earlier presumption of a relation
cept of “insistence on sameness” (resistance to between autism and childhood schizophrenia,
change, idiosyncratic responses to the environ- and provided a useful definition largely re-
ment, motor mannerisms and stereotypes, flecting Rutter’s (1978) approach. The use of a
etc.). Rutter specified that the social and com- multiaxial system also facilitated research.
munication impairments were distinctive and However, some shortcomings with this system
not just a function of associated mental retar- were relatively quickly apparent. The rationale
dation. In contrast, the National Society for for the inclusion of COPDD was apparently to
Autistic Children (NSAC; Ritvo, 1978) in the account for those relatively rare children who
United States proposed a definition that in- developed an autistic-like disorder after age
cluded disturbances in (1) rates and sequences 30 months (Kolvin, 1971); this disorder was
of development, (2) responses to sensory stim- not, however, meant to be analogous with the
uli, (3) speech, language-cognition, and non- concept of Heller’s syndrome (disintegrative
verbal communication, and (4) the capacity to psychosis) since it was assumed (incorrectly)
relate appropriately to people, events, and ob- that the latter was invariably a function of
jects. This definition also emphasized the neu- some related general medical condition (Volk-
robiological basis of autism. While clinically mar, 1992). The definition of autism itself was
providing more detail, the Ritvo-NSAC defini- rather sparse and tended, perhaps not surpris-
tion proved rather less influential than the ingly given the official name of the disorder
Rutter synthesis, probably because the latter (infantile autism), to focus very much on
seemed conceptually clearer and closer to autism as it is exhibited in younger children.
Kanner’s original description. The use of the term residual autism was
16 Diagnosis and Classification

included to account for cases where the child tive impairment in verbal and nonverbal com-
once met the criteria for infantile autism but munication and in imagination, and restricted
no longer did so; this seemed, at some level, to repertoire of activities and interests.
imply that the individual no longer had autism. A small national field trial was conducted
The term atypical PDD was used for sub- to finalize scoring rules for the DSM-III-R
threshold conditions, that is, for a constella- definition of autism (Spitzer & Siegel, 1990).
tion of difficulties that appeared to most Sixteen proposed criteria for autistic disorder
appropriately be placed within the PDD class were grouped into the three broad categories.
but which did not meet criteria for infantile Based on this field trial, the diagnosis of
autism or another explicitly defined condition, autism required that an individual child or
unintentionally suggesting Rank’s earlier adult had to exhibit at least 8 of these 16 crite-
(1949) concept. Individuals with hallucina- ria, in total, with a specified distribution over
tions and delusions were specifically excluded the three areas of disturbance. This require-
from the PDD diagnoses. While it is unlikely ment for an early onset of the condition was
that many persons with autism will develop dropped in DSM-III-R because of the wish to
schizophrenia, it might be anticipated that in- provide a generally applicable criterion set, re-
dividuals with autism would develop schizo- gardless of age, and partly for the philosophi-
phrenia at least as often as other individuals in cal reason that the age of onset should not be
the general population, a hypothesis that considered a diagnostic feature, that is, that
seems to be sustained by available evidence clinicians should rely on present examination
(Volkmar & Cohen, 1991a). rather than history in making the diagnosis.
The multiaxial placement of disorders in This change would make it possible to diagnose
DSM-III also was a source of controversy; that autism in children who, for example, appeared
is, autism and other PDDs were placed on Axis to develop autism or something suggestive of it
I as was mental retardation although other spe- much later in development (Weir & Salisbury,
cific developmental disorders were listed on 1980); such cases have never, however, been
Axis II of the multiaxial system. The problems very common and it seemed problematic that
with DSM-III were widely recognized, and a their uniqueness was not flagged in some way
major revision was undertaken for DSM-III-R (e.g., through diagnostic coding).
(American Psychiatric Association, 1987). DSM III-R was attentive to changes in the
expression of autism with age and developmen-
DSM-III-R tal level. This represented a clear improvement
over DSM-III (Volkmar, Cicchetti, Cohen, &
Preparations for the revision of DSM-III began Bregman, 1992) where the concept of residual
soon after it appeared. What started as revi- autism had been an unsatisfactory attempt to
sion soon became a major renovation. Radical deal with this issue. Criteria in DSM-III-R
changes were introduced into the concept of were offered for autistic disorder and were ap-
autism in DSM-III-R (American Psychiatric plicable to the entire range of the expression of
Association, 1987; see Waterhouse, Wing, the syndrome. Thus, an individual could retain
Spitzer, & Siegel, 1993, for discussion of these the diagnosis of autism even if he or she was
changes). The rapid revision of the official functioning at a higher developmental level or
nosology posed problems for researchers who had experienced an amelioration of symptoms
were required to rediagnose their patients if with age, perhaps as a result of educational in-
they wished to remain au courant. tervention or maturation. The name of the
The definition of autistic disorder in DSM- condition was changed from infantile autism
III-R was more consistent with that of Wing to reflect these changes. Finally, in DSM-III-
(Wing & Gould, 1979) and others who advo- R, the problematic COPDD category was
cated a somewhat broader view of the diagnos- dropped, leaving those children who had car-
tic concept (see Chapter 21, this Handbook, ried this diagnosis suspended in limbo or, in
this volume). Three major domains of dysfunc- practice, placed within the PDD-not other-
tion were still included, with specific criteria wise specified (NOS) category. The term for
provided for each domain: qualitative impair- all subthreshold categories was changed to
ment in reciprocal social interaction, qualita- “Not otherwise specified” (NOS) throughout
Issues in the Classification of Autism and Related Conditions 17

DSM. Individuals with autism were no longer, ICD, which appeared in 1968 (see Rutter et al.,
by definition, excluded from also exhibiting 1975; Spitzer & Williams, 1980). At the same
schizophrenia. time, there was general agreement that future
The ambitious goal of a heuristic definition refinement would be needed and, over the next
in DSM-III-R was a conceptual advance over decade, a series of steps were undertaken to
DSM-III, but carried unforeseen consequences. improve the ICD system (Sartorius, 1988).
DSM-III-R criteria expanded the diagnostic One important aspect was the development of a
concept (Factor, Freeman, & Kardash, 1989; multiaxial system for the psychiatric disorders
Hertzig, Snow, New, & Shapiro, 1990; Szat- of childhood (Rutter et al., 1975). By 1978, the
mari, 1992a; Volkmar et al., 1992). The rate of ninth edition of ICD appeared and plans for a
false-positive cases (if clinician judgment is revision were put into place. The ICD-9 ac-
taken as the standard) diagnosed according to corded official recognition to infantile autism
DSM-III-R was nearly 40% (Rutter & as well as disintegrative psychosis (or what
Schopler, 1992; Spitzer & Siegel, 1990). This would now be termed childhood disintegrative
tendency to overdiagnose autism in more intel- disorders); both conditions were included in a
lectual handicapped individuals likely also had category of childhood psychotic conditions—a
the inadvertent effect of diverting clinical at- category that also included other specific psy-
tention from autism as it appeared in intellec- chotic conditions of childhood and unspecified
tually more able individuals. psychotic conditions. This approach reflected
Other problems with DSM-III-R also were the historical view (then beginning to change)
noted. First, the criteria set was more complex that autism represented one of the first mani-
and detailed, and the inclusion of specific ex- festations of childhood psychosis.
amples within the actual criteria seemed to The plan for revision of ICD-10 was well
limit clinician judgment. The elimination of underway at the time that DSM-IV was being
age of onset as a central diagnostic feature was developed. An important aspect of ICD-10 has
not consistent with Kanner’s original report been its conceptualization as a group of docu-
(1943) nor subsequent research that firmly es- ments written specifically for different users;
tablished that autism was an early-onset dis- for example, in contrast to the DSM-IV ap-
order (e.g., Harper & Williams, 1975; Kolvin, proach, research criteria for disorders are pro-
1971; Short & Schopler, 1988; Volkmar, vided separately from clinical guidelines for
Cohen, Hoshino, Rende, & Paul, 1988; Volk- primary health care providers. ICD-10 offers
mar, Stier, & Cohen, 1985). Probably the main comprehensive descriptions of clinical con-
issue with DSM-III-R, however, was the major cepts underlying the disorder, followed by
changes introduced in the diagnostic concept. points of differential diagnosis, and then pre-
These changes severely complicated the inter- sents the main symptoms that should be pres-
pretation of studies that used different diag- ent for a diagnosis. As a result, the ICD-10
nostic criteria. This issue was particularly system offers, in some important respects,
acute relative to the pending changes in the more flexibility to the clinician; this is partic-
classification of autism and similar conditions ularly valuable given the intended interna-
in the 10th edition of the ICD-10 (WHO, tional and cross-cultural use of the system.
1992), since it appeared that DSM-III-R
markedly overdiagnosed autism relative to the DSM-IV AND ICD-10
draft ICD-10 definition (Volkmar, Cicchetti,
Bregman, & Cohen, 1992). The process of revision in the ICD-10 was
closely related to the development of the
FROM ICD-9 TO ICD-10 DSM-IV (American Psychiatric Association,
1994). The International (ICD) and American
Since it was first introduced toward the end of (DSM) systems are fundamentally related, and
the nineteenth century, the ICD has under- by formal agreements must share, to some de-
gone many revisions (Kramer, 1968). The gree, a common approach to diagnostic coding.
limitations of the psychiatric section were There are, however, important general and
increasingly recognized, and extensive revi- specific differences between the two major di-
sion was undertaken in the eighth edition of agnostic systems (Volkmar & Schwab-Stone,
18 Diagnosis and Classification

1996). For example, the ICD-10 system high- PDD class; there was also agreement about the
lighted the importance of an individual’s his- desirability of compatibility of DSM-IV and
tory in making a diagnosis while DSM-III-R ICD-10 (Rutter & Schopler, 1992).
relied on contemporaneous examination. Also In addition to these literature reviews, a
in contrast to DSM-IV, ICD-10 was specifi- series of data reanalyses were undertaken
cally designed to have one set of research diag- with regard to autism. These reanalyses used
nostic criteria and a separate set of clinical previously collected data and indicated that
guidelines. The American and International the DSM-III-R definition of autistic disorder
approaches would probably have resulted in was overly broad (Volkmar, Cicchetti, &
very different patterns of diagnosis. Bregman, 1992). Several issues were identi-
Preparations for the creation of the new, fied during this process of analysis of the lit-
fourth edition of DSM began very shortly after erature and of available data that needed
DSM-III-R appeared, partly due to the pending clarification for DSM-IV, including issues of
changes in the ICD-10. As part of the revision overdiagnosis in the more intellectually
process, work groups reviewed the current challenged and underdiagnosis in more able
classification systems in light of existing re- individuals. Consistent with the empirical
search and identified areas both of consensus principles guiding the creation of DSM-IV,
and controversy. They considered various is- the working group decided that the clarifica-
sues, including clinical utility, reliability, and tion of these and other issues would be based
descriptive validity of categories and criteria on the findings from a large, multinational
as well as coordination with the ICD-10 revi- field trial (Volkmar, Klin, Siegel, Szatmari,
sion (Frances et al., 1991). As part of the pro- Lord, et al., 1994).
cess of creating DSM-IV, clinical investigators
conducted literature reviews for each of the DSM-IV Field Trial
potential diagnostic categories. These reviews
were particularly helpful for some of the new As part of the DSM-IV field trial for autism,
diagnostic categories. For example, although 21 sites and 125 raters participated from the
childhood disintegrative disorder (Heller’s United States and around the world. By de-
syndrome) is apparently much less common sign, the raters had a range of experience in
than autism, the data supported the view that it the diagnosis of autism and a range of profes-
differed from autism in a number of important sional backgrounds. The field trial included
ways (Volkmar, 1992; Volkmar & Cohen, information on nearly 1,000 cases seen by
1989). Asperger’s disorder was included in one or more raters. In cases where the same
ICD-10, but the text indicated that the validity case was rated by multiple raters to assess re-
of the syndrome as a disorder, distinct from liability, the rating by one clinician was cho-
autism, was not yet fully established (Rutter & sen at random to be included in the main
Schopler, 1992; Szatmari, 1992a, 1992b). The database. The preference for the entire field
absence of official or other generally agreed trial was for cases rated on the basis of con-
upon definitions for Asperger’s disorder had temporaneous examination and not just on re-
contributed to markedly different uses of the view of records. By design, five contributing
term in clinical and research work (see Chap- sites provided ratings on approximately 100
ter 4, this Handbook, this volume). With Rett’s consecutive cases of individuals either with
disorder, the issues revealed by the review pro- autism or other disorders in which the diagno-
cess had less to do with the validity of the di- sis of autism would reasonably be included in
agnostic concept and more with the question of the differential diagnosis while the other 16
whether Rett’s should be included in the PDD sites provided ratings of a minimum of about
class rather than as a neurological disorder 20 cases. Cases were included only if it ap-
(Gillberg, 1994; Rutter, 1994; Tsai, 1992). Al- peared that the case exhibited difficulties
though the literature identified major gaps in that would reasonably include autism in the
knowledge and persistent issues, the consensus differential diagnosis. The availability of
of workers in the field favored the inclusion of clinical ratings of cases seen at clinical cen-
additional diagnostic categories within the ters around the world was of interest in terms
Issues in the Classification of Autism and Related Conditions 19

TABLE 1.2 DSM-IV Autistic Disorder Field Trial

Group Characteristics
Clinically Autistic Other PDDs Non-PDD
(N = 454) (N = 240) (N = 283)
Sex Ratio (MF) 4.491 3.711 2.291
Mute 54% 35% 33%
Age 8.99 9.68 9.72
IQ 58.1 77.2 66.9
Notes: Cases grouped by clinical diagnosis. Diagnoses of the “other PDD”
cases included: Rett syndrome (13 cases), childhood disintegrative disorder
(16 cases), Asperger syndrome (48 cases), PPD-NOS (116 cases), and atypical
autism (47 cases). Diagnoses of the non-PDD cases included mental retardation
(132 cases), language disorder (88 cases), childhood schizophrenia (9 cases),
other disorders (54 cases).

of issues of compatibility between DSM-IV designed to be a research diagnostic system,

and ICD-10. Characteristics of the field trial had, as expected, higher specificity.
sample are presented in Table 1.2. As mentioned earlier, one of the major dif-
Typically, multiple sources of information ferences between DSM-III-R and both DSM-III
were available to the rater, and the quality of and ICD-10 was the failure to include history
the information available to the rater was in the diagnostic process, for example, early
judged to be excellent or good in about 75% of age of onset as an explicit diagnostic feature.
cases. Individuals from a variety of ethnic Reported age of onset of autism was examined.
backgrounds and in various educational set- The mean reported age at onset for autism was
tings were included. This approach differed in early. The data on reported age of onset are
important respects from that employed in presented in Figure 1.1.
DSM-III-R where, for example, children with Age at onset had a modest, positive rela-
conduct disorders (without development disor- tionship with measured intelligence. Individu-
der) were included in the comparison group. als with slightly later onset were more likely
A standard system of coding was used to to have higher IQ scores. If onset by 36
elicit information on basic characteristics of months was added as an essential feature to
the case (age, IQ, communicative ability, edu- DSM-III-R, the sensitivity of that system was
cational placement), the rater, and various di- increased. Thus, inclusion of age of onset as
agnostic criteria. The coding form also
provided possible criteria for Asperger’s dis-
order, Rett’s disorder, and childhood disinte- TABLE 1.3 Table IV-2: Sensitivity
grative disorder, based on the draft ICD-10 (Se)/Specificity (Sp) by IQ Level
definitions. DSM-IIIa DSM-III-R ICD-10b
The field trial provided data for studying
By IQ Level N Se Sp Se Sp Se Sp
the patterns of agreement among the various
diagnostic systems. These results are pre- <25 64 .90 .76 .84 .39 .74 .88
sented in Table 1.3. As shown, the DSM-III 25–39 148 .88 .76 .90 .60 .88 .92
40–54 191 .79 .76 .93 .74 .84 .83
diagnoses of infantile autism and residual 55– 69 167 .86 .78 .84 .77 .78 .89
autism had a reasonable balance of sensitivity 70–85 152 .79 .81 .88 .81 .74 .96
and specificity; the use of the residual autism >85 218 .78 .83 .78 .78 .78 .91
category in DSM-III was associated with Overall .82 .80 .86 .83 .79 .89
other problems. In contrast, DSM-III-R crite- a
“Lifetime” diagnosis (current IA or “residual” IA).
ria had a higher sensitivity but lower speci- b
Original ICD-10 criteria and scoring.
ficity and a relatively high rate of Adapted from “Field Trial for Autistic Disorder
false-positive cases, especially among indi- in DSM-IV,” by F. R. Volkmar et al., 1994, American
viduals with retardation where the rate Journal of Psychiatry, 151, 1361–1367. Used with
reached 60%. The ICD-10 draft definition, permission.
20 Diagnosis and Classification

120 individuals who were assigned a diagnosis of

autism only by DSM-III-R. This instability of
100 diagnostic classification was most apparent for
younger children and for individuals with
Number of Cases

80 lower IQ.
The field trial data were also analyzed
using signal detection methods and principal
components analyses. The various approaches
to the data suggested that certain items could
be eliminated from the ICD-10 definition,
particularly items with low base rates or
strong developmental associations (see later
6 12 18 24 30 36 42 48 54 60 discussion). Before final decisions could be
Report Onset (months) made on the DSM-IV definition, it was neces-
Figure 1.1 Age of onset: Cases with clinical diag-
sary to address the broader issue of whether
nosis of autism. other explicitly defined disorders would be in-
cluded in the PDD class in DSM-IV. While the
DSM-IV autism field trial was not primarily
focused on the definition (much less the valid-
an essential diagnostic feature for autism was ity) of these conditions, the issues of the defi-
supported and was consistent with the ICD-10 nition and validity were relevant to the
draft criteria. DSM-IV and ICD-10 definitions of autism.
Aspects of the reliability of criteria and of The boundaries for autism and the nonautistic
diagnoses made by the various diagnostic sys- PDD were mutually related: A narrow defini-
tems were examined using chance corrected tion of autism would force some cases into the
statistics. Since raters with a range of experi- nonautistic PDD group. The broad definition
ence had participated in the field trial, it was of autism in DSM-III-R had certain advan-
possible to address rater experience in relation tages, for example, in ensuring access to ser-
to reliability. In general, the interrater reliabil- vices; but a narrower definition might be
ity of individual diagnostic criteria was in the important for research studies that require
good to excellent range. Only one criterion had greater homogeneity.
poor interrater reliability. Typically, the more
Definition of Autism in DSM-IV and ICD-10
detailed ICD-10 criteria had, as expected,
greater reliability. Also as expected, experi- The field trial data provided an important em-
enced evaluators usually had excellent agree- pirical basis for constructing the definition of
ment among themselves and were more likely autism for DSM-IV. The data showed that the
to agree with one another than with less expe- DSM-III-R definition could be substantially
rienced raters. The experience of the raters improved by addition of a criterion relating to
rather than their professional discipline had age of onset and by raising the diagnostic
the greatest impact on reliability (Klin, Lang, threshold. Similarly, various combinations of
Cicchetti, & Volkmar, 2000). DSM-III, DSM-III-R, and new criteria all
The temporal stability of ratings was as- could have been used to provide a reasonably
sessed in two ways. A small number of cases balanced diagnostic system. Given the concern
for test-retest reliability were collected as part about the importance of compatibility with
of the field trial; in addition, follow-up infor- ICD-10 and the implications for research of a
mation was available on the cohort of 114 universally accepted definition, the working
cases originally reported earlier (Volkmar, group of DSM-IV considered the benefits of
Bregman, Cohen, & Cicchetti, 1988). Criteria the ICD-10 system. Possible modifications in
and diagnostic assignments were highly stable the ICD-10 system were examined. The goal
over relatively short periods of time in the was to establish a definition for DSM-IV that
range of less than one year. Findings with the balanced clinical and research needs, was rea-
cases followed up by Volkmar et al. (1988) sonably concise and easy to use, provided rea-
suggested more diagnostic instability for those sonable coverage over the range of syndrome
Issues in the Classification of Autism and Related Conditions 21

expression in autism, and was applicable over gesture. The domain of restricted patterns of
the full life span, from early childhood behavior, interests, and activities includes en-
through adulthood. compassing preoccupations that are abnormal
Of the original 20 ICD-10 criteria, four either in focus or intensity, adherence to non-
were identified for possible elimination. Alter- functional routines or rituals, stereotyped
natives to specific criteria were examined, and motor movements, and persistent preoccupa-
a modified definition was developed. This tion with parts of objects.
modified definition worked well both overall
and over different levels of age and associated The Definition of the Nonautistic PDDs
mental retardation; it also could be readily
used by less experienced examiners. In contrast to DSM-III-R, a number of condi-
Diagnostic criteria for autism in DSM-IV tions other than autism and subthreshold
and ICD-10 are presented in Table 1.4. autism (i.e., PDD-NOS) are now officially
For the diagnosis of autism, at least six cri- recognized in both DSM-IV and ICD-10.
teria must be exhibited, including at least two Given that these are newer disorders (at least
criteria relating to social abnormalities (group in terms of their official recognition), it is not
one) and one each relating to impaired commu- surprising that the substantive body of work
nication (group 2) and range of interests and on their definitions is less extensive than that
activities (group 3). In addition, the onset of for autism.
the condition must have been prior to age 3
Rett’s Disorder
years as evidenced by delay or abnormal func-
tioning in social interaction, language as used There were few concerns about the validity of
in social interaction, and symbolic/imagina- the entity explicated by Rett. It was clear that
tive play. In addition, DSM-IV accepted the di- the transient, autistic-like phase of social with-
agnostic convention that the disorder could not drawal occurred early in the child’s develop-
better be accounted for by the diagnosis of ment and presented the primary problem for
Rett’s disorder or childhood disintegrative dis- differentiation from autism (and one of the
order (the definitions of these concepts are main arguments for its placement in the PDD
discussed subsequently). class). However, there were some objections to
Qualitative impairment in social interaction including it in the PDD class (Gillberg, 1994)
can take the form of markedly impaired non- although it was also clear that it should be in-
verbal behaviors, failure in developmentally cluded somewhere (Rutter, 1994). The impor-
expectable peer relationships, lack of shared tance of its inclusion has been underscored by
enjoyment or pleasure, or lack of social- the subsequent discovery of a gene involved in
emotional reciprocity. The stronger weighting the pathogenesis of the disorder (Amir, Van den
of the impairments in socialization was noted Veyver, Wan, Tran, Francke, et al., 1999; also
during the field trial to be important in avoid- see Chapter 5, this Handbook, this volume).
ing overdiagnosis of autism in more intellec-
Childhood Disintegrative Disorder
tually handicapped persons. This is also
consistent with extensive previous clinical Although this condition had been included in
work, from the time of Kanner onward (e.g., ICD-9 the presumption in DSM-III-R was that
Rutter, 1978; D. Cohen, 1980; Siegel, Vukice- individuals with childhood disintegrative dis-
vic, Elliott, & Kraemer, 1989) that high- order (also known as Heller’s syndrome or
lighted social dysfunction as the critical disintegrative psychosis) usually suffered
domain of impairment in autism. from a neurological or other progressive pro-
Impairments in communication can take cess that accounted for their marked behav-
the form of delay or lack of spoken language, ioral and developmental deterioration. The
impairment in conversational ability, stereo- literature, however, did not support this associ-
typed language use, and deficits in imagina- ation (Volkmar, 1992). While rare, childhood
tive play. For persons with autism, the delay disintegrative disorder appeared to be a disor-
or lack of spoken language must not be ac- der that could be distinguished from autism
companied by compensations through other and that was, like autism, of generally un-
communicative means, for example, the use of known etiology. The rationale for including
TABLE 1.4 ICD-10 Criteria for Autism
Childhood Autism (F84.0)
A. Abnormal or impaired development is evident before the age of 3 years in at least one of the following areas:
(1) receptive or expressive language as used in social communication;
(2) the development of selective social attachments or of reciprocal social interaction;
(3) functional or symbolic play.
B. A total of at least six symptoms from (1), (2) and (3) must be present, with at least two from (1) and at least
one from each of (2) and (3)
(1) Qualitative impairment in social interaction are manifest in at least two of the following areas:
(a) failure adequately to use eye-to-eye gaze, facial expression, body postures, and gestures to regulate
social interaction;
( b) failure to develop (in a manner appropriate to mental age, and despite ample opportunities) peer
relationships that involve a mutual sharing of interests, activities and emotions;
(c) lack of socio-emotional reciprocity as shown by an impaired or deviant response to other people’s
emotions; or lack of modulation of behaviour according to social context; or a weak integration of
social, emotional, and communicative behaviors;
(d) lack of spontaneous seeking to share enjoyment, interests, or achievements with other people (e.g., a
lack of showing, bringing, or point out to other people objects of interest to the individual).
(2) Qualitative abnormalities communication as manifest in at least one of the following areas:
(a) delay in or total lack of, development of spoken language that is not accompanied by an attempt to
compensate through the use of gestures or mime as an alternative mode of communication (often
preceded by a lack of communicative babbling);
( b) relative failure to initiate or sustain conversational interchange (at whatever level of language skill
is present), in which there is reciprocal responsiveness to the communications of the other person;
(c) stereotyped and repetitive use of language or idiosyncratic use of words or phrases;
(d) lack of varied spontaneous make-believe play or (when young) social imitative play.
(3) Restricted, repetitive, and stereotyped patterns of behaviour, interests, and activities are manifested in
at least one of the following:
(a) an encompassing preoccupation with one or more stereotyped and restricted patterns of interest that
are abnormal in content or focus; or one or more interests that are abnormal in their intensity and
circumscribed nature though not in their content or focus;
( b) apparently compulsive adherence to specific, nonfunctional routines or rituals;
(c) stereotyped and repetitive motor mannerisms that involve either hand or finger f lapping or twisting
or complex whole body movements;
(d) preoccupations with part-objects or non-functional elements of play materials (such as their odour,
the feel of their surface, or the noise or vibration they generate).
C. The clinical picture is not attributable to the other varieties of pervasive developmental disorders; specific
development disorder of receptive language (F80.2) with secondary socio-emotional problems’ reactive
attachment disorder (F94.1) or disinhibited attachment disorder (F94.2); mental retardation (F70-F72) with
some associated emotional or behavioral disorders; schizophrenia (F20.-) of unusually early Onset; and Rett’s
syndrome (F84.12).

F84.1 Atypical autism

A. Abnormal or impaired development is evident at or after the age of 3 years (criteria as for autism except for
age of manifestation).
B. There are qualitative abnormalities in reciprocal social interaction or in communication, or restricted,
repetitive, and stereotyped patterns of behavior, interests, and activities. (Criteria as for autism except that it
is unnecessary to meet the criteria for number of areas of abnormality.)
C. The disorder does not meet the diagnostic criteria for autism (F84.0).
Autism may be atypical in either age of onset (F84.10) or symptomatology (F84.11); the two types are
differentiated with a fifth character for research purposes. Syndromes that are typical in both respects should
be coded F84.12.

F84.10 Atypicality in age of onset

A. The disorder does not meet criterion A for autism (F84.0); that is, abnormal or impaired development is
evident only at or after age 3 years.
B. The disorder meets criteria B and C for autism (F84.0).

Issues in the Classification of Autism and Related Conditions 23

TABLE 1.4 (Continued)

F84.11 Atypicality in symptomatology

A. The disorder meets criterion A for autism (F84.0); that is abnormal or impaired development is evident
before age 3 years.
B. There are qualitative abnormalities in reciprocal social interactions or in communication, or restricted,
repetitive, and stereotyped patterns of behavior, interests, and activities. (Criteria as for autism except that
it is unnecessary to meet the criteria for number of areas of abnormality.)
C. The disorder meets criterion C for autism (F84.0).
D. The disorder does not fully meet criterion B for autism (F84.0).

F84.12 Atypicality in both age of onset and symptomatology

A. The disorder does not meet criterion A for autism (F84.0); that is, abnormal or impaired development is
evident only at or after age 3 years.
B. There are qualitative abnormalities in reciprocal social interactions or in communication, or restricted,
repetitive, and stereotyped patterns of behavior, interests, and activities. (Criteria as for autism except that
it is unnecessary to meet the criteria for number of areas of abnormality.)
C. The disorder meets criterion C for autism (F84.0).
D. The disorder does not fully meet criterion B for autism (F84.0).

DSM-IV Criteria for Autistic Disorder (299.0)

A. A total of at least six items from (1), (2), and (3), with at least two from (1), and one each from (2) and (3):
(1) Qualitative impairment in social interaction, as manifested by at least two of the following:
(a) marked impairment in the use of multiple nonverbal behaviors such as eye-to-eye gaze, facial
expression, body postures, and gestures to regulate social interaction,
( b) failure to develop peer relationships appropriate to developmental level,
(c) markedly impaired expression of pleasure in other people’s happiness,
(d) lack of social or emotional reciprocity,
(2) Qualitative impairments in communication as manifested by at least one of the following:
(a) delay in or total lack of, the development of spoken language (not accompanied by an attempt to
compensate through alternative modes of communication such as gestures or mime)
( b) in individuals with adequate speech, marked impairment in the ability to initiate or sustain a
conversation with others
(c) stereotyped and repetitive use of language or idiosyncratic language
(d) lack of varied spontaneous make-believe play or social imitative play appropriate to developmental
(3) Restricted repetitive and stereotyped patterns of behavior, interests, and activities, as manifested by at
least one of the following:
(a) encompassing preoccupation with one or more stereotyped and restricted patterns of interest that is
abnormal either in intensity or focus
( b) apparently compulsive adherence to specific, nonfunctional routines or rituals
(c) stereotyped and repetitive motor mannerisms (e.g., hand or finger f lapping or twisting, or complex
whole body movements)
(d) persistent preoccupation with parts of objects
B. Delays or abnormal functioning in at least one of the following areas, with onset prior to age three: (1) social
interaction, (2) language as used in social communication, or (3) symbolic or imaginative play.
C. Not better accounted for by Rett’s Disorder or Childhood Disintegrative Disorder.
Sources: From Diagnostic and Statistical Manual of Mental Disorders, fourth edition, by American Psychiatric As-
sociation, 1994, Washington, DC: Author; and International Classification of Diseases: Diagnostic Criteria for Re-
search, tenth edition, by the World Health Organization, 1992, Geneva, Switzerland: Author. Reprinted with

this condition had less to do with its potential course and prognosis (Volkmar & Rutter,
importance for research, for example, relative 1995) although others (e.g., Hendry, 2000)
to the search for a gene or genes that might be have questioned the recognition of the cate-
involved, than its frequency. The limited data gory; these issues are discussed in greater de-
available also suggested some important po- tail in Chapter 3, this Handbook, of this
tential differences from autism in terms of volume.
24 Diagnosis and Classification

not required. In the DSM-IV field trial, the

Asperger’s Disorder
presence of such interest was one of the fea-
In many ways, the inclusion and definition of tures that discriminated individuals with clini-
this condition have been the source of the cal diagnoses of autism from Asperger’s
greatest continuing confusion and controversy disorder. The limited available data (see Chap-
(e.g., Klin, Sparrow, & Volkmar, 1997; see ter 4, this Handbook, this volume, and Klin
also Chapter 4, this Handbook, this volume). et al., in press) suggest, not surprisingly, rather
Although Asperger’s original paper (Asperger, poor overall agreement of these different diag-
1944) and his subsequent clinical work (Hip- nostic approaches.
pler & Klicpera, 2003) emphasized the pres- To some extent, these disparities in diag-
ence of circumscribed interests and motor nostic approach parallel broader differences in
delays, they were technically not required in the way the disorder is conceptualized. For ex-
either the ICD-10 or DSM-IV definition that ample, is Asperger best thought of as a milder
was eventually adopted. Indeed, in DSM-IV, it form of autism (Leekam et al., 2000), is it
was emphasized that autism should take diag- characterized by a rather different neuropsy-
nostic precedence; difficulties in the use of chological profile than autism (Klin, Volkmar,
these criteria were quickly noted (Miller & Sparrow, Cicchetti, & Rourke, 1995), or are
Ozonoff, 1997, 2000). As a result, final clo- the social difficulties different from autism
sure on the best definition of this disorder has (Tsai, 1992)? Yet another issue is how and
not yet been achieved. whether motor skills problems are taken into
Given the general dissatisfaction with the account (Ghaziuddin & Butler, 1998) or
definition of Asperger’s disorder (see Chapter whether some other feature, for example,
4, this Handbook, this volume), the unfortunate prosody, might differentiate autism and As-
problem of markedly different approaches to the perger’s disorder (Ghaziuddin & Gerstein,
definition of the disorder has continued compli- 1996). Perhaps the one thing that can be said
cating comparisons of results across studies. with certainty about current diagnostic ap-
There are now a least five rather different proaches is that there is general agreement that
conceptualizations of Asperger’s disorder in the current official approach (as in DSM-IV and
addition to those provided by ICD-10 and ICD-10) has not been easy to operationalize
DSM-IV (Ghaziuddin, Tsai, & Ghaziuddin, and has not proven useful for research. Miller
1992; Klin & Volkmar, 1997; Leekam, Libby, and Ozonoff (1997) have raised the cogent
Wing, Gould, & Gillberg, 2000; Szatmari, point that Asperger’s own cases likely would
Bryson, Boyle, Streiner, & Duku, 2003; Tsai, not meet current official criteria for the disor-
1992; Wing, 1981). Unfortunately, these defi- der; a recent report (Hipller & Klicpera, 2003)
nitions are not always easy to operationalize. of cases seen by Asperger may help inform the
Several major sources of disagreement are ap- current debate (see also Eisenmajer et al., 1996;
parent. The first issue has to do with the Howlin, 2003; Szatmari et al., 2003).
precedence rule, which (in DSM-IV and ICD- It must, however, also be noted that even
10) excludes an individual from Asperger’s if given the lack of general agreement on a gen-
the person ever met the criteria for autism. (As eral diagnostic approach, emerging data are
a practical matter, this ends up, largely, revolv- beginning to suggest some important potential
ing around the age at which parents were first differences between Asperger’s and higher
concerned about the child’s development.) The functioning autism, for example, in terms of
second issue concerns the approach to lan- neuropsychological profiles (Klin et al., 1995;
guage delay (usually operationalized by Lincoln, Courchesne, Kilman, Elmasian, &
whether the child spontaneously used mean- Allen, 1998), comorbidity with other psychi-
ingful words by 24 months and phrases by 36 atric disorders (Klin et al., in press), neu-
months; Howlin, 2003; Klin, Schultz, Pauls, & ropsychological profiles and family genetics
Volkmar, in press). A third major issue has to (Volkmar & Klin, 1998) and outcome (Szat-
do with whether the unusual circumscribed in- mari et al., 2003). The critical issue is whether
terests originally described by Asperger Asperger’s can be shown to differ in important
(1944) must be present for diagnosis; in DSM- respects from either autism or PDD-NOS on
IV and ICD-10, these may be present but are measures other than those used in selecting
Issues in the Classification of Autism and Related Conditions 25

cases in the first place; that is, information on has been increasingly recognized in research
the validity of the disorder is needed in areas studies (Bailey et al., 1998). DSM-IV and
such as differences in patterns of comorbidity, ICD-10 take slightly different approaches to
outcome, response to treatment, family history, this category with ICD-10 providing the possi-
or neuropsychological profiles. The relation- bility for more fine-grained distinctions based
ship of Asperger’s disorder to various other on the way in which full criteria for autism or
diagnostic concepts—for example, schizoid another of the explicitly defined PDDs are
disorder, right hemisphere learning disability, not met. An unfortunate editorial change in
and semantic pragmatic processing disorder— DSM-IV produced some difficulties, which
remains an important topic for research (see have now been rectified in DSM-IV-TR.
Klin, Volkmar, & Sparrow, 2000 for a review). Specifically, prior to DSM-IV, an individual
Replication of findings based on the same di- had to have problems in social interaction and
agnostic criteria used across sites is critical in communication or restricted interests. In
for progress to be made in this area. Until the DSM-IV, this criterion was changed leading
time when a consensus on the definition of the to an unintended further broadening of the
condition emerges, it will be critical for re- concept.
searchers to employ very clear, operational de- Table 1.5 provides a concise summary and
pictions to allow for replication of findings. comparison of the various disorders presently
included within the overarching PDD category.
Atypical Autism/PDD-NOS
Somewhat paradoxically, studies of what is un- CURRENT CONTROVERSIES
doubtedly the more frequent of the PDDs are IN DIAGNOSIS
uncommon (see Chapter 6, this Handbook, this
volume). This subthreshold category receives Although considerable progress has been made
considerable clinical use, and its importance further work is needed in several areas.

TABLE 1.5 Dif ferential Diagnostic Features of Autism and Nonautistic Pervasive
Developmental Disorders

Childhood Pervasive
Autistic Disintegrative Developmental
Feature Disorder Asperger’s Rett’s Disorder Disorder-NOS
Age at recognition 0–36 Usually >36 5–30 >24 Variable
Sex ratio M>F M>F F (?M) M>F M>F
Loss of skills Variable Usually not Marked Marked Usually not
Social skills Very poor Poor Varies with age Very poor Variable
Communication skills Usually poor Fair Very poor Very poor Fair to good
Circumscribed Variable Marked NA NA Variable
interests (mechanical) (facts)
Family history— Sometimes Frequent Not usually No Unknown
similar problems
Seizure disorder Common Uncommon Frequent Common Uncommon
Head growth No No Yes No No
IQ range Severe MR Mild MR to Severe MR Severe MR Severe MR to
to normal normal normal
Outcome Poor to good Fair to good Very poor Very poor Fair to good
Adapted from “Nonautistic Pervasive Developmental Disorders,” chap. 27.2, p. 4, by F. R. Volkmar & D. Cohen, in
Psychiatry, R. Michaels et al., eds. Used with permission from Lippincott-Raven Publishers. NA = Not Applicable.
26 Diagnosis and Classification

Comorbid Conditions and Autism diagnosis in DSM-IV and ICD-10. Both sys-
tems are meant to be comprehensive in cover-
The issue of comorbidity with autism has as- age. However, any system that attempts to
sumed increasing importance in recent years; move past the level of symptom description
it is intimately related to the search for sub- must deal with complicated problems of ensur-
groups of autism. It appears likely that having ing clinical utility, reliability, and validity. As
any serious disability—such as autism or intel- a practical matter, this leads to decisions,
lectual disability—increases the risk for other sometimes fairly obvious and sometimes much
problems, and it is likely that, in the past, less so, about relationships between cate-
autism has tended to overshadow the presence gories, including whether one condition takes
of other difficulties (see Dykens, 2000). precedence over another in a diagnostic hierar-
Autism has now been reported to co-occur chy. The ICD-10 system reflects a nosological
with various other developmental, psychiatric, tradition of searching for a single, parsimo-
and medical conditions (Gillberg & Coleman, nious diagnostic label to explain a patient’s
2000). However, much of this literature rests problems. This top-down approach tends to be
on case reports, and this literature fails to ad- concerned with broader, heuristic diagnoses
dress the more central question of whether as- and is less focused on symptoms as such. On
sociations are observed at greater than chance the other hand, DSM-IV and its immediate pre-
levels and, when this is done, results are gener- decessors have tended to be more bottom up in
ally much less striking (Rutter et al., 1994). orientation. They start with symptoms and
An additional problem is that only positive as- move toward broader categories. No single di-
sociations are typically reported; for example, agnosis is expected to convey the entire range
it is somewhat surprising that failure to thrive of a patient’s major problems, and there is
in infancy is so uncommonly reported in in- more comfort with multiple categorical diag-
fants who go on to have autism. noses, each covering a smaller domain of dif-
Evolving diagnostic concepts and research ficulties. In other words, ICD may miss some
findings have sometimes clarified such associ- trees, and DSM may not capture the forest:
ations. For example, Kanner’s original impres- Each approach has inherent advantages and
sion (1943) that persons with autism had limitations (see Volkmar & Schwab-Stone,
normal intellectual potential has been shown 1996). The DSM-IV approach has some advan-
to be incorrect; although the pattern of cogni- tages for clinical utility; that is, important
tive and adaptive abilities in autism is unusual, symptoms are less likely to be overlooked. It
for the majority of children with autism, over- also does not prejudge the issue of comorbid
all scores on cognitive testing are stable relationships. The ICD-10 approach has the
within the mentally retarded range (see Chap- advantage of providing a more robust big pic-
ter 29, this Handbook, Volume 2). On the other ture less focused on single symptoms and
hand, a substantial minority of persons with minimizing what are often spurious or mean-
autism has cognitive abilities in the average or ingless associations.
above-average range. Similarly, it is now well The issue of comorbidity in relation to
recognized that seizure disorders of various autism is further complicated by the nature of
types are associated with autism in about 25% the syndrome. While autism is a lifelong disor-
of cases (see Chapter 18, this Handbook, this der and probably one of the best examples of a
volume). A much smaller proportion of autis- disorder in psychiatry, symptoms change with
tic individuals exhibit fragile X syndrome or age and developmental level. If the approach to
tuberous sclerosis (see Chapter 18, this Hand- diagnosis focuses on symptoms, an individual
book, this volume). Apart from these well- with autism will receive a large number of ad-
recognized associations, the association of ditional diagnoses over the course of the life
autism with other medical and behavioral con- span, including diagnoses that focus on anxi-
ditions is much less convincing (Rutter, Bai- ety, language, social problems, and the like.
ley, et al., 1994). Such a list of additional diagnoses might serve
Issues relating to comorbidity arise from a useful function by cataloging behaviors in
a major difference between approaches to need of clinical attention. But the list does not
Issues in the Classification of Autism and Related Conditions 27

basically change the fundamental conception autistic persons are largely or entirely mute,
that the person has autism. and for some disorders, this presents a pro-
Given the wide range and severity of the found diagnostic problem (Tsai, 1996). For ex-
disabilities experienced by individuals with ample, early investigators incorrectly assumed
autism, it is not surprising that they are vulner- continuity between autism and schizophrenia.
able to many types of behavioral difficulties, While persons with autism may also develop
including hyperactivity, obsessive-compulsive schizophrenia (Petty, Ornitz, Michelman, &
phenomena, self-injury and stereotypy, tics, Zimmerman, 1985), this does not appear to be
and affective symptoms (Brasic, Barnett, Ka- above the level expected in the general popula-
plan, Sheitman, Aisemberg et al., 1994; Ghaz- tion (Volkmar & Cohen, 1991a). Similarly, the
iuddin et al., 1992; Ghaziuddin, Alessi, & issue of comorbid obsessive-compulsive disor-
Greden, 1995; Jaselskis, Cook, & Fletcher, der and autism has been of interest given the
1992; Nelson & Pribor, 1993; Poustka & use of new pharmacological treatments such as
Lisch, 1993; Quintana et al., 1995; Realmuto the selective serotonin reuptake inhibitors
& Main, 1982). Interpretation of the available (SSRIs; see Chapter 44, this Handbook, Vol-
data is more complex when you move past the ume 2; Gordon, Rapoport, Hamburger, State,
level of behavioral observation and try to con- & Mannheim, 1992; Gordon, State, Nelson,
sider these associations within a causal Hamburger, & Rapoport, 1993; McDougle,
framework. For example, the diagnosis of Price, Volkmar, & Goodman, 1992). While
Tourette’s syndrome requires only the history phenomena suggestive of obsessions or com-
of motor and vocal tics for a year or more. Do pulsions are often observed in adults with
the compulsive behaviors and vocalizations autism (Rumsey, Rapoport, & Sceery, 1985),
emitted by many individuals with autism and levels of such phenomena vary considerably
intellectual disability warrant a second diag- across samples (Brasic et al., 1994; Fom-
nosis of Tourette’s syndrome? When should bonne, 1992; McDougle et al., 1995), and re-
obsessive-compulsive disorder be diagnosed sponse to medication may not be specific to
in a retarded, autistic individual with many diagnosis. In general, it appears that the ritu-
perseverative behaviors? alistic phenomena of autism and typical ob-
Diagnostic systems like DSM-IV and ICD-10 sessions and compulsions cannot simply be
strive for logical consistency in their approach equated (Baron-Cohen, 1989).
to the problem of diagnosis; this usually means Stereotyped motor movements and other
that some degree of hierarchical decision mannerisms are very common in autism but do
must be employed when, for example, fea- not qualify a case for the additional diagnosis
tures that are part of the definition of autism of stereotyped movement disorder. However, a
are observed in other disorders. Thus, since number of case reports and some case series
stereotyped behaviors are common in autism have suggested a potentially more interesting
and are included as a diagnostic feature in association between autism and Tourette’s
both DSM-IV and ICD-10, persons with disorder. In the latter condition, the child ex-
autism cannot also receive a diagnosis of hibits persistent motor and vocal tics (Burd,
stereotyped movement disorder. Similarly, di- Fisher, Kerbeshian, & Arnold, 1987; Leck-
agnostic problems arise with difficulties that man, Peterson, Pauls, & Cohen, 1997; Nelson
are commonly observed to be “associated fea- & Pribor, 1993; Realmuto & Main, 1982). It
tures” of autism, for example, unusual affective remains to be seen whether such an associa-
responses. On the other hand, mental retarda- tion is more frequent than would be expected
tion is not an essential diagnostic feature of by chance alone, particularly since differenti-
autism, and it is thus possible (and important) ation of tics and stereotyped motor manner-
for this diagnosis and one of autism to be made isms can be confusing for less experienced
when both sets of criteria are satisfied. clinicians.
The task of moving from the level of be- Affective symptoms are frequently ob-
havioral problems and symptoms to formal served in persons with autism. These symp-
psychiatric/developmental diagnosis is com- toms include affective lability, inappropriate
plicated by the nature of autism itself. Half of affective responses, anxiety, and depression.
28 Diagnosis and Classification

For higher functioning autistic persons, an which may suggest disorders of attention.
awareness of their difficulties may result in Hellgren, Gillberg, and Gillberg (1994) have
overt clinical depression. There is some sug- described a putative condition characterized
gestion that adolescents with Asperger’s are at by problems in attention, motor control and
particularly high risk for depression (Klin, perception (DAMP) with features of both
Volkmar, & Sparrow, 2000). Bipolar disorders PDD and attention deficit disorder.
have also been reported and may respond to Autistic individuals are not immune to any
drug treatment (Gillberg, 1985; Kerbeshian, other known medical conditions (Chapters 16
Burd, & Fisher, 1987; Komoto, Usui, & Hirata, & 18, this Handbook, this volume). Yet, spe-
1984; Lainhart & Folstein, 1994; Steingard & cific associations between autism and general
Biederman, 1987). medical conditions generally have not been
Given the characteristic difficulties in so- sustained by formal research. Although some
cial interaction and communication, as well as investigators (e.g., Gillberg, 1990) suggest that
the frequent association of autism with mental many different associations are common, stud-
retardation, it is not surprising that deploy- ies that employ stringent diagnostic criteria
ment and sustaining of attention would be have not supported this view (e.g., Rutter, Bai-
problematic for individuals with autism (see ley, Bolton, & Le Couter, 1994). In one sense,
Chapter 13, this Handbook, this volume). In this issue is simply definitional. If you take a
DSM-III-R, the convention was established very broad view of autism, a large number of
that autism and attention deficit disorder were persons with profound intellectual disability
made mutually exclusive diagnoses. This was will be included in samples of autistic individ-
based on the clinical belief that attentional uals; this population has a marked increase in
problems in autism were better viewed as an the number of medical conditions that may be
aspect of the autistic condition and develop- significantly involved in the person’s develop-
mental level; there was a clinical impression mental difficulties. The difficulties inherent
that stimulant medications used in the treat- in including such cases among those with more
ment of attention deficit disorder often led to strictly defined autism are exemplified in the
deterioration in the behavior of individuals early reports about the association of autism
with autism. The latter notion has now been with congenital rubella. Children with congen-
called into question (see Towbin, 2003, for a ital rubella initially were reported to have
review), and there is little doubt that atten- many autistic-like features and to be very low
tional difficulties are observed in children functioning; over time, however, the diagnoses
with autism (Charman, 1998), but the question of these cases have proven questionable.
of whether such difficulties are sufficient to
justify an additional diagnosis of attention Subtypes of Autism
deficit disorder remains unclear. Attentional
difficulties may be intrinsically associated Investigators have used various approaches to
with developmental problems and may reflect subtype autism and the broader PDD class of
broader difficulties in cognitive organization conditions. Essentially, these attempts have
(Iacoboni, 2000) without necessarily implying fallen into two broad categories. The more
attention deficit disorder. While some have common approach rests on clinical experience
suggested that attention deficit / hyperactivity and the ability of clinician-investigators to no-
disorder should be considered an additional di- tice features that are then used to delineate a
agnosis and target of treatment in persons with specific diagnostic concept. Kanner’s descrip-
autism (Tsai, 1999), firm empirical data on tion of autism and the work of Asperger, Rett,
this issue are lacking. and Heller are all examples of this approach.
Barkely (1990) has noted that the issue of More recent examples include the proposed
attentional problem is of much greater interest typology based on social characteristics pro-
in children with PDD-NOS. Such children do posed by Wing and colleagues (Wing & Gould,
not exhibit classical autism but have persistent 1979). The major alternative is to utilize more
problems in social interaction and the regula- complex statistical procedures to derive sub-
tion of affective responses and behavior, groups or subtypes empirically. It might seem
Issues in the Classification of Autism and Related Conditions 29

more likely that the latter approach would be analyses. It is a testament to the creativity of
more productive, but, somewhat surprisingly, engaged clinicians and to the human capacity
this really has not been the case. to notice regularities that at least so far the di-
agnostic concepts we are presently familiar
Statistical Approaches to Subtyping
with have emerged from clinical work and not
Complex statistical approaches have been from complex statistical analyses. On the other
helpful in developing and validating screening hand, such analyses may be helpful in examin-
and assessment instruments, as well as in de- ing current diagnostic concepts and alternative
veloping criteria to operationalize diagnostic ways to conceptualize syndrome boundaries. It
concepts. Their value in developing new diag- is possible, in the future, that better diagnostic
nostic categories has been limited by several concepts will be derived, for example, within
factors. Approaches such as cluster and factor the broad category of PDD-NOS.
analysis, in the first place, are very dependent Despite these problems, cluster and factor
on the characteristic of the sample being stud- analytic approaches have been used with some
ied and on the information originally pro- frequency. For example, in an early study,
vided; you cannot identify relevant variables or Prior and colleagues (Prior, Boulton, Gajzago,
combinations of variables if they are not mea- & Perry, 1975) observed two clusters of cases.
sured in the sample in the first place. Since One cluster was more similar to Kanner’s orig-
our knowledge regarding the underlying neuro- inal syndrome in terms of early onset and clin-
pathological basis of autism and its relation- ical features and the other with later onset and
ship to development and behavior remains more complex features. Similarly, Siegel, An-
limited, it is not clear exactly what measures ders, Ciaranello, Bienenstock, and Kramer
would best be included in such analyses. An- (1986) identified four possible subgroups in a
other set of issues surrounds a set of interre- larger group of children with PDDs. Two
lated problems: the marked range in syndrome groups appeared to correspond roughly to low
expression associated with age and develop- and higher functioning autism while the other
mental level and issues related to sample selec- two groups were characterized either by
tion and sample size. Nosological research schizotypal features or affective symptoms
using complex statistical models generally re- and behavior problems. Dahl, Cohen, and
quires large and representative samples of pa- Provence (1986) identified two clusters of
tients. Unfortunately, the samples used in most children in the PDD spectrum who had similar
studies are small and not representative. Re- behavior problems but somewhat different pat-
sults may be highly dependent on the original terns of language functioning and onset. De-
sample and may not generalize to other sam- pending on sample and range of variables
ples. This problem is compounded by the fact included in the analyses, various numbers of
that the meaning of behaviors may change with clusters have been derived. The less robust
age and with developmental level. The diagno- clusters—those with fewer cases and very
sis of autism may be particularly difficult to complex clinical features—are less likely to be
make in very young children below the age of observed in subsequent studies. Eaves, Eaves,
3 years. You might assume that the purest form and Ho (1994) used data from over 150 chil-
of autism is exhibited at this young age. How- dren with autism spectrum disorders. In their
ever, as Lord (1995) has shown, the character- sample, four meaningful subtypes emerged
istic symptoms of autism such as repetitive with different behavioral and cognitive pro-
behaviors often do not clearly develop before files. Over half the sample fell into the subtype
age 3 years while significant social deficits, described as typically autistic; approximately
suggestive of autism, may markedly improve 20% were also autistic but were lower func-
after the first two years of life (see also tioning cognitively. The remaining cases
Rogers, 2001). formed two subtypes: One was a higher func-
The strong developmental nature of changes tioning group with similarities to Asperger’s
in syndrome expression means that variables and another with less severe difficulties. Fein,
such as age, developmental level, or IQ them- Waterhouse, Lucci, and Snyder (1985) identi-
selves become important variables in statistical fied eight cognitive profiles that could be
30 Diagnosis and Classification

related to handedness (Soper et al., 1986) but associated medical condition do not simply
not to more usual autistic features. More re- correspond to obvious behavioral subtypes
cently, Waterhouse and colleagues (1996) (Rutter, 1996). As Rutter has noted (2000),
studied a relatively large group of children conditions such as autism are defined on the
with some form of PDD not associated with an basis of their clinical features, and it is likely
overt medical condition; they suggested that at that complex, multifactorial models will be
least two overlapping continua were present, needed to understand underlying pathophysiol-
corresponding roughly to lower and higher ogy. That is, systems such as DSM and ICD are
functioning autism. strongly influenced by pathophysiology when
Methods other than cluster and factor this is known but should not simply be thought
analysis have been employed as well in the of as classifying by cause.
search for subgroups. For example, I. Cohen, As with the more statistically based ap-
Sudhalter, Landon-Jimenez, and Keogh (1993) proaches, clinically inspired approaches also
utilized a novel system of pattern recognition must deal with the major confounding problem
(neural networks) as well as discriminant of intellectual level. For example, the three-
analyses; they argued that the neural network group subtyping (aloof, passive, active-but-
procedure was superior in correctly identify- odd) proposed by Wing and Gould (1979)
ing whether autism was or was not present. In appears to sort children into relatively reliable
a well-controlled study by Cicchetti, Volkmar, groups; the typology has some measure of va-
Klin, and Showalter (1995), however, the lidity as well as potential benefits for planning
neural networks procedure was not as effective interventions (Borden & Ollendick, 1994;
as the simple diagnostic algorithm proposed in Castelloe & Dawson, 1993; Volkmar & Cohen,
ICD-10 and DSM-IV. 1989). However, differences among the sub-
Multivariate methods have also been uti- groups appear to be largely a function of asso-
lized to validate existing diagnostic groupings ciated IQ. When IQ is controlled for,
and new possible subgroups, for example, differences among the groups largely vanish
within the broad PDD-NOS category (see also (Volkmar & Cohen, 1989).
Chapter 6, this Handbook, this volume). Van Individuals with profound mental retarda-
der Gaag et al. (1995) utilized a multivariate tion exhibit a number of autistic-like features
cluster analysis and demonstrated differences (Wing & Gould, 1979) without, however,
between cases with autistic disorder and a spe- meeting full criteria for autism. Such cases
cific subtype of PDD-NOS (multiplex or mul- have many of the same service needs as those
tiple complex developmental disorder) on the with more strictly defined autism. Various in-
basis of clinical and developmental features. vestigators have, accordingly, proposed a dis-
tinction among primary, higher, and lower
Clinical Approaches to Subtyping
functioning autism given the very different
The issue of subtypes has also been approached patterns of educational need, associated med-
from a clinical standpoint. Wing and Gould ical problems, outcome, family history, and so
(1979) proposed a classification scheme based forth associated with lower and higher IQ
on the nature of observed patterns of social in- (Cohen, Paul, & Volkmar, 1986; Rutter, 1996;
teraction (aloof, passive, active-but-odd; see Tsai, 1992; Waterhouse et al., 1996). This im-
also Chapter 7, this Handbook, this volume). portant issue remains unresolved. Similarly, it
Other classifications have focused on cognitive is clear that, over time, children with severe
profiles (Fein et al., 1985), language problems developmental language disorders go on to ex-
(Rapin, 1991; Rapin & Allen, 1983), presence hibit marked social difficulties (Howlin,
of signs of overt central nervous system dys- Mawhood, & Rutter, 2000) so that the issue of
function (Tsai, Tsai, & August, 1985), and so the connection between language disorders
forth. A decade ago, it appeared that possible and autism remains an important area of
associations of autism with various medical study.
conditions would have major implications for
Developmental Regression
understanding subtypes and etiology. At pres-
ent, however, it appears that distinctions based Various studies have suggested that perhaps
on the presence of a strictly defined etiology or 20% to 25% of children with autism have some
Issues in the Classification of Autism and Related Conditions 31

degree of developmental regression (see Chap- older (and make communicative gains).
ter 3, this Handbook, this volume). Unfortu- Stereotyped mannerisms also become some-
nately, this phenomenon remains poorly what more common when children become
understood and, in part as a result, controver- older while other features (e.g., persistent pre-
sial. Most studies have utilized parent report occupation with parts of objects) are consis-
with all the attendant problems of definition, tently observed.
reliability, and validity. In some cases, parents Examination of some of the items not in-
report a pattern less of regression and more cluded in DSM-IV/ICD-10 also illustrates this
one of developmental stagnation; in other issue. Abnormal pitch/tone is largely a phe-
cases, the report is of a regression but the nomenon observed in older individuals while
history may also be remarkable for prior devel- attachments to unusual objects are less com-
opmental delays. Finally, in some cases, a dra- monly observed in older individuals. Similarly,
matic regression is observed (Siperstein & the phenomenon of hyper- or hyposensitivity to
Volkmar, 2004). The most common pattern is the inanimate environment has a complicated
one in which a few words are apparently ac- developmental course with features exhibited
quired and then lost. The more dramatic cases at some ages and not others (see Chapter 32,
(e.g., where hundreds of words are acquired this Handbook, Volume 2).
and then lost) are often more consistent with a
diagnosis of childhood disintegrative disorder; Autism in Infants and Young Children
however, the latter condition, by definition,
has its onset after age 2. It is possible that Increased awareness (on the part of both the
some of the earlier and more dramatic cases general public and health care providers) and
of regression are expressions of the earliest advances in early diagnosis have led to a
forms of childhood disintegrative disorder. In change in the age at which autism is first diag-
any event, the study of this phenomenon (ide- nosed. A decade ago, diagnosis at age 4 was
ally at the time it happens) using various relatively typical (Siegel, Pliner, Eschler, &
methodologies (genetics, neuroimaging, EEG, Elliott, 1988)—even when parents had been
etc.) is critically needed. concerned much earlier. It is now more com-
mon for specialized diagnostic centers to see
Developmental Change
children at age 2 years (Lord, 1995; Moore &
Important issues of developmental change in Goodson, 2003) or even younger (Klin, Ca-
syndrome expression (over both age and IQ hawarska, Paul, Rubin, Morgan, et al., 2004).
level) have been recognized for many years The increased interest in early diagnosis and
(Rutter, 1970). Diagnostic systems such as the increasing numbers of younger children
DSM-IV and ICD-10 have generally adopted presenting for assessment present special prob-
the stance of providing criteria that are specif- lems for diagnosis. In contrast to older individ-
ically meant to cover this range of syndrome uals, the diagnosis of infants and very young
expression. An alternative, if rather unwieldy, children is more complex (Charman & Baird,
approach is to provide different diagnostic cri- 2002; Cox et al., 1999; Stone et al., 1999) with
teria either for different age groups or for dif- diagnostic stability increasing after about age
ferent levels of impairment (e.g., depending on 2 years (Courchesne, 2002; Dawson et al.,
level of communicative ability). 2002). However, developmental changes in this
Examination of the data from the DSM-IV age group can be marked (Szatmari, Merette,
field trial illustrates some of these issues. For Bryson, Thivierge, Roy, et al., 2002). For ex-
example, if we utilize the phi statistic to evalu- ample, the repetitive behaviors typical of older
ate the ability of criteria to predict autism, the children are much less common in very young
criteria included in DSM-IV and ICD-10 are children (Charman & Baird, 2002; Cox et al.,
generally comparably powerful predictors 1999; Lord, 1995; Moore & Goodson, 2003;
across age and developmental level with some Stone et al., 1999). Social abnormalities may
expectable but not overly dramatic exceptions; become more striking as the child matures
for example, stereotyped language use and (Lord, Storoschuk, Rutter, & Pickles, 1993).
problems in conversation would be expected to A few studies have addressed the appli-
become more common as children become cability of DSM-IV and ICD-10 criteria in
32 Diagnosis and Classification

infants and young children. It appears that critically needed. Given the very limited liter-
some young children will meet criteria for ature on the topic of cultural factors, this area
autism, but some may not necessarily fulfill is one ripe for future research. Chapter 48
the required repetitive behavior criteria until (this Handbook, Volume 2) provides an inter-
around their third birthday (Lord, 1996). Less national perspective on this problem.
commonly, a child appears to meet criteria for
Defining the Broader Phenotype
autism but then, over time, makes substantial
gain. Some alternatives to DSM-IV and ICD- Somewhat paradoxically as the definition of
10 have been proposed (e.g., National Center autism has become more elaborated, interest
for Clinical Infant Programs [NCCIP], 1994) has also increased in the broader spectrum of
but have not met with wide acceptance due to difficulties apparently inherited in families
both practical and theoretical concerns. (see Chapter 16, this Handbook, this volume).
Considerable efforts have gone into the de- Most investigators would now agree what is
velopment of methods to facilitate screening transmitted genetically includes not only clas-
and early diagnosis (see Chapter 27, this Hand- sical autism (Kanner, 1943) but a broader
book, Volume 2). Given the apparent associa- range of difficulties variously impacting on
tion of early identification and intervention social development, communication, and/or
with improved outcome (NRC, 2001) the issues behavior. Attempts are now being made to
of early diagnosis have assumed increasing im- stratify families based on various measures
portance. In addition to the various approaches initially designed for use in more stringently
for screening based on history and direct ob- diagnostic autistic samples (Bishop, 1998;
servation, new approaches are needed in which Constantino & Todd, 2003; Lord, 1990; Lord
screening becomes more behavioral and less et al., 2000; Shao et al., 2002; Tadevosyan-
subjective (and thus more readily available in Leyfer et al., 2003; Tanguay, Robertson, &
nonspecialist settings; see Chawarska, Klin, & Derrick, 1998). Such approaches hold promise
Volkmar, 2003). for identifying broader dimensions of func-
tion /dysfunction in families. The development
Cultural Issues and Diagnosis
of new methods for assessing the broader phe-
The issue of cultural factors in the diagnosis of notype (e.g., Bishop, 1998; Constantino &
autism has been the subject of remarkably lit- Todd, 2003) is of great interest in this regard.
tle discussion. As Brown and Rogers (2003) In addition to both the more strictly de-
point out, this is somewhat paradoxical given fined cases of autism, the broader range of
the various governmental and other mandates autism spectrum disorders includes difficul-
for the study of cultural factors. While by no ties that do not fit neatly into our current clas-
means excusing the dearth of studies, several sification scheme. Such cases of atypical
factors likely have operated to reduce interest autism test the boundaries of our classification
in this area. First, the general impression of system but also serve to underscore the impor-
clinicians seeing children from a range of cul- tant point that individuals with these condi-
tures and subcultures around the world is one tions have not always read the textbooks and
of how much more alike than different chil- may exhibit unusual patterns of difficulty sug-
dren are. While variations in treatment and, to gestive of autism in some ways but also with
some extent, theoretical conceptualizations important differences. Children reared in pro-
differ (see Chapter 48, this Handbook, Volume foundly impoverished environments may ex-
2), it is a testament to the robustness of autism hibit marked social difficulties and other
as a diagnostic concept that cultural influences problems suggestive of autism (Rutter, 1999).
are not more striking. One potential exception Similar issues arise with respect to children
(although one tending to prove the rule) relates who are congenitally blind (Hobson & Bishop,
to the high levels of autistic-like behavior in 2003). Yet another set of issues arises with re-
individuals who suffer severe early institu- gard to children who, at least initially, seem to
tional deprivation (Rutter, 1999). More rigor- exhibit problems more suggestive of a lan-
ous and well-controlled studies on the issue of guage disorder but, over time, exhibit a course
social-cultural factors in autism are clearly and outcome in some ways more suggestive of
Issues in the Classification of Autism and Related Conditions 33

autism (Mawhood, Howlin, & Rutter, 2000). disorder (Bishop, 1989, 2002), also remain to
Issues with regard to differentiation of autism be clearly established.
and Asperger’s and language disorders have While DSM-IV and ICD-10 are the most re-
been noted (Bishop, 2000; Bishop & Norbury, cent and most extensively evaluated diagnostic
2002). Cases with unusual features or presen- approaches for autism, they are undoubtedly
tations are of great interest in that they may not the last word on diagnosis. The present
help to clarify syndrome boundaries, under- DSM-IV and ICD-10 systems have the consid-
score areas where knowledge is lacking, and erable advantage of being based on a relatively
may clarify alternative mechanisms or devel- extensive set of data; they have clearly facili-
opmental pathways. For example, while there tated research and service. The dual-use con-
is little disagreement that higher functioning straints on DSM, that is, the use of the same
autism and Asperger’s disorder both are char- criteria for both research and service, meant
acterized by significant problems in social in- that brevity and ease of use were important
teraction in the face of average overall considerations. The ICD-10 system does not,
cognitive ability, the social difficulties appear at least for the research definitions, have this
to arise in the context of rather different devel- constraint. It remains to be seen whether the
opmental pathways and trajectories, for exam- more detailed ICD-10 research definition will,
ple, with preservation of language skills early in the end, predominate. From the point of
on, and possibly later, in Asperger’s but not in view of research, the attempt to link diagnostic
higher functioning autism (see Chapter 5, this instruments specially to diagnostic criteria is a
Handbook, this volume). considerable advantage and may mean that for
research purposes, in effect, the more detailed
CONCLUSION research definition will come to dominate.
Probably the greatest nosological need at
Leo Kanner’s description (1943) of the syn- present is the classification of conditions that
drome of early infantile autism has proven to appear to fall within the broad class of the
be robust and enduring. To a remarkable de- PDDs but do not meet criteria for presently
gree, his observations and intuitions remain recognized disorders. This group of condi-
fresh and inspiring. False leads in the original tions, referred to either as “atypical autism” or
work have been clarified by research. We are “pervasive developmental disorder not other-
also aware of how much work remains 60 wise specified,” includes a larger number of
years later. children than those who are stringently de-
Studies have clarified that the disintegra- fined as autistic. Their nosological status is
tive PDDs (Rett’s disorder and childhood dis- much less well defined (see Chapter 6, this
integrative disorder) differ from strictly Handbook, this volume). Concepts such as
defined autism in various ways (Tsai, 1992; multiplex developmental disorder have been
Volkmar & Rutter, 1995); the study of these proposed for some of these individuals. A large
unusual conditions may be helpful in clarify- subgroup of such cases is associated with se-
ing mechanisms of pathogenesis relevant vere mental handicap. These conditions re-
to autism (see Chapters 3 & 5, this Hand- quire special services similar to those required
book, this volume). The validity of the for autism (Wing & Gould, 1979); their rela-
newest PDD—Asperger’s disorder—apart tionship to strictly defined autism remains an
from higher functioning autism is less clearly area of considerable interest and may have par-
established and results contradictory (al- ticular importance for family-genetic studies
though often based on markedly differing de- (Rutter, 1996). Biological and behavioral re-
finitions of the disorder; Gilchrist et al., 2001; search depends on well-defined groups of pa-
Klin et al., 1995; Manjiviona & Prior, 1999; tients and rigorous application of diagnostic
Miller & Ozonoff, 2000; Ozonoff, Pennington, methodologies. For example, genetic studies
& Rogers, 1991). The boundaries of As- require clear definition of affected individuals
perger’s disorder with autism and other disor- and exclusion of false-positive cases. In turn,
ders, such as schizoid disorder of childhood we can hope that future nosologies will be en-
(Wolff, 1998, 2000) and semantic-pragmatic riched by the inclusion of other types of data,
34 Diagnosis and Classification

including genetic, neuroimaging, neurochemi- Bartak, L., Rutter, M., & Cox, A. (1977). A com-
cal, and other behavioral and biological mark- parative study of infantile autism and specific
ers. Thus, there is a critical dialectic between developmental receptive language disorders—
research in nosology and research of other III. Discriminant function analysis. Journal of
types. Advances in both fields are mutually Autism and Childhood Schizophrenia, 7,
dependent and have the same goal: enhancing
Bender, L. (1946). Childhood schizophrenia. Amer-
the understanding and care of individuals and ican Journal of Orthopsychiatry, 17, 40–56.
advancing our understanding of autism and re- Bishop, D. V. (1989). Autism, Asperger’s syndrome
lated conditions (Rutter, 1999). and semantic-pragmatic disorder: Where are
the boundaries? [Special issue: Autism].
Cross-References British Journal of Disorders of Communica-
tion, 24(2), 107–121.
Other syndromes presently included as PDDs Bishop, D. V. (1998). Development of the Chil-
are discussed in Chapters 3 through 6; Chapter dren’s Communication Checklist (CCC): A
21 provides an alternative view of issues of di- method for assessing qualitative aspects of
communicative impairment in children. Jour-
agnosis and classification; changes in syn-
nal of Child Psychology and Psychiatry and
drome expression are discussed in Chapters 8
Allied Disciplines, 39(6), 879–891.
through 10. Bishop, D. V. (2000). What’s so special about As-
perger syndrome? The need for further explo-
REFERENCES ration of the borderlands of autism. In A. Klin
& F. R. Volkmar (Eds.), Asperger syndrome
American Psychiatric Association. (1952). Diag- (pp. 254–277). New York: Guilford Press.
nostic and statistical manual of mental disor- Bishop, D. V., & Norbury, C. F. (2002). Exploring
ders. Washington, DC: Author. the borderlands of autistic disorder and spe-
American Psychiatric Association. (1968). Diag- cific language impairment: A study using
nostic and statistical manual of mental disor- standardised diagnostic instruments. Journal
ders (2nd ed.). Washington, DC: Author. of Child Psychology and Psychiatry and Allied
American Psychiatric Association. (1980). Diag- Disciplines, 43(7), 917–929.
nostic and statistical manual of mental disor- Bleuler, E. (1951). Dementia Praecox, or the Group
ders (3rd ed.). Washington, DC: Author. of Schizophrenia. Translated by J. Zinkin. New
American Psychiatric Association. (1987). Diag- York: International Universities Press. (Origi-
nostic and statistical manual of mental disor- nal work published 1911)
ders (3rd ed., rev.). Washington, DC: Author. Borden, M. C., & Ollendick, T. H. (1994). An ex-
American Psychiatric Association. (1994). Diag- amination of the validity of social subtypes in
nostic and statistical manual of mental disor- autism. Journal of Autism and Developmental
ders (4th ed.). Washington, DC: Author. Disorders, 24(1), 23–37.
Amir, R. E., Van den Veyver, I. B., Wan, M., Tran, Brasic, J. R., Barnett, J. Y., Kaplan, D., Sheitman,
C. Q., Francke, U., & Zoghbi, H. Y. (1999). B. B., Aisemberg, P., Lafargue, R. T., et al.
Rett syndrome is caused by mutations in X- (1994). Clomipramine ameliorates adventitious
linked MECP2, encoding methyl-CpG-binding movements and compulsions in prepubertal
protein 2 [See comment]. Nature Genetics, boys with autistic disorder and severe mental
23(2), 185–188. retardation. Neurology, 44(7), 1309–1312.
Asperger, H. (1944). Die “autistichen Psychopa- Brown, J. R., & Rogers, S. J. (2003). Cultural is-
then” im Kindersalter. Archive fur psychiatrie sues in autism. In R. L. Hendren, S. Ozonoff,
und Nervenkrankheiten, 117, 76–136. & S. Rogers (Eds.), Autism spectrum disorders
Bailey, A., Palferman, S., Heavey, L., & Le Cou- (pp. 209–226). Washington, DC: American
teur, A. (1998). Autism: The phenotype in rel- Psychiatric Press.
atives. Journal of Autism and Developmental Buitelaar, J. K., Van der Gaag, R., Klin, A., & Volk-
Disorders, 28(5), 369–392. mar, F. (1999). Exploring the boundaries of
Barkely, R. A. (1990). Attention deficit hyperactiv- Pervasive Developmental Disorder Not Other-
ity disorder: A handbook for diagnosis and wise Specified: Analyses of data from the
treatment. New York: Guilford Press. DSM-IV autistic field trial. Journal of Autism
Baron-Cohen, S. (1989). Do autistic children have and Developmental Disorders, 29(1), 33–43.
obsessions and compulsions? British Journal of Burd, L., Fisher, W. W., Kerbeshian, J., & Arnold,
Clinical Psychology, 28(3), 193–200. M. E. (1987). Is development of Tourette dis-
Issues in the Classification of Autism and Related Conditions 35

order a marker for improvement in patients Cohen, I. L., Sudhalter, V., Landon-Jimenez, D., &
with autism and other pervasive developmen- Keogh, M. (1993). A neural network approach
tal disorders? Journal of the American Acad- to the classification of autism. Journal of
emy of Child and Adolescent Psychiatry, 26(2), Autism and Developmental Disorders, 23(3),
162–165. 443–466.
Candland, D. K. (1993). Feral children and clever Constantino, J. N., & Todd, R. D. (2003). Autistic
animals: Ref lections on human nature. Oxford traits in the general population: A twin study.
University Press, 411. Archives of General Psychiatry, 60(5),
Cantwell, D. P. (1996). Classification of Child and 524–530.
Adolescent Psychopathology. Journal of Child Courchesne, E. (2002). Abnormal early brain de-
Psychiatry and Psychology, 37, 3–12. velopment in autism. Molecular Psychiatry,
Cantwell, D. P., Baker, L., & Rutter, M. (1979). 7(Suppl. 2), S21–S23.
Families of autistic children and dysphasic Cox, A., Klein, K., Charman, T., Baird, G., Baron-
children: Family life and interaction patterns. Cohen, S., Swettenham, J., et al. (1999).
Archives of General Psychiatry, 36, 682–687. Autism spectrum disorders at 20 and 42
Castelloe, P., & Dawson, G. (1993). Subclassifica- months of age: Stability of clinical and ADI-R
tion of children with autism and pervasive de- diagnosis. Journal of Child Psychology and
velopmental disorder: A questionnaire based Psychiatry and Allied Disciplines, 40(5),
on Wing’s subgrouping scheme. Journal of 719–732.
Autism and Developmental Disorders, 23(2), Dahl, E. K., Cohen, D. J., & Provence, S. (1986).
229–241. Clinical and multivariate approaches to the
Charman, T. (1998). Specifying the nature and nosology of pervasive developmental disor-
course of the joint attention impairment in ders. Journal of the American Academy of
autism in the preschool years: Implications for Child Psychiatry, 25(2), 170–180.
diagnosis and intervention. Autism: Interna- Dawson, G., Munson, J., Estes, A., Osterling, J.,
tional Journal of Research and Practice, 2(1), McPartland, J., Toth, K., et al. (2002). Neu-
61–79. rocognitive function and joint attention ability
Charman, T., & Baird, G. (2002). Practitioner re- in young children with autism spectrum disor-
view: Diagnosis of autism spectrum disorder der versus developmental delay. Child Devel-
in 2- and 3-year-old children. Journal of Child opment, 73(2), 345–358.
Psychology and Psychiatry and Allied Disci- DeMyer, M. K., Hingtgen, J. N., & Jackson, R. K.
plines, 43(3), 289–305. (1981). Infantile autism reviewed: A decade of
Chawarska, K., Klin, A., & Volkmar, F. (2003). research. Schizophrenia Bulletin, 7(3), 388–451.
Automatic attention cueing through eye move- Dykens, E. M. (2000). Psychopathology in children
ment in 2-year-old children with autism. Child with intellectual disability. Journal of Child
Development, 74(4), 1108–1122. Psychology and Psychiatry and Allied Disci-
Chess, S., Fernandez, P., & Korn, S. (1978). Behav- plines, 41(4), 407–417.
ioral consequences of congenital rubella. Jour- Eaves, L. C., Eaves, D. M., & Ho, H. H. (1994).
nal of Pediatrics, 93(4), 699–703. Subtypes of autism by cluster analysis. Journal
Cicchetti, D. V., Volkmar, F. R., Klin, A., & of Autism and Developmental Disorders, 24(1),
Showalter, D. (1995). Diagnosing autism using 3–22.
ICD-10 criteria: A comparison of neural net- Eisenmajer, R., Prior, M., Leekam, S., Wing, L.,
works and standard multi variate procedures. Ong, B., Gould, J., et al. (1996). Comparison
Journal of Child Neuropsychology, 1(1), 26–37. of clinical symptoms in autism and Asperger’s
Cohen, D. J. (1976). The diagnostic process in disorder. Journal of the American Academy of
child psychiatry. Psychiatric Annals, 6(9), Child and Adolescent Psychiatry, 35(11),
29–56. 1523–1531.
Cohen, D. J. (1980). The pathology of the self in pri- Factor, D. C., Freeman, N. L., & Kardash, A. (1989,
mary childhood autism and Gilles de la December). A comparison of DSM-III and
Tourette syndrome. In B. Blinder (Ed.), Psychi- DSM-III-R criteria for autism. Journal of
atric clinics of North America (pp. 383–402). Autism and Developmental Disorders, 19(4),
Philadelphia: Saunders. 637–640.
Cohen, D. J., Paul, R., & Volkmar, F. R. (1986). Is- Fein, D., Waterhouse, L., Lucci, D., & Snyder, D.
sues in the classification of pervasive and other (1985). Cognitive subtypes in developmentally
developmental disorders: Toward DSM-IV. disabled children: A pilot study. Journal of
Journal of the American Academy of Child Psy- Autism and Developmental Disorders, 15(1),
chiatry, 25(2), 213–220. 77–95.
36 Diagnosis and Classification

Fombonne, E. (1992). Diagnostic assessment in a double-blind comparison of clomipramine, de-

sample of autistic and developmentally im- sipramine, and placebo in the treatment of
paired adolescents [Special issue: Classifica- autistic disorder. Archives of General Psychia-
tion and diagnosis]. Journal of Autism and try, 50(6), 441–447.
Developmental Disorders, 22(4), 563–581. Harper, J., & Williams, S. (1975). Age and type of
Frances, A. J., Widiger, T. A., & Pincus, H. A. onset as critical variables in early infantile
(1989). The development of DSM-IV. Archives autism. Journal of Autism and Childhood
of General Psychiatry, 46, 373–375. Schizophrenia, 5, 25–35.
Ghaziuddin, M., Alessi, N., & Greden, J. F. (1995). Heller, T. (1908). Dementia Infantilis. Zeitschrift
Life events and depression in children with fur die Erforschung und Behandlung des Ju-
pervasive disorders. Journal of Autism and De- genlichen. Schwachsinns, 2, 141–165.
velopmental Disorders, 25(5), 495–502. Hellgren, L., Gillberg, C., & Gillberg, I. C. (1994).
Ghaziuddin, M., & Butler, E. (1998). Clumsiness in Children with deficits in attention, motor con-
autism and Asperger syndrome: A further re- trol and perception (DAMP) almost grown up:
port. Journal of Intellectual Disability Re- The contribution of various background to
search, 42(Pt. 1), 43–48. outcome at age 16 years. European Child and
Ghaziuddin, M., & Gerstein, L. (1996). Pedantic Adolescent Psychiatry, 3(1), 1–15.
speaking style differentiates Asperger syn- Hendry, C. N. (2000). Childhood disintegrative dis-
drome from high-functioning autism. Journal order: Should it be considered a distinct diag-
of Autism and Developmental Disorders, 26(6), nosis. Clinical Psychology Review, 20, 77–90.
585–595. Hertzig, M. E., Snow, M. E., New, E., & Shapiro, T.
Ghaziuddin, M., Tsai, L., & Ghaziuddin, N. (1992). (1990). DSM-III and DSM-III-R diagnosis of
Comorbidity of autistic disorder in children autism and pervasive. Journal of the American
and adolescents. European Child and Adoles- Academy of Child and Adolescent Psychiatry,
cent Psychiatry, 1(4), 209–213. 29(1), 123–126.
Gilchrist, A., Green, J., Cox, A., Burton, D., Rut- Hippler, K., & Klicpera, C. (2003). A retrospective
ter, M., & Le Couteur, A. (2001). Development analysis of the clinical case records of “autis-
and current functioning in adolescents with tic psychopaths” diagnosed by Hans Asperger
Asperger syndrome: A comparative study. and his team at the University Children’s Hos-
Journal of Child Psychology and Psychiatry pital, Vienna. In U. Frith & E. Hill (Eds.),
and Allied Disciplines, 42(2), 227–240. Autism: Mind and Brain (pp. 21–42). Oxford,
Gillberg, C. (1985). Asperger’s syndrome and re- England: Oxford University Press.
current psychosis: A case study. Journal of Hobbs, N. (1975). Issues in the classification of
Autism and Developmental Disorders, 15(4), children. San Francisco: Jossey-Bass.
389–397. Hobson, R. P., & Bishop, M. (2003). The pathogen-
Gillberg, C. (1990). Medical work-up in children esis of autism: Insights from congenital blind-
with autism and Asperger syndrome. Brain ness. Philosophical Transactions of the Royal
Dysfunction, 3(5/6), 249–260. Society of London—Series B: Biological Sci-
Gillberg, C. (1991). Debate and argument: Is ences, 358(1430), 335–344.
autism a pervasive developmental disorder? Howlin, P. (2003). Outcome in high-functioning
Journal of Child Psychology and Psychiatry adults with autism with and without early lan-
and Allied Disciplines, 32(7), 1169–1170. guage delays: Implications for the differentia-
Gillberg, C. (1994). Debate and argument: Having tion between autism and Asperger syndrome.
Rett syndrome in the ICD-10 PDD category Journal of Autism and Developmental Disor-
does not make sense. Journal of Child Psychol- ders, 33(1), 3–13.
ogy and Psychiatry and Allied Disciplines, Howlin, P., Mawhood, L., & Rutter, M. (2000).
35(2), 377–378. Autism and developmental receptive language
Gillberg, C., & M. Coleman. (2000). The biology of disorders—a follow-up comparison in early
the autistic syndromes (3rd ed.). London: Mac adult life: II. Social, behavioural, and psychi-
Keith Press. atric outcomes. Journal of Child Psychology
Gordon, C. T., Rapoport, J. L., Hamburger, S. D., and Psychiatry, 41, 561–578.
State, R. C., & Mannheim, G. B. (1992). Dif- Iacoboni, M. (2000). Attention and sensorimotor
ferential response of seven subjects with autis- integration: Mapping the embodied mind. In
tic disorder to clomipramine and desipramine. A. W. Toga & J. C. Mazziotta (Eds.), Brain
American Journal of Psychiatry, 149(3), mapping: The systems (pp. 463–490). San
363–366. Diego, CA: Academic Press.
Gordon, C. T., State, R. C., Nelson, J. E., Ham- Jaselskis, C. A., Cook, E. H., & Fletcher, K. E.
burger, S. D., & Rapoport, J. L. (1993). A (1992). Clonidine treatment of hyperactive
Issues in the Classification of Autism and Related Conditions 37

and impulsive children with autistic disorder. pp. xi–xx). Washington, DC: American Psychi-
Journal of Clinical Psychopharmacology, atric Association.
12(5), 322–327. Lainhart, J. E., & Folstein, S. E. (1994). Affective
Kanner, L. (1943). Autistic disturbances of affec- disorders in people with autism: A review of
tive contact. Nervous Child, 2, 217–250. published cases. Journal of Autism and Devel-
Kerbeshian, J., Burd, L., & Fisher, W. (1987). opmental Disorders, 24(5), 587–601.
Lithium carbonate in the treatment of two pa- Leckman, J. F., Peterson, B. S., Pauls, D. L., &
tients with infantile autism and atypical bipo- Cohen, D. J. (1997). Tic disorders. Psychiatric
lar symptomatology. Journal of Clinical Clinics of North America, 20(4), 839–861.
Psychopharmacology, 7(6), 401–405. Leekam, S., Libby, S., Wing, L., Gould, J., & Gill-
Klin, A., Chawarska, K., Paul, R., Rubin, E., Mor- berg, C. (2000). Comparison of ICD-10 and
gan, T., Wiesner, L., & Volkmar, F. R. (2004). Gillberg’s criteria for Asperger syndrome
Autism in a 15 month old child. American [Special issue: Asperger syndrome]. Autism:
Journal of Psychiatry, 161, 1981–1988. Journal of Research and Practice, 4(1), 11–28.
Klin, A., Jones, W., Schultz, R., & Volkmar, F. Lincoln, A. E., Courchesne, E., Kilman, B. A., El-
(2003). The enactive mind, or from actions to masian, R., & Allen, M. (1998). Neurobiology
cognition: Lessons from autism. Philosophical of Asperger syndrome: Seven case studies and
Transactions of the Royal Society of London— quantitative magnetic resonance imaging find-
Series B: Biological Sciences, 358(1430), ings. In E. Schopler, G. B. Mesibov, & L. J.
345–360. Kunc (Eds.), Asperger syndrome or high func-
Klin, A., Lang, J., Cicchetti, D. V., & Volkmar, tioning autism? (pp. 145–166). New York:
F. R. (2000). Brief report: Interrater reliability Plenum Press.
of clinical diagnosis and DSM-IV criteria for Lockyer, L., & Rutter, M. (1969). A five to fifteen
autistic disorder: Results of the DSM-IV year follow-up study of infantile psychosis:
autism field trial. Journal of Autism and Devel- III. Psychological Aspects. British Journal of
opmental Disorders, 30(2), 163–167. Psychiatry, 115, 865–882.
Klin, A., Schultz, R., Pauls, D., & Volkmar, F. R. Lockyer, L., & Rutter, M. (1970). A five to fifteen
(in press). Three diagnostic approaches to As- year follow-up study of infantile psychosis.
perger’s Disorder. Journal of Autism and De- British Journal of Social and Clinical Psychol-
velopmental Disorders. ogy, 9, 152–163.
Klin, A., Sparrow, S., & Volkmar, F. R. (1997). As- Lord, C. (1995). Follow-up of two year olds re-
perger’s syndrome. New York: Guilford Press. ferred for possible autism. Journal of Child
Klin, A., & Volkmar, F. R. (1997). Asperger’s Syn- Psychology and Psychiatry, 36, 1365–1382.
drome. In D. J. Cohen & F. R. Volkmar (Eds.), Lord, C., Storoschuk, S., Rutter, M., & Pickles, A.
Handbook of Autism and Pervasive Develop- (1993). Using the ADI-R to diagnose autism in
mental Disorders (2nd ed., pp. 94–122). New preschool children. Infant Mental Health Jour-
York: Wiley. nal, 14(3), 234–252.
Klin, A., Volkmar, F. R., & Sparrow, S. S. (Eds.). Magnusson, P., & Saemundsen, E. (2001). Preva-
(2000). Asperger syndrome. New York: Guil- lence of autism in Iceland. Journal of Autism
ford Press. and Developmental Disorders, 31(2), 153–163.
Klin, A., Volkmar, F. R., Sparrow, S. S., Cicchetti, Mahler, M. (1952). On child psychoses and schizo-
D. V., & Rourke, B. P. (1995). Validity and phrenia: Autistic and symbiotic infantiles psy-
neuropsychological characterization of As- choses. Psychoanalytic Study of the Child, 7,
perger syndrome: Convergence with nonverbal 286–305.
learning disabilities syndrome. Journal of Manjiviona, J., & Prior, M. (1999). Neuropsycho-
Child Psychology and Psychiatry, 36(7), logical profiles of children with Asperger
1127–1140. syndrome and autism. Autism: International
Kolvin, I. (1971). Studies in childhood psychoses: Journal of Research and Practice, 3(4),
Diagnostic criteria and classification. British 327–356.
Journal of Psychiatry, 118, 381–384. Mawhood, L., Howlin, P., & Rutter, M. (2000).
Komoto, J., Usui, S., & Hirata, J. (1984). Infantile Autism and developmental receptive language
autism and affective disorder. Journal of disorder—a comparative follow-up in early
Autism and Developmental Disorders, 14(1), adult life. I: Cognitive and language outcomes.
81–84. Journal of Child Psychology and Psychiatry
Kramer, M. (1968). The history of the efforts to and Allied Disciplines, 41(5), 547–559.
agree on an international classification of McDougle, C. J., Kresch, L. E., Goodman, W. K.,
mental disorders. In Diagnostic and statistical Naylor, S. T., Volkmar, F. R., Cohen, D. J., et al.
manual of mental disorders (2nd ed., (1995). A case-controlled study of repetitive
38 Diagnosis and Classification

thoughts and behavior in adults with autistic Evidence from a family history study of multiple-
disorder and obsessive-compulsive disorder. incidence autism families. American Journal
American Journal of Psychiatry, 152(5), of Psychiatry, 154(2), 185–190.
772–777. Poustka, F., & Lisch, S. (1993). Autistic behaviour
McDougle, C. J., Price, L. H., Volkmar, F. R., & domains and their relation to self-injurious be-
Goodman, W. K. E. (1992). Clomipramine in haviour. Acta Paedopsychiatrica: International
autism: Preliminary evidence of efficacy. Journal of Child and Adolescent Psychiatry,
Journal of the American Academy of Child and 56(2), 69–73.
Adolescent Psychiatry, 31(4), 746–750. Prior, M., Boulton, D., Gajzago, C., & Perry, D.
Miller, J. N., & Ozonoff, S. (1997). Did Asperger’s (1975). The classification of childhood psy-
cases have Asperger disorder? A research choses by numerical taxonomy. Journal of
note. Journal of Child Psychology and Psychia- Child Psychology and Psychiatry, 16(4),
try, 38(2), 247–251. 321–330.
Miller, J. N., & Ozonoff, S. (2000). The external va- Provence, S., & Dahl, K. (1987). Disorders of atyp-
lidity of Asperger disorder: Lack of evidence ical development: Diagnostic issues raised by a
from the domain of neuropsychology. Journal spectrum disorder. In D. Cohen & A. Donnel-
of Abnormal Psychology, 109(2), 227–238. lan (Eds.), Handbook of autism and pervasive
Moore, V., & Goodson, S. (2003). How well does developmental disorders (pp. 677–689). New
early diagnosis of autism stand the test of York: Wiley.
time? Follow-up study of children assessed for Quintana, H., Birmaher, B., Stedge, D., Lennon, S.,
autism at age 2 and development of an early di- Freed, J., Bridge, J., et al. (1995). Use of
agnostic service. Autism: International Jour- methylphenidate in the treatment of children
nal of Research and Practice, 7(1), 47–63. with autistic disorder. Journal of Autism and
Mundy, P., Sigman, M. D., Ungerer, J., & Sherman, Developmental Disorders, 25(3), 283–294.
T. (1986). Defining the social deficits of Rank, B. (1949). Adaptation of the psychoanalytic
autism: The contribution of non-verbal. Jour- technique for the treatment of young children
nal of Child Psychology and Psychiatry and with atypical development. American Journal
Allied Disciplines, 27(5), 657–669. of Orthopsychiatry, 19, 130–139.
National Center for Clinical Infant Programs. Rapin, I. (1991). Autistic children: Diagnosis and
(1994). Diagnostic classification: 0–3. Na- clinical features. Pediatrics, 87(5, Pt. 2),
tional Center for Clinical Infant Programs. 751–760.
Washington, DC: Author. Rapin, I., & Allen, D. A. (1983). Developmental
National Research Council. (2001). Educating language disorders: Nosologic considerations.
young children with Autism. Washington, DC: In U. Kirk (Ed.), Neuropsychology of language,
National Academy Press. reading and spelling (pp. 155–184). New York:
Nelson, E. C., & Pribor, E. F. (1993). A calendar Academic Press.
savant with autism and Tourette syndrome: Realmuto, G. M., & Main, B. (1982). Coincidence
Response to treatment and thoughts on the in- of Tourette’s disorder and infantile autism.
terrelationships of these conditions. Annals of Journal of Autism and Developmental Disor-
Clinical Psychiatry, 5(2), 135–140. ders, 12(4), 367–372.
Ozonoff, S., Pennington, B. F., & Rogers, S. J. Rett, A. (1966). Uber ein eigenartiges hirn-
(1991). Executive function deficits in high- tophisces Syndroem bei hyperammonie im
functioning autistic individuals: Relationship Kindersalter. Wein Medizinische Wochen-
to theory of mind. Journal of Child Psychology schrift, 118, 723–726.
and Psychiatry, 32(7), 1081–1105. Ritvo, E. R., & Freeman, B. J. (1977). National So-
Petty, L. K., Ornitz, E. M., Michelman, J. D., & ciety for Autistic Children definition of the
Zimmerman, E. G. (1985). Autistic children syndrome of. Journal of Pediatric Psychology,
who become schizophrenic. Annual Progress in 2(4), 142–145.
Child Psychiatry and Child Development. Rogers, S. (2001). Diagnosis of autism before the
Pickles, A., Starr, E., Kazak, S., Bolton, P., Pa- age of 3. International review of research in
panikolaou, K., Bailey, A., et al. (2000). Vari- mental retardation: Autism (23), 1–31.
able expression of the autism broader Rogers, S., & Ozonoff, S. (in press). What do we
phenotype: Findings from extended pedigrees. know about sensory dysfunction in autism? A
Journal of Child Psychology and Psychiatry critical review of the empirical evidence. Jour-
and Allied Disciplines, 41(4), 491–502. nal of Child Psychology and Psychiatry.
Piven, J., Palmer, P., Jacobi, D., Childress, D., & Rumsey, J. M., Rapoport, J. L., & Sceery, W. R.
Arndt, S. (1997). Broader autism phenotype: (1985). Autistic children as adults: Psychiatric,
Issues in the Classification of Autism and Related Conditions 39

social, and behavioral. Journal of the American Sartorius, N. (1988). International perspectives of
Academy of Child Psychiatry, 24(4), 465–473. psychiatric classification. British Journal of
Rutter, M. (1970). Autistic children: Infancy to Psychiatry, 152(Suppl.), 9–14.
adulthood. Seminars in Psychiatry, 2, 435–450. Shao, Y., Raiford, K. L., Wolpert, C. M., Cope,
Rutter, M. (1978). Diagnostic validity in child psy- H. A., Ravan, S. A., Ashley-Koch, A. A., et al.
chiatry. Advances in Biological Psychiatry, 2, (2002). Phenotypic homogeneity provides in-
2–22. creased support for linkage on chromosome 2
Rutter, M. (1994). Debate and argument: There are in autistic disorder. American Journal of
connections between brain and mind and it is Human Genetics, 70(4), 1058–1061.
important that Rett syndrome be classified Short, A., & Schopler, E. (1988). Factors relat-
somewhere. Journal of Child Psychology and ing to age of onset in autism. Journal of
Psychiatry and Allied Disciplines, 35(2), Autism and Developmental Disorders, 18,
379–381. 207–216.
Rutter, M. (1996). Autism research: Prospects and Siegel, B., Anders, T. F., Ciaranello, R. D., Bienen-
priorities. Journal of Autism and Developmen- stock, B., & Kraemer, H. C. (1986). Empiri-
tal Disorders, 26, 257–276. cally derived subclassification of the autistic
Rutter, M. (1999). The Emmanual Miller Memor- syndrome. Journal of Autism and Developmen-
ial Lectur 1998: Autism: Two-way interplay tal Disorders, 16(3), 275–293.
between research and clinical world. Journal Siegel, B., Pliner, C., Eschler, J., & Elliott, G. R.
of Child Psychology and Psychiatry, 40, (1988). How children with autism are diag-
169–188. nosed: Difficulties in identification of chil-
Rutter, M. (2000). Genetic studies of autism: From dren with multiple developmental delays.
the 1970s into the millennium. Journal of Ab- Journal of Developmental Behavioral Pedi-
normal Child Psychology, 28(1), 3–14. atrics, 9(4), 199–204.
Rutter, M. (2002). Classification: Conceptual is- Siegel, B., Vukicevic, J., Elliott, G. R., & Kraemer,
sues and substantive findings. In E. Taylor & H. C. (1989). The use of signal detection the-
M. Rutter (Eds.), Child and adolescent psychi- ory to assess DSM-III-R criteria for autistic
atry (4th ed., pp. 3–17). Oxford, England: disorder. Journal of the American Academy of
Blackwell. Child and Adolescent Psychiatry, 28(4),
Rutter, M., Bailey, A., Bolton, P., & Le Couter, A. 542–548.
(1994). Autism and known medical conditions: Simon, N. (1978). Kaspar Hauser’s recovery and
Myth and substance. Journal of Child Psychol- autopsy: A perspective on neurological and
ogy and Psychiatry and Allied Disciplines, sociological requirements for language devel-
35(2), 311–322. opment. Journal of Autism and Childhood
Rutter, M., & Garmezy, N. (1983). Developmental Schizophrenia, 8(2), 209–217.
psychopathology. In E. M. Hetherington (Ed.), Siperstein, R., & Volkmar, F. R. (2004). Parental
Mussen’s handbook of child psychology: Vol. 4. reporting of regression in children with perva-
Socialization, personality and child develop- sive developmental disorders. Journal of
ment (pp. 755–911). New York: Wiley. Autism and Developmental Disorders.
Rutter M., & Gould, M. (1985). Classification. In Soper, H. V., Satz, P., Orsini, D. L., Henry, R. R.,
M. Rutter & L. Hersov (Eds.), Child and Zvi, J. C., & Schulman, M. (1986). Handed-
Adolescent Psychiatry: Modern Approaches ness patterns in autism suggest subtypes. Jour-
(2nd ed., pp. 304–321). Cambridge, MA: nal of Autism and Developmental Disorders,
Blackwell. 16(2), 155–167.
Rutter, M., Le Couteur, A., & Lord, C. (2003). Spitzer, R. L., Endicott, J. E., & Robins, E. (1978).
Autism diagnostic interview-Revised. Los An- Research diagnostic criteria. Archives of Gen-
geles: Western Psychological Services. eral Psychiatry, 35, 773–782.
Rutter, M., & Schopler, E. (1992). Classification of Spitzer, R. L., & Siegel, B. (1990). The DSM-III-R
pervasive developmental disorders: Some con- field trial of pervasive developmental disor-
cepts and practical considerations [Special ders. Journal of the American Academy of
issue: Classification and diagnosis]. Journal of Child and Adolescent Psychiatry, 29(6),
Autism and Developmental Disorders, 22(4), 855–862.
459–482. Spitzer, R. L., & Williams, J. B. W. (1980). Classi-
Rutter, M., Shaffer, D., & Shepherd, M. (1975). A fication in psychiatry. In H. Kaplan, A. Freed-
multiaxial classification of child psychiatric man, & B. Sadock (Eds.), The comprehensive
disorders. Geneva, Switzerland: World Health textbook of psychiatry (3rd ed.). Baltimore:
Organization. Williams & Wilkins.
40 Diagnosis and Classification

Sponheim, E. (1996). Changing criteria of autistic of the American Academy of Child and Adoles-
disorders: A comparison of the ICD-10 re- cent Psychiatry, 42(7), 864–872.
search criteria and DSM-IV with DSM-III-R, Tanguay, P. E., Robertson, J., & Derrick, A. (1998).
CARS, and ABC. Journal of Autism and Devel- A dimensional classification of autism spectrum
opmental Disorders, 26(5), 513–525. disorder by social communication domains.
Sponheim, E., & Skjeldal, O. (1998). Autism and Journal of the American Academy of Child and
related disorders: Epidemiological findings in Adolescent Psychiatry, 37(3), 271–277.
a Norwegian study using ICD-10 diagnostic Towbin, K. E. (2003). Strategies for pharmaco-
criteria. Journal of Autism and Developmental logic treatment of high functioning autism
Disorders, 28(3), 217–227. and Asperger syndrome. Child and Adolescent
Steingard, R., & Biederman, J. (1987). Lithium re- Psychiatric Clinics of North America, 12(1)
sponsive manic-like symptoms in two individ- 23–45.
uals with autism. Journal of the American Tsai, L. (1992). Is Rett syndrome a subtype of per-
Academy of Child and Adolescent Psychiatry, vasive developmental disorder? Journal of
26(6), 932–935. Autism and Developmental Disorders, 22,
Stone, W. L., Lee, E. B., Ashford, L., Brissie, J., 551–561.
Hepburn, S. L., Coonrod, E. E., et al. (1999). Tsai, L. (1996). Brief report: Comorbid psychiatric
Can autism be diagnosed accurately in chil- disorders of autistic disorders. Journal of
dren under 3 years? Journal of Child Psychol- Autism and Developmental Disorders, 26,
ogy and Psychiatry and Allied Disciplines, 159–164.
40(2), 219–226. Tsai, L. Y. (1999). Psychopharmacology in autism.
Szatmari, P. (1992a). A review of the DSM-III-R Psychosomatic Medicine, 61(5), 651–665.
criteria for autistic disorder [Special issue: Tsai, L. Y., Tsai, M. C., & August, G. J. (1985).
Classification and diagnosis]. Journal of Brief report: Implication of EEG diagnoses in
Autism and Developmental Disorders, 22(4), the subclassification of infantile autism. Jour-
507–523. nal of Autism and Developmental Disorders,
Szatmari, P. (1992b). The validity of autistic spec- 15(3), 339–344.
trum disorders: A literature review [Special Van der Gaag, R. J., Buitelaar, J., Van den Ban, E.,
issue: Classification and diagnosis]. Journal of Bezemer, M., Njio, L., & Van Engeland, H.
Autism and Developmental Disorders, 22(4), (1995). A controlled multivariate chart review
583–600. of multiple complex developmental disorder.
Szatmari, P. (2000). Perspectives on the classifi- Journal of the American Academy of Child and
cation of Asperger syndrome. In A. Klin & Adolescent Psychiatry, 34(8), 1096–1106.
F. R. Volkmar (Eds.), Asperger syndrome Volkmar, F. R. (1992). Childhood disintegrative
(pp. 403–417). New York: Guilford Press. disorder: Issues for DSM-IV. Journal of Autism
Szatmari, P., Bryson, S. E., Boyle, M. H., Streiner, and Developmental Disorders, 22(4), 625–642.
D. L., & Duku, E. (2003). Predictors of out- Volkmar, F. R., Cicchetti, D. V., Bregman, J., &
come among high functioning children with Cohen, D. J. (1992). Three diagnostic systems
autism and Asperger syndrome. Journal of for autism: DSM-III, DSM-III-R, and ICD-10
Child Psychology and Psychiatry and Allied [Special issue: Classification and diagnosis].
Disciplines, 44(4), 520–528. Journal of Autism and Developmental Disor-
Szatmari, P., Merette, C., Bryson, S. E., Thivierge, ders, 22(4), 483–492.
J., Roy, M. A., Cayer, M., et al. (2002). Quan- Volkmar, F. R., Cicchetti, D. V., Cohen, D. J., &
tifying dimensions in autism: a factor-analytic Bregman, J. (1992). Brief report: Developmen-
study [Comment]. Journal of the American tal aspects of DSM-III-R criteria for autism
Academy of Child and Adolescent Psychiatry, [Special issue: Classification and diagnosis].
41(4), 467–474. Journal of Autism and Developmental Disor-
Szatmari, P., Volkmar, F., & Walter, S. (1995). ders, 22(4), 657–662.
Evaluation of diagnostic criteria for autism Volkmar, F. R., & Cohen, D. J. (1989). Disintegra-
using latent class models. Journal of the Amer- tive disorder or “late onset ” autism. Journal of
ican Academy of Child and Adolescent Psychi- Child Psychology and Psychiatry and Allied
atry, 34(2), 216–222. Disciplines, 30(5), 717–724.
Tadevosyan-Leyfer, O., Dowd, M., Mankoski, R., Volkmar, F. R., & Cohen, D. J. (1991a). Comorbid
Winklosky, B., Putnam, S., McGrath, L., et al. association of autism and schizophrenia.
(2003). A principal components analysis of the American Journal of Psychiatry, 148(12),
Autism Diagnostic Interview-Revised. Journal 1705–1707.
Issues in the Classification of Autism and Related Conditions 41

Volkmar, F. R., & Cohen, D. J. (1991b). Debate and (Ed.), Child and adolescent psychiatry: A com-
argument: The utility of the term pervasive prehensive textbook (3rd ed., pp. 499–505).
developmental disorder. Journal of Child Psy- Baltimore: Williams & Wilkins.
chology and Psychiatry and Allied Disciplines, Volkmar, F. R., Stier, D. M., & Cohen, D. J. (1985).
32(7), 1171–1172. Age of recognition of pervasive developmental
Volkmar, F., Cohen, D., Hoshino, K., Rende, R., & disorder. American Journal of Psychiatry,
Paul, R. (1988). Phenomenology and classifi- 142(12), 1450–1452.
cation of the childhood psychoses. Psychologi- Waterhouse, L., Morris, R., Allen, D., Dunn, M.,
cal Medicine, 18(1), 191–201. Fein, D., Feinstein, C., et al. (1996). Diagnosis
Volkmar, F. R., & Klin, A. (1998). Asperger syn- and classification in autism. Journal of Autism
drome and nonverbal learning disabilities. In and Developmental Disorders, 26(1), 59–86.
E. Schopler & G. B. Mesibov (Eds.), Asperger Waterhouse, L., Wing, L., Spitzer, R., & Siegel, B.
syndrome or high-functioning autism? Current (1993). Pervasive developmental disorders:
issues in autism (pp. 107–121). New York: From DSM-III to DSM-III-R [Special issue:
Plenum Press. Classification and diagnosis]. Journal of
Volkmar, F. R., Klin, A., & Pauls, D. (1998). Noso- Autism and Developmental Disorders, 22(4),
logical and genetic aspects of Asperger Syn- 525–549.
drome. Journal of Autism and Developmental Weir, K., & Salisbury, D. M. (1980). Acute onset of
Disorders, 28(5), 457–463. autistic features following brain damage in a
Volkmar, F. R., Klin, A., Siegel, B., Szatmari, P., ten-year-old. Journal of Autism and Develop-
Lord, C., Campbell, M., et al. (1994). Field mental Disorders, 10, 185–191.
trial for autistic disorder in DSM-IV. American Wing, L. (1980). Childhood autism and social
Journal of Psychiatry, 151(9), 1361–1367. class: A question of selection? British Journal
Volkmar, F. R., Lord, C., Bailey, A., Schultz, R. T., of Psychiatry, 137, 410–417.
& Klin, A. J. (2004). Autism and Pervasive Wing, L. (1981). Asperger’s syndrome: A clinical
Developmental Disorders. Journal of Child account. Psychological Medicine, 11(1),
Psychology and Psychiatry, 45(1), 135–170. 115–129.
Volkmar, F. R., & Nelson, D. S. (1990). Seizure Wing, L., & Gould, J. (1979). Severe impairments
disorders in autism. Journal of the American of social interaction and associated abnormal-
Academy of Child and Adolescent Psychiatry, ities in children: Epidemiology and classifica-
29(1), 127–129. tion. Journal of Autism and Developmental
Volkmar, F. R., & Rutter, M. (1995). Childhood Disorders, 9(1), 11–29.
disintegrative disorder: Results of the DSM-IV Wolff, S. (1998). Schizoid personality in child-
Autism Field Trial. Journal of the American hood: The links with Asperger syndrome,
Academy of Child and Adolescent Psychiatry, schizophrenia spectrum disorders, and elec-
34(8), 1092–1095. tive mutism. In E. Schopler & G. B. Mesibov
Volkmar, F. R., & Schwab-Stone, M. (1996). Anno- (Eds.), Asperger syndrome or high-function-
tation: Childhood disorders in DSM-IV. Jour- ing autism? Current issues in autism
nal of Child Psychology and Psychiatry and (pp. 123–142). New York: Plenum Press.
Allied Disciplines, 37(7), 779–784. World Health Organization. (1992). International
Volkmar F. R., Schwab-Stone, M., & First, M. classification of diseases: Diagnostic criteria
(2002). Classification in child and adolescent for research (10th ed.). Geneva, Switzerland:
Psychiatry: Principles and issues. In M. Lewis Author.

Epidemiological Studies of Pervasive

Developmental Disorders


Epidemiological surveys of autism started of autism-spectrum disorders. This chapter

in the mid-1960s in England (Lotter, 1966, addresses the following five questions:
1967) and have since been conducted in many
countries. Most of these surveys have focused 1. What is the range of prevalence estimates
on a categorical-diagnostic approach to autism for autism and related disorders?
that has relied over time on different sets of 2. What proportion of autism cases is attrib-
criteria. All surveys, however, used a defini- utable to specific associated medical dis-
tion of autism that comprised severe impair- orders?
ments in communication and language, social 3. Is the incidence of autism increasing?
interactions, and play and behavior. This chap- 4. What are the other correlates of autistic-
ter focuses on autism defined as a severe de- spectrum disorders, particularly with re-
velopmental disorder. It does not deal with spect to race and ethnicity?
subtle autistic features or symptoms that 5. What is the role, if any, of cluster reports in
occur as part of other, more specific, develop- causal investigations of autism?
mental disorders, as unusual personality traits,
or as components of the lesser variant of DESIGN OF
autism thought to index genetic liability to EPIDEMIOLOGICAL STUDIES
autism in relatives. With the exception of re-
cent studies, other pervasive developmental Epidemiology is concerned with the study of
disorders (PDD) falling short of diagnostic the repartition of diseases in human popula-
criteria for autistic disorder (PDDNOS, As- tions and of the factors that influence it. Epi-
perger syndrome) were generally not included demiologists use several measures of disease
in the case definition used in earlier surveys occurrence. Incidence rate refers to the number
although several epidemiological investiga- of new cases (numerator) of a disease occurring
tions yielded useful information on the rates over a specified period in those at risk of devel-
of these particular PDDs. These data are sum- oping the disease in the population (denomina-
marized separately. This chapter provides an tor, in person × years). Cumulative incidence
up-to-date review of the methodological fea- is the proportion of those who were free of the
tures and substantive results of published epi- disease at the beginning of the observation
demiological surveys. It also updates our period and developed the disease during that
previous review (Fombonne, 2003a) with the period. Measures of incidence are required to
inclusion of eight new studies made available properly estimate morbidity due to a disease,
since then. A key feature of the review was to its possible changes over time, and the risk fac-
rely on summary statistics throughout to derive tors underlying disease status. Prevalence is a
quantitative estimates for rates and correlates measure used in cross-sectional surveys (there

Epidemiological Studies of Pervasive Developmental Disorders 43

is no passage of time) and reflects the propor- the concept of autism, a shift from autism to
tion of subjects in a given population who, PDD, the recognition of autism in subjects of
at that point in time, suffer from the disease. normal intelligence, and other similar factors
Most epidemiological studies of autism have will all contribute to increase (a) (subjects who
been cross-sectional and are not informative on are now regarded as cases, a, whereas, previ-
incidence (with a few recent exceptions). As a ously, they were included in b, or even d). Thus,
result, prevalence rates have been used to de- even in the absence of a change in the incidence
scribe autism in populations. of the disorder, prevalence estimates (a /N) can
It is useful to summarize how data are col- go up merely for methodological reasons.
lected in a prevalence study (refer to Table 2.1;
and also Fombonne, 2002a). The investigators Selection of Studies
first select a population of a given size (N),
often in a circumscribed geographic area. The studies were identified through systematic
Then, one or more screening stages are orga- searches from the major scientific literature
nized to identify possible cases designated as databases (MEDLINE, PSYCINFO) and from
screen positives (a + b). In a second, diagnostic prior reviews (Fombonne, 1999, 2003a; Wing,
stage, the screen positives (a + b) undergo a 1993). Only studies published in the English
thorough evaluation and are finally classified language were included. Surveys that relied
as cases (a, or true positives) or noncases ( b, on a questionnaire-based approach to define
false positives). The probability that a screen whether a subject was a case or not a case were
positive is a case (a /a + b) is called the positive also excluded because the validity of the diag-
predictive value. The prevalence is then calcu- nosis is unsatisfactory in these studies. Overall,
lated by dividing the number of cases identi- 42 studies published between 1966 and 2003
fied in the diagnostic stage by the size of the were selected that surveyed PDDs in clearly de-
population (a /N). However, the imperfection of marcated, nonoverlapping samples. Of these, 36
the screening process means that this calcula- studies provided information on rates of autis-
tion does not take into account the false nega- tic disorder, 3 studies provided estimates only
tives (c), true cases who were missed in the on all PDDs combined, and 3 studies provided
screening stage. In published autism surveys, data only on high-functioning PDDs. For sev-
there is often no way to estimate (c), although eral studies, the publication listed in the tables
techniques exist that could allow for this. As a is the most detailed account or the earliest one.
result, the prevalence estimate can be seri- When appropriate, however, other published ar-
ously underestimated. When comparing sur- ticles were used to extract relevant information
veys over time, two factors may jeopardize the from the same study.
comparison. Better awareness of the disorder,
improved screening techniques, and detection Survey Descriptions
all contribute to reduce the false negatives (c)
(and as a consequence to increase a). Changes Surveys were conducted in 14 countries, and
in case definition, especially a broadening of half of the results have been published since
1997. Details on the precise sociodemographic
composition and economical activities of the
Table 2.1 Hypothetical Prevalence Study area surveyed in each study were generally
of Autism lacking. Most studies were conducted in pre-
Cases Noncases dominantly urban or mixed areas, however,
(D) (D) with only two surveys (6 and 11) carried out in
Screen positive a b a+b
predominantly rural areas. The proportion of
Screen negative c d c+d children from immigrant families was generally
a+c b+d a+b+c+d=N not available and very low in five surveyed pop-
False negatives (FN) = d; False positives (FP) = b; Pos-
ulations (Studies 11, 12, 19, 23, and 26). Only
itive predictive value (PPV); Prevalence (P) = a + c/N = in Studies 4, 34, and 38 was there a substantial
p(D); Sensitivity (Se, rate of true positives) = a /a + c; minority of children with either an immigrant
Specificity (Sp, rate of true negatives) = d/ b + d. or different ethnic background living in the
44 Diagnosis and Classification

area. The age range of the population included (Study 25) to 35% (Study 40), with a median
in the surveys is spread from birth to early adult value of 14%. Fewer studies could examine
life, with an overall median age of 8.0. Simi- the extent to which uncooperative participa-
larly, in 39 studies, there is huge variation in tion or outright refusal to participate in sur-
the size of the population surveyed (range: veys is associated with the likelihood that the
826–4,590,000), with a median population size corresponding children have autism. Bryson,
of 63,860 subjects (mean = 255,000). About Clark, and Smith (1988; Study 12) provided
half of the studies relied on targeted popula- some evidence that those families who refused
tions ranging in size from 15,870 to 166,860. to cooperate in the intensive assessment phase
had children with ABC scores similar to other
Study Designs false positives in their study, suggesting that
these children were unlikely to have autism.
A few studies have relied on existing admin- Webb, Morey, et al. (2003; Study 40) similarly
istrative databases (i.e., Croen, Grether, produced data showing increasing refusal rate
Hoogstrate, & Selvin, 2002; Gurney et al., in those with fewer ICD-10 PDD symptoms.
2003) or on national registers (Madsen et al., By contrast, in a Japanese study (Sugiyama &
2002) for case identification. Most investiga- Abe, 1989; Study 13) where 17.3% of parents
tions have relied on a two-stage or multistage refused further investigations for their 18-
approach to identify cases in underlying popu- month-old children who had failed a devel-
lations. The first screening stage of these opmental check, follow-up data at age 3 sug-
studies often consisted of sending letters or gested that half of these children still dis-
brief screening scales requesting school and played developmental problems. Whether
health professionals to identify possible cases these problems were connected to autism is
of autism. Each investigation varied in several unknown, but this study points to the possibil-
key aspects of this screening stage. First, the ity that higher rates of developmental disor-
coverage of the population varied enormously ders exist among nonparticipants to surveys.
from one study to another. In some studies Similarly, in Lotter’s study (1966; Study 1),
(3, 17, 20, 24, 33), only cases already known 58 questionnaires covering schools for handi-
from educational or medical authorities could capped children were returned out of the 76
be identified. In other surveys, investigators forms sent out, and an independent review of
achieved extensive coverage of the entire the records showed that 4 of the 18 missing
population, including children attending nor- forms corresponded to autistic children. It
mal schools (Studies 1, 25, 40) or children un- is difficult to draw firm conclusions from
dergoing systematic developmental checks these different accounts. Although there is no
(Studies 13, 19, 22, 32, 36). In addition, the consistent evidence that parental refusal to co-
surveyed areas varied in terms of service de- operate is associated with autism in their off-
velopment as a function of the specific educa- spring, a small proportion of cases may be
tional or health care systems of each country missed in some surveys as a consequence of
and of the year of investigation. Second, the noncooperation at the screening stage. One
type of information sent out to professionals study (40) included a weighting procedure to
invited to identify children varied from simple compensate for nonresponse.
letters including a few clinical descriptors Only two studies (1 and 30) provided an
of autism-related symptoms or diagnostic estimate of the reliability of the screening
checklists rephrased in nontechnical terms, to procedure. The sensitivity of the screening
more systematic screening based on question- methodology (a /(a + c) in Table 2.1) is also dif-
naires or rating scales of known reliability and ficult to gauge in autism surveys. The usual
validity. Third, participation rates in the first epidemiological approach of sampling screened
screening stages provide another source of negative subjects at random to estimate the
variation in the screening efficiency of sur- proportion of false negatives (c/(a + c) in Table
veys. Refusal rates were available for 13 stud- 2.1) has not been used in these surveys because
ies (1, 5, 6, 9, 12, 14, 19, 20, 23, 25, 30, 37, the low frequency of the disorder would make
and 40); the rate of refusal ranged from 0% undertaking such estimations both imprecise
Epidemiological Studies of Pervasive Developmental Disorders 45

and costly. The cases that were missed as a re- case determination. It is, furthermore, difficult
sult of noncooperation or imperfect sensitivity to assess the impact of a specific diagnostic
of the screening procedure make it necessary scheme or of a particular diagnostic criterion
to view the prevalence estimates as underesti- on the estimate of prevalence since other pow-
mates of the true rates. The magnitude of this erful method factors confound between-studies
underestimation is unknown in each survey. comparisons of rates. Surprisingly, few studies
Similar considerations about the method- have built in a reliability assessment of the di-
ological variability across studies apply to agnostic procedure; reliability during the in-
the intensive assessment phases. Participation tensive assessment phase was high in seven
rates in these second-stage assessments were surveys (4, 13, 16, 23, 24, 32, 36) and moder-
not always available, either because they had ate in another one (14).
simply not been calculated, or because the de-
sign and/or method of data collection did not CHARACTERISTICS OF
lead easily to their estimation. When available AUTISTIC SAMPLES
(Studies 1, 5, 8, 12, 13, 15, 22, 23, 25, 29, 30,
32, 36), they were generally high, ranging from Data on children with autistic disorders were
76.1% (Study 12) to 98.6% (Study 25). The in- available in 36 surveys (1 to 36; see Table 2.2).
formation used to determine final diagnostic In total, 7,514 subjects were considered to suf-
status usually involved a combination of infor- fer from autism; this number ranged from 6
mants and data sources, with a direct assess- (Studies 18 and 25) to 5,038 (Study 34) across
ment of the person with autism in 21 studies. studies (median: 48; mean: 209). An assessment
The assessments were conducted with vari- of intellectual function was obtained in 21
ous diagnostic instruments, ranging from a studies. These assessments were conducted with
classical clinical examination to the use of bat- various tests and instruments; furthermore, re-
teries of standardized measures. The Autism sults were pooled in broad bands of intellectual
Diagnostic Interview (Le Couteur et al., 1989) level that did not share the same boundaries
and/or the Autism Diagnostic Observational across studies. As a consequence, differences
Schedule (Lord, Risi, et al., 2000) were used in rates of cognitive impairment between stud-
in the most recent surveys. The precise diag- ies should be interpreted with caution. Despite
nostic criteria retained to define caseness vary these caveats, some general conclusions can be
according to the study and, to a large extent, reached (Table 2.2). The median proportion of
reflect historical changes in classification sys- subjects without intellectual impairment is
tems. Thus, Kanner’s criteria and Lotter’s and 29.6% (range: 0% to 60%).1 The corresponding
Rutter’s definitions were used in Studies 1 to figures are 29.3% (range: 6.6% to 100%) for
8 (all conducted before 1982), whereas DSM- mild-to-moderate intellectual impairments,
based definitions took over thereafter as well as and 38.5% (range: 0% to 81.3%) for severe-to-
ICD-10 since 1990. Some studies have relaxed profound mental retardation. Gender reparti-
partially some diagnostic criteria such as an age tion among subjects with autism was reported
of onset before 30 months (Study 6) or the ab- in 32 studies totaling 6,963 subjects with
sence of schizophrenic-like symptoms (Studies autism, and the male/female sex ratio varied
13 and 14). However, most surveys have relied from 1.33 (Study 7) to 16.0 (Study 4), with a
on the clinical judgment of experts to arrive at mean malefemale ratio of 4.31. Thus, no epi-
the final case groupings. It is worth underlining demiological study ever identified more girls
that field trials for recent classifications such than boys with autism, a finding that parallels
as DSM-III-R (Spitzer & Siegel, 1990) or DSM- the gender differences found in clinically re-
IV/ICD-10 (Volkmar, Klin, et al., 1994) have ferred samples (Lord, Schopler, & Revecki,
also relied on the judgment of clinical experts, 1982). Gender differences were more pro-
taken as a gold standard to diagnose autism nounced when autism was not associated with
and calibrate diagnostic algorithms. There-
fore, the heterogeneity of diagnostic criteria
used across surveys is somewhat mitigated by 1
Study 23, which relied on different IQ groupings,
reliance on expert clinical judgment for final has been excluded.
TABLE 2.2 Prevalence Surveys of Autistic Disorder
Number of
Size of Subjects Percentage Gender
Target with Diagnostic with Normal Ratio Prevalence
No. Study Country Area Population Age Autism Criteria IQ (MF) Rate/10,000 95% CI

1 Lotter, 1966 United Kingdom Middlesex 78,000 8–10 32 Rating scale 15.6 2.6 (23/9) 4.1 2.7 ; 5.5
2 Brask, 1970 Denmark Aarhus 46,500 2–14 20 Clinical — 1.4 (12/7) 4.3 2.4 ; 6.2
3 Treffert, 1970 United States Wisconsin 899,750 3–12 69 Kanner — 3.06 0.7 0.6 ; 0.9
4 Wing, Yeates, United Kingdom Camberwell 25,000 5–14 17a 24 items rating 30 16 (16/1) 4.8 b 2.1 ; 7.5
Brierly, & Gould, scale of Lotter
5 Hoshino et al., 1982 Japan Fukushima- 609,848 0–18 142 Kanner’s — 9.9 2.33 1.9 ; 2.7
Ken criteria (129/13)

6 Bohman, Bohman, Sweden County of 69,000 0–20 39 Rutter criteria 20.5 1.6 5.6 3.9 ; 7.4
Björck, & Sjöholm, Västerbotten (24/15)
7 McCarthy, Ireland East 65,000 8–10 28 Kanner — 1.33 4.3 2.7 ; 5.9
Fitzgerald, & Smith, (16/12)
8 Steinhausen, Göbel, Germany West Berlin 279,616 0–14 52 Rutter 55.8 2.25 1.9 1.4 ; 2.4
Breinlinger, & (36/16)
Wohlloben, 1986
9 Burd, Fisher, & United States North Dakota 180,986 2–18 59 DSM-III — 2.7 3.26 2.4 ; 4.1
Kerbeshan, 1987 (43/16)
10 Matsuishi et al., Japan Kurume City 32,834 4–12 51 DSM-III — 4.7 (42/9) 15.5 11.3 ; 19.8
11 Tanoue, Oda, Asano, Japan Southern 95,394 7 132 DSM-III — 4.07 13.8 11.5 ; 16.2
& Kawashima, 1988 Ibaraki (106/26)
12 Bryson, Clark, & Canada Part of Nova- 20,800 6–14 21 New RDC 23.8 2.5 (15/6) 10.1 5.8 ; 14.4
Smith, 1988 Scotia
13 Sugiyama & Abe, Japan Nagoya 12,263 3 16 DSM-III — — 13.0 6.7 ; 19.4
14 Cialdella & France 1 district 135,180 3–9 61 DSM-III-like — 2.3 4.5 3.4 ; 5.6
Mamelle, 1989 (Rhône)
15 Ritvo et al., 1989 United States Utah 769,620 3–27 241 DSM-III 34 3.73 2.47 2.1 ; 2.8
16 Gillberg, Sweden South-West 78,106 4–13 74 DSM-III-R 18 2.7 9.5 7.3 ; 11.6
Steffenburg, & Gothenburg + (54/20)
Schaumann, 1991d Bohuslän
17 Fombonne & du France 4 regions, 14 274,816 9 and 13 154 Clinical- 13.3 2.1 4.9 4.1 ; 5.7
Mazaubrun, 1992 districts ICD-10-like (105/49)
18 Wignyosumarto, Indonesia Yogyakarita 5,120 4–7 6 CARS 0 2.0 (4/2) 11.7 2.3 ; 21.1
Mukhlas, & (SE of
Skirataki, 1992 Jakarta)
19 Honda, Shimizu, Japan Yokohama 8,537 5 18 ICD-10 50.0 2.6 (13.5) 21.08 11.4 ; 30.8
Misumi, Niimi, &
Ohashi, 1996
20 Fombonne, du France 3 districts 325,347 8–16 174 Clinical 12.1 1.81 5.35 4.6 ; 6.1
Mazaubrun, Cans, & ICD-10-like (112/62)
Grandjean, 1997

TABLE 2.2 (Continued)

Number of
Size of Subjects Percentage Gender
Target with Diagnostic with Normal Ratio Prevalence
No. Study Country Area Population Age Autism Criteria IQ (MF) Rate/10,000 95% CI
21 Webb, Lobo, Hervas, United Kingdom South 73,301 3–15 53 DSM-III-R — 6.57 7.2 5.3 ; 9.3
Scourfield, & Fraser, Glamorgan, (46/7)
1997 Wales

22 Arvidsson, Sweden (West Mölnlycke 1,941 3–6 9 ICD-10 22.2 3.5 (7/2) 46.4 16.1 ; 76.6
Danielsson, coast)
Forsberg, Gillberg,
& Johansson, 1997
23 Sponheim & Norway Akershus 65,688 3–14 34 ICD-10 47.1 c 2.09 5.2 3.4 ; 6.9
Skjeldal, 1998 County (23/11)
24 Taylor et al., 1999 United Kingdom North 490,000 0–16 427 ICD-10 — — 8.7 7.9 ; 9.5
25 Kadesjö, Gillberg, & Sweden Karlstad 826 6.7–7.7 6 DSM-III-R / 50.0 5.0 (5/1) 72.6 14.7 ; 130.6
Hagberg, 1999 (Central) ICD-10
26 Baird, Charman, & United Kingdom South-East 16,235 — 50 ICD-10 60 15.7 30.8 22.9 ; 40.6
Baron-Cohen, 2000 Thames (47/3)
27 Powell et al., 2000 United Kingdom West 25,377 — 62 Clinical / — — 7.8 5.8 ; 10.5
Midlands ICD-10/DSM-IV

28 Kielinen, Linna, & Finland North (Oulu 152,732 — 187 ICD-8/ICD-9/ 49.8 4.12 12.2 10.5 ; 14.0
Moilanen, 2000 et Lapland) ICD-10 (156/50)
29 Bertrand et al., 2001 United States Brick 8,896 — 36 DSM-IV 36.7 2.2 40.5 28.0 ; 56.0
Township, (25/11)
New Jersey
30 Fombonne, Simmons, United Kingdom Angleterre et 10,438 5–15 27 DSM-IV/ 55.5 8.0 (24/3) 26.1 16.2 ; 36.0
Ford, Meltzer, & Pays de ICD-10
Goodman, 2001 Galles
31 Magnússon & Iceland Whole Island 43,153 5–14 57 Mostly ICD-10 15.8 4.2 13.2 9.8 ; 16.6
Saemundsen, 2001 (46/11)
32 Chakrabarti & United Kingdom Staffordshire 15,500 2.5–6.5 26 ICD-10/ 29.2 3.3 (20/6) 16.8 10.3 ; 23.2
Fombonne, 2001 (Midlands) DSM-IV
33 Davidovitch, Israel Haiffa 26,160 7–11 26 DSM-III-R / — 4.2 (21/5) 10.0 6.6 ; 14.4
Holtzman, & Tirosh, DSM-IV
34 Croen, Grether, United States California 4,950,333 5–12 5,038 CDER (Full 62.8 e 4.47 11.0 10.7 ; 11.3
Hoogstrate, & DDS syndrome) (4,116/921)
Selvin, 2002a
35 Madsen et al., 2002 Denmark National 63,859 8 46 ICD-10 — — 7.2 5.0–10.0
36 Chakrabarti & United Kingdom Staffordshire 10,903 4–7 24 ICD-10/ 33.3 3.8 (19/5) 22.0 14.4 ; 32.2
Fombonne, 2004 (Midlands) DSM-IV
This number corresponds to the sample described in Wing and Gould (1979).
This rate corresponds to the first published paper on this survey and is based on 12 subjects among children age 5 to 14 years.
In this study, mild mental retardation was combined with normal IQ, whereas moderate and severe mental retardation were grouped together.
For the Goteborg surveys by Gillberg et al. (Gillberg, 1984; Gillberg et al., 1991; Steffenburg & Gillberg, 1986), a detailed examination showed that there was overlap
among the samples included in the three surveys; consequently only the last survey has been included in this table.
This proportion is likely to be overestimated and to ref lect an underreporting of mental retardation in the CDER evaluations.

50 Diagnosis and Classification

mental retardation. In 13 studies (865 subjects) calculations, we arbitrarily adopted the mid-
where the sex ratio was available within the nor- point of this interval as the working rate for
mal band of intellectual functioning, the median autism prevalence, that is, the value of
sex ratio was 5.51. Conversely, in 12 studies 13/10,000.
(813 subjects), the median sex ratio was 1.951
in the group with autism and moderate-to- Associated Medical Conditions
severe mental retardation.
Rates of medical conditions associated with
Prevalence Estimations for autism were reported in 15 surveys and the
Autistic Disorder findings are summarized in Table 2.3. These
medical conditions were investigated by very
Prevalence estimates ranged from 0.7/10,000 different means ranging from questionnaires
to 72.6/10,000 (Table 2.2). Confidence inter- to full medical workups.
vals were computed for each estimate; their Conditions such as congenital rubella
width (difference between the upper and lower and PKU account for almost no cases of
limit of the 95% confidence interval) indicates autism. Prior studies suggesting an association
the variation in sample sizes and in the preci- of congenital rubella (Chess, 1971) and
sion achieved in each study (range: 0.3 −115.9; PKU (Knobloch & Pasamanick, 1975; Lowe,
mean = 11.3). Prevalence rates were nega- Tanaka, Seashore, Young, & Cohen, 1980)
tively correlated with sample size (Spearman with autism were conducted before implemen-
r = −.73; p < .01); small-scale studies tended tation of systematic prevention measures.
to report higher prevalence rates. Likewise, our nil estimate of 0% for autism
When surveys were combined in two groups and neurofibromatosis is consistent with
according to the median year of publication the 0.3% rate found in a large series of 341
(1994), the median prevalence rate for 18 sur- referred cases (Mouridsen, Bachmann-
veys published in the period 1966 to 1993 was Andersen, Sörensen, Rich, & Isager, 1992).
4.7/10,000, and the median rate for the 18 sur- Similarly, the rates found for cerebral palsy
veys published in the period 1994 to 2004 and Down syndrome equally suggest no partic-
was 12.7/10,000. Indeed, the correlation be- ular association. Recent reports (Bregman &
tween prevalence rate and year of publication Volkmar, 1988; Ghaziuddin, Tsai, & Ghaziud-
reached statistical significance (Spearman din, 1992; Howlin, Wing, & Gould, 1995) have
r = .65; p < .01); and the results of the 22 sur- focused on the co-occurrence of Down syn-
veys with prevalence rates over 7/10,000 were drome and autism in some individuals. The
all published since 1987. These findings point epidemiological findings give further support
toward an increase in prevalence estimates
in the past 15 to 20 years. To derive a best es-
timate of the current prevalence of autism, TABLE 2.3 Medical Disorders Associated with
it was therefore deemed appropriate to restrict Autism in Recent Epidemiological Surveys
the analysis to 28 surveys published since Number of Median
1987. The prevalence estimates ranged from Studies Rate Range
2.5 to 72.6/10,000 (average 95% CI width: Cerebral palsy 7 1.4 0–4.8
14.1), with an average rate of 16.2/10,000 and Fragile X 9 0.0 0–8.1
a median rate of 11.3/10,000. Similar values Tuberous sclerosis 11 1.1 0–3.8
were obtained when slightly different rules Phenylketonuria 8 0 0–0
and time cutpoints were used, with median Neurofibromatosis 7 0 0–1.4
and mean rates fluctuating between 10 and 13 Congenital rubella 11 0.0 0–5.9
Down syndrome 12 0.7 0–16.7
and 13 and 18/10,000 respectively. From
these results, a conservative estimate for the At least one disorder 16 5.5 0–16.7
current prevalence of autistic disorder is most Epilepsy 12 16.7 0–26.4
consistent with values lying somewhere be- Hearing deficits 8 1.3 0–5.9
Visual deficits 6 0.7 0–11.1
tween 10/10,000 and 16/10,000. For further
Epidemiological Studies of Pervasive Developmental Disorders 51

to the validity of these clinical descriptions figure for any medical disorder (excluding
(that the two conditions co-occur in some chil- epilepsy and sensory impairments). Although
dren), although they do not suggest that the this figure does not incorporate other medical
rate of comorbidity is higher than that ex- events of potential etiological significance,
pected by chance once the effects of mental such as encephalitis, congenital anomalies, and
retardation are taken into account. For fragile other rare medical syndromes, it is similar to
X, the low rate available in epidemiological that reported in a recent review of the question
studies is almost certainly an underestimate (Rutter, Bailey, Bolton, & Le Couteur, 1994). It
because fragile X was not recognized until rel- is worth noting that epidemiological surveys of
atively recently, and the most recent surveys autism in very large samples (Studies 15, 17,
did not always include systematic screening and 20) provided estimates in line with our
for fragile X. In line with prior reports (Smal- conservative summary statistics. By contrast,
ley, Tanguay, Smith, & Guitierrez, 1992), claims of average rates of medical conditions
tuberous sclerosis (TS) has a consistently high as high as 24% appear to apply to studies of
frequency among autistic samples. Assuming smaller size and to rely on a broadened defini-
a population prevalence of 1/10,000 for TS tion of autism (Gillberg & Coleman, 1996).
(Ahlsen, Gillberg, Lindblom, & Gillberg, Rates of epilepsy are high among autism sam-
1994; Hunt & Lindenbaum, 1984; Shepherd, ples. The proportion suffering from epilepsy
Beard, Gomez, Kurland, & Whisnant, 1991), it tends also to be higher in studies that have
appears that the rate of TS is about 100 times higher rates of severe mental retardation (as in
higher than that expected under the hypothesis Studies 16, 17, and 20). Age-specific rates for
of no association. Whether epilepsy, localized the prevalence of epilepsy were not available.
brain lesions, or direct genetic effects mediate The samples where high rates of epilepsy were
the association between TS and autism is a reported tended to have a higher median age, al-
matter for ongoing research (Smalley, 1998). though these rates seemed mostly to apply to
The overall proportion of cases of autism school-age children. Thus, in light of the in-
that could be causally attributed to known med- creased incidence of seizures during adoles-
ical disorders therefore remains low. From the cence among subjects with autism (Deykin &
16 surveys where rates of one of seven clear-cut MacMahon, 1979; Rutter, 1970), the epidemio-
medical disorders potentially causally associ- logical rates should be regarded as underesti-
ated with autism (cerebral palsy, fragile X, TS, mates of the lifetime risk of epilepsy in autism.
PKU, neurofibromatosis, congenital rubella, These rates are nonetheless high and support the
and Down syndrome) were available, we com- findings of a bimodal peak of incidence of
puted the proportion of subjects with at least epilepsy in autistic samples, with a first peak of
one of these recognizable disorders. Because incidence in the first years of life (Volkmar &
the overlap between these conditions is ex- Nelson, 1990).
pected to be low and because the information
about multiply-handicapped subjects was not RATES OF OTHER PERVASIVE
available, this overall rate was obtained by sum- DEVELOPMENTAL DISORDERS
ming directly the rates for each individual
condition within each study. The resulting rate Several studies have provided useful informa-
might, therefore, be slightly overestimated. tion on rates of syndromes that are similar to
The fraction of cases of autism with a known autism, but fall short of strict diagnostic criteria
medical condition that was potentially etiologi- for autistic disorder (Table 2.4). Because the
cally significant ranged from 0% to 16.7%, screening procedures and subsequent diagnos-
with a median and mean values of 5.5% and tic assessments differed from one study to an-
5.9% respectively. Even if some adjustment other, these groups of disorders are not strictly
were made to account for the underestimation comparable across studies. In addition, as they
of the rate of fragile X in epidemiological sur- were not the group on which the attention was
veys of autism, the attributable proportion of focused, details are often lacking on their phe-
cases of autism would not exceed the 10% nomenological features in the available reports.
TABLE 2.4 Relative Rates of Autism and Other Pervasive Developmental Disorders

Combined Rate
Prevalence Rate of Autism and Prevalence
No. Study Rates of Autism of Other PDDs Other PDDs Rate Ratio a Case Definition for Other PDDs

1 Lotter, 1966 4.1 3.3 7.8 0.90 Children with some behavior similar to that of
autistic children
2 Brask, 1970 4.3 1.9 6.2 0.44 Children with “other psychoses” or “ borderline
4 Wing, Yates, Brierly, & 4.9 16.3 21.2 3.33 Socially impaired (triad of impairments)
Gould, 1976
5 Hoshino et al., 1982 2.33 2.92 5.25 1.25 Autistic mental retardation

9 Burd, Fisher, & Kerbeshan, 3.26 >7.79 b >11.05 b 2.39 Children referred by professionals with “autistic-
1987 like” symptoms, not meeting DSM-III criteria for IA,
COPDD, or atypical PDD
14 Cialdella & Mamelle, 1989 4.5 4.7 9.2 1.04 Children meeting criteria for other forms of
“infantile psychosis” than autism, or a broadened
definition of DSM-III
17 Fombonne & du Mazaubrun, 4.6 6.6 11.2 1.43 Children with mixed developmental disorders
1992 c
20 Fombonne, du Mazaubrun, 5.3 10.94 16.3 2.05 Children with mixed developmental disorders
Cans, & Grandjean, 1997
26 Baird, Charman, & Baron- 30.8 27.1 57.9 0.9 Children with other PDDs
Cohen, 2000
27 Powell et al., 2000 7.8 13.0 20.8 1.7 Children with other PDDs
29 Bertrand et al., 2001 40.5 27.0 67.4 0.7 Children with PDDNOS and Asperger disorder
32 Chakrabarti & Fombonne, 16.8 36.1 52.9 2.15 Children with PDDNOS
35 Madsen et al., 2002 7.2 22.2 29.4 3.08

36 Chakrabarti & Fombonne, 22.0 24.8 46.8 1.13 Children with PDDNOS
Other PDD rate divided by autism rate.
Computed by the author.
These rates are derived from the complete results of the survey of three birth cohorts of French children (Rumeau-Rouquette et al., 1994).
Epidemiological Studies of Pervasive Developmental Disorders 53

Unspecified Pervasive Only a handful (N < 5) of cases were identified

Developmental Disorders in these surveys, with the resulting estimates
of 28 and 48/10,000 being extremely impre-
Different labels (see Table 2.4) have been used cise. By contrast, other recent autism surveys
to characterize these conditions, such as the have consistently identified smaller numbers
triad of impairments involving impairments in of children with AS than those with autism
reciprocal social interaction, communication, within the same survey. In Studies 23 to 27 and
and imagination (Wing & Gould, 1979). These 32 (reviewed in Fombonne & Tidmarsh, 2003)
groups would be overlapping with current di- and Study 36, the ratio of autism to AS rates in
agnostic labels such as atypical autism and each survey was above unity, suggesting that
pervasive developmental disorders not other- the rate of AS was consistently lower than that
wise specified (PDDNOS). Fourteen of the 36 for autism (Table 2.5). How much lower is dif-
surveys yielded separate estimates of the ficult to establish from existing data, but a
prevalence of these developmental disorders, ratio of 51 would appear to be an acceptable,
with 10 studies showing higher rates for the albeit conservative, conclusion based on this
nonautism disorders than the rates for autism. limited available evidence. Taking 13/10,000
The ratio of the rate of nonautistic PDD to the as the rate for autism, this translates into a
rate of autism varied between from 0.44 to rate for AS that would be 2.6/10,000, a figure
3.33 (Table 2.4) with a mean value of 1.6, used for subsequent calculations. A recent
which translates into an average prevalence es- survey of high-functioning PDDs in Welsh
timate of 20.8/10,000 if one takes 13/10,000 as mainstream primary schools has yielded a rel-
the rate for autism. In other words, for two atively high (uncorrected) prevalence estimate
children with autism assessed in epidemiologi- of 14.5/10,000. Of the 17 children contributing
cal surveys, three children were found with se- to this figure, 10 had either Asperger’s disor-
vere impairments that had a similar nature but der or high-functioning autism as a primary di-
that fell short of strict diagnostic criteria for agnosis. Assuming than half of these would
autism. This group has been much less studied have Asperger’s disorder, we could extrapolate
in previous epidemiological studies, but pro- a 4.3/10,000 prevalence, a figure that is in line
gressive recognition of its importance and rel- with other studies. However, much caution
evance to autism has led to changes in the should be applied to this calculation as it is
design of more recent epidemiological surveys based on several assumptions that are impossi-
(see later in this chapter). They now include ble to verify.
these less typical children in the case defini-
tion adopted in surveys. It should be clear from Childhood Disintegrative Disorder
these figures that they represent a substantial
group of children whose treatment needs are Few surveys have provided data on childhood
likely to be as important as those of children disintegrative disorder (CDD), also known
with autism. as Heller syndrome, disintegrative psychosis
(ICD-9), or late-onset autism (see Volkmar,
Asperger Syndrome and Childhood 1992). In addition to the four studies (9, 23, 31,
Disintegrative Disorder 32) of our previous review (Fombonne, 2002b),
another survey has provided new data on CDD
The reader is referred to recent epidemiological (36). Taking the five studies into account
reviews for these two conditions (Fombonne, (Table 2.6), prevalence estimates ranged from
2002b; Fombonne & Tidmarsh, 2003). Epi- 1.1 to 9.2/100,000. The pooled estimate based
demiological studies of Asperger syndrome on seven identified cases and a surveyed popu-
(AS) are sparse, probably because it was ac- lation of 358,633 children, was 1.9/100,000.
knowledged as a separate diagnostic category The upper-bound limit of the associated confi-
only recently in both ICD-10 and DSM-IV. Only dence interval (4.15/100,000) indicates that
two epidemiological surveys have specifically CDD is a rare condition, with 1 case occurring
investigated its prevalence (Ehlers & Gillberg, for every 65 cases of autistic disorder. As cases
1993; Kadesjö, Gillberg, & Hagberg, 1999). of CDD were both rare and already included in
TABLE 2.5 Asperger Syndrome (AS) in Recent Autism Surveys

Assessment Autism Asperger Syndrome

Autism /
Size of Age Rate/ Rate/ AS
Study Population Group Informants Instruments Diagnostic Criteria N 10,000 N 10,000 Ratio

Sponheim & Skjeldal, 1998 65,688 3–14 Parent Parental interview + direct ICD-10 32 4.9 2 0.3 16.0
Child observation, CARS, ABC
Taylor et al., 1999 490,000 0–16 Record Rating of all data available in child ICD-10 427 8.7 71 1.4 6.0
Kadesjö, Gillberg, & Hagberg, 826 6.7–7.7 Child ADI-R, Griffiths Scale or WISC, DSM-III-R /ICD-10 6 72.6 4 48.4 1.5
1999 Parent Asperger Syndrome Screening Gillberg’s criteria
Professional Questionnaire (Asperger syndrome)
Powell et al., 2000 25,377 1– 4.9 Records ADI-R DSM-III-R 54 — 16 — 3.4
Available data DSM-IV
Baird, Charman, & Baron- 16,235 7 Parents ADI-R ICD-10 45 27.7 5 3.1 9.0
Cohen, 2000 Child Psychometry DSM-IV
Other data
Chakrabarti & Fombonne, 15,500 2.5–6.5 Child ADI-R, 2 weeks multidisciplinary ICD-10 26 16.8 13 8.4 2.0
2001 Parent assessment, Merrill-Palmer, WPPSI DSM-IV
Chakrabarti & Fombonne, 10,903 2.5–6.5 Child ADI-R, 2 weeks multidisciplinary ICD-10 24 22.0 12 11.0 2.0
2004 Parent assessment, Merrill-Palmer, WPPSI DSM-IV
Overall 614 123 5.0
TABLE 2.6 Surveys of Childhood Disintegrative Disorder (CDD)

Size of Prevalence
Target Age Estimate 95% Cl1
Study Country (Region /State) Population Group Assessment N M/ F (/100,000) (/100,000)

Burd, Fisher, & Kerbeshan, United States 180,986 2–18 Structured parental interview and 2 2/— 1.11 0.13 ; 3.4
1987 (North Dakota) review of all data available–DSM-III
Sponheim & Skjeldal, 1998 Norway 65,688 3–14 Parental interview and direct 1 ? 1.52 0.04 ; 8.5
(Akershus County) observation (CARS, ABC)
Magnusson & Saemundsen, Iceland 85,556 5–14 ADI-R, CARS, and psychological 2 2/— 2.34 0.3 ; 8.4
2001 (whole island) tests—mostly ICD-10
Chakrabarti & Fombonne, United Kingdom 15,500 2.5– 6.5 ADI-R, 2 weeks’ multidisciplinary 1 1/— 6.4 0.16 ; 35.9
2001 (Staffordshire, Midlands) assessment, Merrill-Palmer, WPPSI—
Chakrabarti & Fombonne, 2004 United Kingdom 10,903 2.5– 6.5 ADI-R, 2 weeks’ multidisciplinary 1 1/— 9.2 0–58.6
(Staffordshire, Midlands) assessment, Merrill-Palmer, WPPSI—
Pooled estimates 358,633 7 6/— 1.9 0.87–4.15

56 Diagnosis and Classification

the numerator alongside autism cases in most fifth birthday, thereby optimizing sensitivity of
surveys, we do not provide separate estimates case-finding procedures. Furthermore, the size
of the numbers of subjects suffering from CDD of targeted populations was reasonably small
in subsequent calculations. ( between 9,000 and 16,000), probably allowing
for the most efficient use of research resources.
Prevalence for Combined PDDs Conducted in different regions and countries by
different teams, the convergence of estimates is
Taking the aforementioned conservative esti- striking. Two further results are worth noting.
mates, the prevalence for all PDDs is at least First, in sharp contrast with the prevalence for
36.4/10,000 (the sum of estimates for autism combined PDDs, the separate estimates for
[13/10,000], PDDNOS [20.8/10,000], and AS autistic disorder and PDD-NOS vary widely in
[2.6/10,000]). This global estimate is derived studies where separate figures were available.
from a conservative analysis of existing data. It appears that the reliability of the differentia-
However, six out of eight recent epidemio- tion between autistic disorder and PDD-NOS
logical surveys yielded even higher rates (Table was mediocre at that young age, despite the use
2.7). The two surveys that did not show higher of up-to-date standardized measures. Second,
rates might have underestimated them. In the the rate of mental retardation was, overall,
Danish investigation (Study 35), case finding much lower than in previous surveys of autism.
depended on notification to a National Reg- Although this should not be a surprise for chil-
istry, a method that is usually associated with dren in the PDD-NOS/AS groups, this trend
lower sensitivity for case finding. The Atlanta was also noticeable within samples diagnosed
survey by the Centers for Disease Control and with autistic disorder. To what extent this trend
Prevention (CDC; Study 38) was based on a reflects the previously mentioned differential
very large population (which typically yields classification issues between autism and PDD-
lower prevalence, as described earlier) and NOS or a genuine trend over time toward de-
age-specific rates were, in fact, in the 40-to- creased rate of mental retardation within
45/10,000 range in some birth cohorts (Fom- children with autistic disorder (possibly as a re-
bonne, 2003b). The common design features of sult as earlier diagnosis and intervention) re-
the four other epidemiological inquiries (Stud- mains to be established.
ies 26, 29, 32, 36) that yielded higher rates are In conclusion, the convergence of recent
worthy of mention. First, the case definition surveys around an estimate of 60/10,000 for
chosen for these investigations was that of a all PDD combined is striking, especially when
pervasive developmental disorder as opposed to coming from studies with improved methods.
the narrower approach focusing on autistic dis- This estimate appears now to be the best esti-
order typical of previous surveys. Investigators mate for the prevalence of PDDs currently
were concerned with any combination of severe available.
developmental abnormalities occurring in one
or more of the three symptomatic domains TIME TRENDS
defining PDD and autism. Second, case-finding
techniques employed in these surveys were The debate on the hypothesis of a secular in-
proactive, relying on multiple and repeated crease in rates of autism has been obscured by a
screening phases, involving both different in- lack of clarity in the measures of disease occur-
formants at each phase and surveying the same rence used by investigators, or rather in their
cohorts at different ages, which certainly max- interpretation. In particular, it is crucial to
imized the sensitivity of case identification. differentiate prevalence (the proportion of
Third, assessments were performed with stan- individuals in a population who suffer from a
dardized diagnostic measures (Autism Diag- defined disorder) from incidence (the number
nostic Interview-Revised [ADI-R] and Autism of new cases occurring in a population over
Diagnostic Observation Schedule [ADOS]), time). Prevalence is useful for estimating needs
which match well the dimensional approach re- and planning services; only incidence rates can
tained for case definition. Finally, these sam- be used for causal research. Both prevalence
ples comprised young children around their and incidence estimates will be inflated when
TABLE 2.7 Newer Epidemiological Surveys of PDDs
Rate/ Gender Ratio IQ Normal Rate/ Gender Ratio IQ Normal Rate/
No. Study Age 10,000 (MF) (%) 10,000 (MF) (%) 10,000
26 Baird, Charman, & Baron-Cohen, 2000 7 30.8 15.7 60 27.1 4.5 — 57.9
29 Bertrand et al., 2001 3–10 40.5 2.2 37 27.0 3.7 51 67.5
32 Chakrabarti & Fombonne, 2001 4–7 16.8 3.3 29 44.5 4.3 94 61.3
35 Madsen et al., 2002 8 7.7 — — 22.2 — — 30.0
36 Chakrabarti & Fombonne, 2004 4–7 22.0 4.0 33.3 35.8 8.7 91.6 58.7
37 Scott, Baron-Cohen, Bolton, & Brayne, 2002 5–11 — — — — — — 58.3 a
38 Yeargin-Allsopp et al., 2003 3–10 — — — — — — 34.0
39 Gurney et al., 2003 6–11 — — — — — — 52.0
Computed by the author.

58 Diagnosis and Classification

case definition is broadened and case ascertain- 2003). The population of 0- to 19-year-olds of
ment is improved. Time trends in rates can California was 10,462,273 in July 2002. If one
therefore only be gauged in investigations that applies a somewhat conservative PDD rate of
hold these parameters under strict control over 30/10,000, one would expect to have 31,386
time. These methodological requirements must PDD subjects within this age group living in
be borne in mind while reviewing the evidence California. These calculations do not support
for a secular increase in rates of PDDs. the “epidemic” interpretation, but instead sug-
Five approaches to assess this question have gest that children identified in the DDS data-
been used in the literature: (1) referral statis- base were only a subset of the population
tics, (2) comparison of cross-sectional epidemi- prevalence pool. The increasing numbers re-
ological surveys, (3) repeat surveys in defined flect merely an increasing proportion of chil-
geographic areas, (4) successive birth cohorts, dren accessing services.
and (5) incidence studies. Third, no attempt was ever made to adjust
the trends for changes in diagnostic concepts
Referral Statistics and definitions. However, major nosographic
modifications were introduced during the cor-
Increasing numbers of children referred to spe- responding years, with a general tendency in
cialist services or known to special education most classifications to broaden the concept of
registers have been taken as evidence for an autism (as embodied in the terms autism spec-
increased incidence of autism-spectrum dis- trum or pervasive developmental disorder).
orders. However, trends over time in referred Fourth, age characteristics of the subjects
samples are confounded by many factors such recorded in official statistics were portrayed
as referral patterns, availability of services, in a confusing manner where the preponder-
heightened public awareness, decreasing age at ance of young subjects was presented as evi-
diagnosis, and changes over time in diagnostic dence of increasing rates in successive birth
concepts and practices, to name only a few. cohorts (see Fombonne, 2001). The problems
Failure to control for these confounding factors associated with disentangling age from period
was obvious in some recent reports (Fombonne, and cohort effects in such observational data
2001), such as the widely quoted reports from are well known in the epidemiological litera-
California educational services (Department of ture and deserve better statistical handling.
Developmental Services, 1999, 2003). Fifth, the decreasing age at diagnosis leads
First, these reports applied to numbers, not to increasing numbers of young children being
to rates, and failure to relate these numbers to identified in official statistics or referred to al-
meaningful denominators left the interpreta- ready busy specialist services. Earlier identifi-
tion of an upward trend vulnerable to changes cation of children from the prevalence pool
in the composition of the underlying popula- may result in increased service activity; how-
tion. For example, the population of California ever, it does not mean increased incidence.
was 19,971,000 in 1970 and rose to 35,116,000 Another study of this dataset was subse-
as of July 1, 2002, a change of 75.8%. Thus, quently launched to demonstrate the validity
part of the increase in numbers of subjects of the epidemic hypothesis (MIND Institute,
identified with autism merely reflects the 2002). The investigation was, however, flawed
change in population size, but the DDS reports in its design. The authors relied on DDS data
have ignored or not adequately accounted for and aimed at ruling out changes in diagnostic
this change. practices and immigration into California as
Second, the focus on the year-to-year changes factors explaining the increased numbers.
in absolute numbers of subjects known to Cali- While immigration was reasonably ruled out,
fornia state-funded services detracts from more the study comparing diagnoses of autism and
meaningful comparisons. As of December 2002, mental retardation over time was impossible to
the total number of subjects with a PDD diagno- interpret in light of the extremely low (<20%)
sis was 17,748 in the 0-to-19 age group (includ- response rates. Furthermore, a study only
ing 16,108 autism codes 1 and 2 and 1,640 other based on cases registered for services cannot
PDDs; Department of Developmental Services, rule out that the proportion of cases within the
Epidemiological Studies of Pervasive Developmental Disorders 59

general population who registered with ser- increased efficiency over time in case identifi-
vices has changed over time. Assuming a con- cation methods used in surveys as well as
stant incidence and prevalence at two different changes in diagnostic concepts and practices.
time points (meaning there is no epidemic), the Thus, changes in diagnostic practices were re-
number of cases known to a public agency de- ported in Magnusson and Saemundsen’s study
livering services could well increase by 200% (2001) where ICD-9 rates for the oldest co-
if the proportion of cases from the community horts born in the years 1964 to 1983 were
referred to services rises from 25% to 75% in lower than the ICD-10 rates of the most recent
the interval. To rule out this ( likely, as men- 1984 to 1992 birth cohorts. Similarly, lower
tioned) explanation, data over time are needed rates in the oldest birth cohorts were thought
both on referred subjects and on nonreferred to reflect changes in diagnostic practices and
(or referred to other services) subjects. Failure boundaries in Webb, Lobo, Hervas, Scourfield,
to do that precludes drawing any inference to and Fraser’s study (1997). One large survey
the California population from a study of the recently conducted in the United Kingdom
DDS database (Fombonne, 2003b). The con- (Study 24) also documented a steep rise in the
clusions of this report were therefore simply number of cases diagnosed with autism or
unwarranted. atypical autism, and a similar trend for AS.
A recent reanalysis of the California data The interpretation of these trends is, however,
has in fact strongly suggested that switches in unclear because there was no control of drift
diagnostic practices from mental retardation over time in diagnostic practices nor of changes
to autism could also account for increased in service development.
numbers of subjects with an autism diagnosis in The most convincing evidence that method
the California DDS datasets (Croen, Grether, factors could account for most of the variabil-
Hoogstrate, & Selvin, 2002). Jick and Kaye ity in published prevalence estimates comes
(2003) obtained similar data in the United from a direct comparison of eight recent sur-
Kingdom. They showed that the incidence of veys conducted in the United Kingdom and the
specific developmental disorders (including United States (Table 2.8). In each country,
language disorders) decreased by about the four surveys were conducted around the same
same amount that the incidence of diagnoses year and with similar age groups. As there is
of autism increased in boys born from 1990 no reason to expect huge between-area differ-
to 1997. ences in rates, prevalence estimates should
On the whole, evidence from these referral therefore be comparable within each country.
statistics is very weak and certainly does not However, an inspection of estimates obtained
deserve the media attention that it has re- in each set of studies (Table 2.8: right-hand
ceived. Accordingly, proper epidemiological column) shows a 6-fold variation in rates for
studies are needed to assess secular changes in U.K. surveys, and a 14-fold variation in U.S.
the incidence of a disorder. rates. In each set of studies, high rates derive
from surveys that used intensive population-
Comparison of Cross-Sectional based screening techniques, whereas lower
Epidemiological Surveys rates were obtained from studies relying on
administrative methods for case finding. Since
Due to their cross-sectional methodology, no passage of time was involved, the magni-
most epidemiological investigations of autism tude of these gradients in rates can only be at-
have been concerned with prevalence. As tributed to differences in case identification
shown earlier, each epidemiological survey of methods across surveys, and the replication of
autism possesses unique design features that the pattern in two countries provides even
could account almost entirely for between- more confidence in this interpretation. This
studies variations in rates; time trends in rates analysis of recent and contemporaneous stud-
of autism are therefore difficult to gauge from ies shows that no inference on trends in the in-
published prevalence rates. The significant cidence of PDDs can be derived from a simple
correlation mentioned between prevalence rate comparison of prevalence rates over time,
and year of publication could merely reflect since studies conducted at different periods
60 Diagnosis and Classification

TABLE 2.8 Study Design Impact on Prevalence

Age PDD Rate/

Location Size Group Method 10,000

U.K. Studies
Chakrabarti & Staffordshire 15,500 21⁄ 2 –61⁄ 2 Intense screening and 62.6
Fombonne, 2001 assessment
Baird, Charman, & South East Thames 16,235 7 Early screening and 57.9
Baron-Cohen, 2000 follow-up identification
Fombonne, Simmons, England and Wales 10,438 5–15 National household 26.1
Ford, Meltzer, & survey of psychiatric
Goodman, 2001 disorders
Taylor et al., 1999 North Thames 490,000 0–16 Administrative records 10.1

U.S. Studies
Bertrand et al., 2001 Brick Township, New 8,896 3–10 Multiple sources of 67
Jersey ascertainment
Sturmey & James, Texas 3,564,577 6–18 Educational services 16
DDS, 1999 California 3,215,000 4–9 Educational services 15

Hillman, Kanafani, Missouri — 5–9 Educational services 4.8

Takahashi, & Miles,

are likely to differ even more with respect to age groups were included in each survey. The
their methodology. rate in the first survey for the youngest age
The next two approaches are in essence group (which resembles more closely the chil-
equivalent to a comparison of cross-sectional dren included in the other two surveys) was
surveys although specific attempts are made 5.1/10,000. Second, the increased prevalence
to maintain constant some design features in in the second survey was explained by im-
the surveys. proved detection among the mentally retarded,
and that of the third survey by cases born to
Repeat Surveys in Defined immigrant parents. That the majority of the
Geographic Areas latter group was born abroad suggests that mi-
gration into the area could be a key explana-
Repeated surveys, using the same methodology tion. Taken in conjunction with a change in
and conducted in the same geographic area local services and a progressive broadening of
at different points in time, can potentially the definition of autism over time acknowl-
yield useful information on time trends if the edged by the authors (Gillberg, Steffenburg, &
methods are kept relatively constant. The Schaumann, 1991), these findings do not pro-
Göteborg studies (Gillberg, 1984; Gillberg, vide evidence for an increased incidence of
Steffenburg, & Schaumann, 1991; Steffenburg autism.
& Gillberg, 1986) provided three prevalence Two separate surveys conducted on children
estimates that increased over a short period born 1992 to 1995 and 1996 to 1998 in
from 4.0 (1980) to 6.6 (1984) and 9.5/10,000 Staffordshire in the United Kingdom (Table
(1988). The gradient is even steeper if rates for 2.2: Studies 32 and 36) were performed with
the urban area alone are considered (4.0, 7.5, rigorously identical methods for case definition
and 11.6/10,000; Gillberg, Steffenburg, & and case identification. The prevalence for
Schaumann, 1991). However, comparisons of combined PDD was comparable and not statisti-
these rates are not straightforward as different cally different in the two surveys (Chakrabarti
Epidemiological Studies of Pervasive Developmental Disorders 61

& Fombonne, in press), suggesting no upward category including ADHD. The large sample
trend in overall rates of PDD during the time size allowed the authors to assess age, period,
interval of the studies. and cohort effects. Prevalence increased regu-
larly in successive birth cohorts; for example,
Successive Birth Cohorts among 7-year-olds, the prevalence rose from
18/10,000 in those born in 1989, to 29/10,000
In large surveys encompassing a wide age in those born in 1991 and to 55/10,000 in those
range, increasing prevalence rates among the born in 1993, suggestive of birth cohort ef-
most recent birth cohorts could be interpreted fects. Within the same birth cohorts, age ef-
as indicating a secular increase in the incidence fects were also apparent since for children born
of the disorder, provided that alternative expla- in 1989 the prevalence rose with age from
nations can confidently be ruled out. This 13/10,000 at age 6, to 21/10,000 at age 9, and
analysis was used in two French surveys (17 33/10,000 at age 11. As argued by the authors,
and 20), which derived from large sample sizes. this pattern is not consistent with what one
In the first study (17), prevalence estimates would expect from a chronic nonfatal condition
were available for the two birth cohorts of chil- diagnosed in the first years of life. Their analy-
dren born in 1972 and 1976 and surveyed in sis also showed a marked period effect that
1985 and 1986. The rates were similar (5.1 and identified the early 1990s as the period when
4.9/10,000) and not statistically different rates started to go up in all ages and birth co-
(Fombonne & du Mazaubrun, 1992). Further- horts. Gurney et al. (2003) further argued that
more, in a subsequent investigation conducted this phenomenon coincided closely with the in-
in 1989 and 1990 in exactly the same areas, the clusion of PDDs in the federal Individual with
age-specific rate of autism for the 1981 birth Disabilities Educational Act (IDEA) funding
cohort was slightly lower (3.1/10,000; Rumeau- and reporting mechanism in the United States.
Rouquette et al., 1994). In any event, the find- A similar interpretation of upward trends
ings were not suggestive of increasing rates in had been put forward by Croen, Grether,
the most recent cohorts. Another survey con- Hoogstrate, and Selvin (2002) in their analysis
ducted with the same methodology but in dif- of the California DDS data.
ferent French regions a few years later (Study
20) led to a similar overall prevalence estimate Incidence Studies
as the first survey (Table 2.2). The latter survey
included consecutive birth cohorts from 1976 Only three studies provided recent incidence
to 1985, and pooling the data of both surveys estimates (Kaye, Melero-Montes, & Jick, 2001;
showed no upward trend in age-specific rates Powell et al., 2000; Smeeth et al., 2004). All
(Fombonne, du Mazaubrun, Cans, & Grand- studies showed an upward trend in incidence
jean, 1997). Some weight should be given to over short periods. In the largest study of 1,410
these results as they derive from a total target subjects, there was a 10-fold increase in the
population of 735,000 children, 389 of whom rate of first recorded diagnoses of PDDs in
had autism. However, the most retarded children United Kingdom general practice medical
with autism were reflected in these studies and, records from 1988 to 1992 and from 2000 to
as a consequence, any upward trend that would 2001 (Smeeth et al., 2004). The increase was
apply specifically to high-functioning subjects more marked for PDDs other than autism
might have gone undetected. but the increase in autism also was obvious.
An analysis of special educational disability However, none of these investigations could de-
from Minnesota showed a 16-fold increase in termine the effect on the upward trend of
the number of children identified with a PDD changes over time in diagnostic criteria, im-
from 2001 to 2002, as compared with 1991 to proved awareness, and service availability.
1992 (Gurney et al., 2003; Study 39). The in-
crease was not specific to autism since during Conclusion on Time Trends
the same period an increase of 50% was ob-
served for all disability categories (except se- The available epidemiological evidence does
vere mental handicap), especially for the not strongly support the hypothesis that the
62 Diagnosis and Classification

incidence of autism has increased. As it stands constant. Finally, good psychometric data on
now, the recent upward trend in rates of preva- cognitive functioning will also be needed to as-
lence cannot be directly attributed to an in- sess trends in various subgroups in light of
crease in the incidence of the disorder. There is the preliminary evidence that patterns of men-
some evidence that diagnostic substitution and tal retardation in autism may be changing. Ob-
changes in the policies for special education as viously, surveillance programs should also
well as the increasing availability of services incorporate measures of risk factors hypothe-
are responsible for the higher prevalence fig- sized to exert causal influences for this group
ures. Most of the existing epidemiological data of disorders.
are inadequate to properly test hypotheses on
changes in the incidence of autism in human IMMIGRANT STATUS, ETHNICITY,
populations. Moreover, the low frequency of SOCIAL CLASS, AND OTHER
autism and PDDs seriously limits power in CORRELATES
most investigations and variations of small
magnitude in the incidence of the disorder are Some investigators have mentioned the possi-
very likely to go undetected. Future investiga- bility that rates of autism might be higher
tions should aim at setting up surveillance pro- among immigrants (Gillberg, 1987; Gillberg,
grams that will allow estimates of the incidence Schaumann, & Gillberg, 1995; Gillberg, Stef-
of PDDs (as opposed to autism only) and the fenburg, & Schaumann, 1991; Wing, 1980).
monitoring of its changes over time. It will be Five of the 17 children with autism identified
crucial to set up parallel investigations in dif- in the Camberwell study were of Caribbean ori-
ferent geographic areas to replicate findings gin (Study 4; Wing, 1980), and the estimated
across areas as a validating tool. Such programs rate of autism was 6.3/10,000 for this group
should focus on age groups where the identifi- compared with 4.4/10,000 for the rest of the
cation and diagnosis of the range of PDDs is population (Wing, 1993). However, the wide
less likely to fluctuate over time. Rapid changes confidence intervals associated with rates from
in the age at first diagnosis and concerns about this study (Table 2.2) indicate no statistically
the validity and stability of diagnostic assess- significant difference. This area of London had
ments among preschool samples require inves- received a large proportion of immigrants from
tigators to focus on older age groups. On the the Caribbean region in the 1960s; and when
other hand, changes in the autistic symptoma- there is migration flux in and out of an area, es-
tology in adolescence and difficulties in service timation of population rates should be viewed
delivery to teenagers (and therefore in case with much caution. Yet, Afro-Caribbean chil-
identification) suggest a focus on younger dren referred from the same area were recently
children. The school-age years (7 to 12 years) found to have higher rates of autism than re-
should therefore be selected for efficient ferred controls (Goodman & Richards, 1995).
monitoring. Mandatory education at that age The sample again was very small (N = 18) and
would facilitate identification and would min- differential referral patterns to a tertiary center
imize potential difficulties in diagnosing high- also providing services for the local area could
functioning subjects at the upper end of this not be ruled out. Only one child was born from
age range. Diagnostic assessments should rely British-born Afro-Caribbean parents in a re-
on standardized measures of known reliability cent U.K. survey (Study 21; Webb, Lobo, et al.,
and validity. Furthermore, developmental and 1997), providing little support to this particular
phenomenological data should be collected at a hypothesis. Similarly, the findings from the
symptomatic level, and uniformly across the Göteborg studies paralleled an increased mi-
whole spectrum of PDDs, remaining free of gration flux in the early 1980s in this area
particular nosological contingencies. Secondary (Gillberg, 1987); they, too, were based on rela-
application of diagnostic algorithms (current tively small numbers (19 children from immi-
and/or future) on datasets containing detailed grant parents). In the same geographic area,
developmental and symptomatic data will then Arvidsson et al. (1997; Study 22) had five chil-
allow for performing meaningful comparisons dren out of nine in their sample with either both
over time, while holding diagnostic groupings parents (N = 2) or one parent (N = 3) having im-
Epidemiological Studies of Pervasive Developmental Disorders 63

migrated to Sweden. However, there were no cific methodological context of these investi-
systematic comparisons with rates of immi- gations. Most studies had low numbers of
grants in the population. A positive family his- identified cases, and especially small numbers
tory for developmental disorders was reported of autistic children born from immigrant
in three such cases and a chromosomal abnor- parents, and many authors in these studies re-
mality in one further case. In the Icelandic sur- lied on broadened definitions of autism. Sta-
vey (Study 31), 2.5% of the autism parents were tistical testing was not rigorously conducted
from non-European origin compared to a 0.5% and doubts could be raised in several studies
corresponding rate in the whole population, but about the appropriateness of the comparison
it was unclear if this represented a significant data that were used. Thus, the overall propor-
difference. In Study 23, the proportion of chil- tion of immigrants in the population may be an
dren with autism and a non-European origin inappropriate figure with which to compare
was marginally but not significantly raised observed rates of children from immigrant
compared with the population rate of immi- parents among autistic series. Fertility rates of
grants (8% vs. 2.3%), but this was based on a immigrant families are likely to be different
very small sample (two children of non-Euro- from those in the host populations and call
pean origin). A U.K. survey found comparable for strict age-adjusted comparisons of individ-
rates in areas with contrasting ethnic compo- uals at risk for the disorder. The proportion of
sition (Powell et al., 2000). In the Utah sur- immigrants in the entire population might se-
vey, where a clear breakdown by race was riously underestimate that for younger age
achieved (Ritvo et al., 1989; Study 15, Table groups, and, in turn, this could have given rise
2.2), the autism parents showed no deviation to false positive results. In addition, studies
from the racial distribution of this state. The sampling children through services or clinical
proportion of non-Whites in this study and sources may be biased because ethnicity, race
state was, however, noticeably low, providing and social class are likely to differentially af-
little power to detect departures from the null fect access to these settings. Finally, studies
hypothesis. Other studies have not systemati- were generally poor in their definition of im-
cally reported the proportion of immigrant or migrant status, with unclear amalgamation of
ethnic groups in the areas surveyed. In four information on country of origin, citizenship,
studies where the proportions of immigrant immigrant status, race, and ethnicity. Finally,
groups were low (11, 12, 19, 21), rates of it is unclear what common mechanism could
autism were in the upper range of rates. Con- explain the putative association between im-
versely, in studies of other populations (14, migrant status and autism, since the origins of
17, and 20) where immigrants contributed the immigrant parents (especially in Study 16;
substantially to the denominators, rates were see also Gillberg & Gillberg, 1996) were di-
in the rather low band. The analysis of a large verse and represented in fact all continents.
sample (N = 4,356) of Californian PDD chil- With this heterogeneity in mind, what com-
dren showed a lower risk of autism in children mon biological features might these immigrant
of Mexico-born mothers and a similar risk for families share and what would be a plausible
children of mothers born outside the United mechanism explaining the putative association
States compared with California-born moth- between autism and immigrant status? The
ers (Croen, Grether, & Selvin, 2002). In this possibility of an increased vulnerability to in-
study, the risk of PDD was raised in African trauterine infections in nonimmunized im-
American mothers with an adjusted rate ratio migrant mothers was raised, but not supported,
of 1.6 (95% CI: 1.5 to 1.8); by contrast, the in a detailed analysis of 15 autistic children
prevalence was similar in White, Black, and from immigrant parents (Gillberg & Gillberg,
other races in the population-based survey of 1996). These authors instead posited that
Atlanta (Yeargin-Allsopp et al., 2003), where parents, and in particular fathers, affected
case ascertainment is likely to be more com- with autistic traits would be inclined to travel
plete than in the previous study. abroad to find female partners more naïve
Taken altogether, the combined results of to their social difficulties. This speculation
these reports should be interpreted in the spe- was based, however, on three observations
64 Diagnosis and Classification

only, and assessment of the autistic traits in increased incidence or prevalence rate ratio in
two parents was not independently obtained. a cluster does not prove anything; this erro-
The hypothesis of an association between neous approach has been referred to in the lit-
immigrant status or race and autism, therefore, erature as the “ Texan sharpshooter ” effect,
remains largely unsupported by the empirical referring to the gunman who shot first and
results. Most of the claims about these possi- then painted a target around the bullet hole. On
ble correlates of autism were derived from the other hand, a negative finding would cer-
post hoc observations of very small samples tainly suggest a random phenomenon.
and were not subjected to rigorous statistical When cluster alarms are associated to a
testing. Large studies have generally failed to possible causal mechanism, it is recommended
detect such associations. to perform focused tests of clustering at other
suspected sources of risk exposure. The clus-
Autism and Social Class ter alarms for childhood leukemia occurring
near a nuclear plant in England were followed
Twelve studies provided information on the so- by investigations of disease incidence at other
cial class of the families of autistic children. nuclear plants, which proved to be negative
Of these, four studies (1, 2, 3, and 5) suggested (Hoffmann & Schlattmann, 1999). However,
an association between autism and social class the potential source of the cluster alarm is
or parental education. The year of data collec- not always identified and, in these instances,
tion for these four investigations was before it is suggested to monitor the incidence of fu-
1980 (Table 2.2), and all studies conducted ture cases in the area of first alarm. Chen,
thereafter provided no evidence for the associ- Connelly, and Mantel (1993) have outlined
ation. Thus, the epidemiological results sug- postalarm monitoring techniques that allow
gest that the earlier findings were probably investigators to confirm or reject alarms,
due to artifacts in the availability of services based on the observation of the time intervals
and in the case-finding methods, as already preceding each of the first five cases diag-
shown in other samples (Schopler, Andrews, & nosed subsequent to the alarm. The approach is
Strupp, 1979; Wing, 1980). a confirmatory technique that ignores the clus-
ter alarm data and thus avoids the aforemen-
Cluster Reports tioned preselection bias. Other techniques,
such as space-time scan statistics (Kulldorff,
Occasional reports of space or time clustering 1999), can confirm or reject a cluster alarm
of cases of autism have raised concerns in the by extending the investigation to a larger area
general public. In fact, only one such report while avoiding selection biases, adjusting for
has been published in the professional litera- population density, confounding variables and
ture (Baron-Cohen, Saunders, & Chakrabarti, multiple testing, and allowing for the precise
1999) that described seven children with location of clusters. They require, however, the
either autism or PDD-NOS living within a few availability of regional or national geocoded
streets from each other in a small town of the data that are usually not available for autism.
Midlands (United Kingdom). The cluster was Other general statistical techniques to assess
first identified by a parent, and the subsequent time and space clustering are reviewed in spe-
analysis was uninformed with proper statisti- cialist journals.
cal procedures and inconclusive as to whether Cluster alarms are likely to represent ran-
this cluster could have occurred by chance dom occurrence in most instances, as illus-
only. The comparison of the incidence or trated by several recent investigations of cluster
prevalence rate within the cluster to that of the alarms for other rare disorders of childhood.
general population (as performed by Baron- Cluster alarms in autism have not been investi-
Cohen & Wheelwright, 1999) is an inappropri- gated with scientific rigor, whereas research
ate technique to assess cluster alarms. The strategies and ad hoc statistical procedures
reason is that, by definition, a preselection exist for that purpose. The approach to such
bias occurs in the delineation of the cluster cluster alarms should be to confirm the alarm
boundaries (Kulldorff, 1999). Thus, finding an in the first place, using the available techniques
Epidemiological Studies of Pervasive Developmental Disorders 65

to assess the significance of clusters and to ex- most likely represents changes in the con-
clude random noise in spatial and time distribu- cepts, definitions, service availability, and
tion of the disorder. Only when an alarm has awareness of autistic-spectrum disorders in
been confirmed should researchers set up more both the lay and professional public. To assess
complex epidemiological investigations to in- whether the incidence has increased, method
vestigate risk factors and causal mechanisms. factors that account for an important propor-
tion of the variability in rates must be tightly
CONCLUSION controlled.
Taking 35/10,000 and 60/10,000 as two
Epidemiological surveys of autism and PDDs working rates for the combination of all PDDs,
have now been carried out in several countries. and using U.S. population figures as of July 1,
Methodological differences in case definition 2002, it can be estimated that about 284,000
and case-finding procedures make between- and up to 486,000 subjects under the age of 20
survey comparisons difficult to perform. suffer from a PDD in the United States. These
Despite these differences, some common char- figures carry straightforward implications for
acteristics of autism and PDDs in population current and future needs in services and early
surveys have consistently emerged. Autism is educational intervention programs.
associated with mental retardation in about
70% of the cases and is overrepresented among Cross-References
males (with a male/female ratio of 4.31).
Autism is found in association with some rare Issues of diagnosis of autism spectrum disor-
and genetically determined medical conditions, ders are addressed in Chapter 1 and Chapters
such as tuberous sclerosis. Overall, the median 3 through 6; medical aspects of autism are dis-
value of about 5.5% for the combined rate of cussed in Chapter 20.
medical disorders in autism derived from this
review is consistent with the 5% (Tuchman,
Rapin, & Shinnar, 1991) to 10% (Rutter, Bai-
ley, Bolton, & Le Couteur, 1994) figures avail- Ahlsen, G., Gillberg, C., Lindblom, R., & Gillberg,
able from other investigations. A majority of C. (1994). Tuberous sclerosis in Western Swe-
surveys has ruled out social class as a risk den: A population study of cases with early
factor for autism, a result once supported by childhood onset. Archives of Neurology, 51,
studies of clinical, that is, less representative, 76–81.
samples. The putative association of autism Arvidsson, T., Danielsson, B., Forsberg, P., Gillberg,
with immigrant status or race is, so far, not C., Johansson, M., & Kjellgren, G. (1997).
borne out by epidemiological studies. The con- Autism in 3–6-year-old children in a suburb of
Goteborg, Sweden. Autism, 2, 163–173.
clusion of a lack of variation in the incidence
Baird, G., Charman, T., Baron-Cohen, S., Cox, A.,
of autism according to race or ethnicity is Swettenham, J., Wheelwright, S., et al. (2000).
reached, however, from a weak empirical base, A screening instrument for autism at 18 months
and future studies might address this issue of age: A 6 year follow-up study. Journal of the
more efficiently. In fact, epidemiological stud- American Academy of Child and Adolescent
ies of autism and PDDs have generally been Psychiatry, 39, 694–702.
lacking sophistication in their investigation of Baron-Cohen, S., Saunders, K., & Chakrabarti, S.
most other risk factors. (1999). Does autism cluster geographically? A
The same considerations apply to the issue research note. Autism, 3, 39–43.
of secular changes in the incidence of autism. Bertrand, J., Mars, A., Boyle, C., Bove, F., Yeargin-
Allsopp, M., & Decouf le, P. (2001). Preva-
The little evidence that exists does not support
lence of autism in a United States population:
this hypothesis, but power to detect time
The Brick Township, New Jersey, investiga-
trends is seriously limited in existing tion. Pediatrics, 108, 1155–1161.
datasets. The debate has been largely con- Bohman, M., Bohman, I. L., Björck, P. O., &
founded by confusion between prevalence and Sjöholm, E. (1983). Childhood psychosis in a
incidence. Although prevalence estimates ap- northern Swedish county: Some preliminary
pear to have gone up over time, this increase findings from an epidemiological survey. In
66 Diagnosis and Classification

M. H. Schmidt & H. Remschmidt (Eds.), Epi- 1987 through 1998. Retrieved July 16, 2004,
demiological approaches in child psychiatry from http://www.dds.ca.gov.
(pp. 164–173). New York: George Thieme Ver- Department of Developmental Services. (2003).
lag Stuttgart. Autism spectrum disorders: Changes in the Cali-
Brask, B. H. (1972). A prevalence investigation of fornia caseload—an update 1999 through 2002.
childhood psychoses. In Nordic Symposium on Retrieved July 16, 2004, from http://www.dds
the Care of Psychotic Children. Oslo: Barne- .ca.gov/Autism /pdf/AutismReport2003.pdf.
psychiatrist Forening. Deykin, E. Y., & MacMahon, B. (1979). The inci-
Bregman, J. D., & Volkmar, F. R. (1988). Autistic dence of seizures among children with autistic
social dysfunction and Down’s syndrome. symptoms. American Journal of Psychiatry,
Journal of the American Academy of Child and 136(10), 1310–1312.
Adolescent Psychiatry, 27, 440–441. Ehlers, S., & Gillberg, C. (1993). The epidemiol-
Bryson, S. E., Clark, B. S., & Smith, I. M. (1988). ogy of Asperger syndrome: A total population
First report of a Canadian epidemiological study. Journal of Child Psychology and Psychi-
study of autistic syndromes. Journal of Child atry, 34, 1327–1350.
Psychology and Psychiatry, 4, 433–445. Fombonne, E. (1999). The epidemiology of autism:
Burd, L., Fisher, W., & Kerbeshan, J. (1987). A A review. Psychological Medicine, 29, 769–786.
prevalence study of pervasive developmental Fombonne, E. (2001). Is there an epidemic of
disorders in North Dakota. Journal of the autism? Pediatrics, 107, 411–413.
American Academy of Child and Adolescent Fombonne, E. (2002a). Case identification in an
Psychiatry, 26(5), 700–703. epidemiological context. In M. Rutter & E.
Chakrabarti, S., & Fombonne, E. (2001). Pervasive Taylor (Eds.), Child and adolescent psychiatry
developmental disorders in preschool children. (pp. 52–69). Oxford, England: Blackwell.
Journal of the American Medical Association, Fombonne, E. (2002b). Prevalence of childhood
285(24), 3093–3099. disintegrative disorder (CDD). Autism, 6(2),
Chakrabarti, S., & Fombonne, E. (in press). Perva- 147–155.
sive developmental disorders in preschool chil- Fombonne, E. (2003a). Epidemiological surveys of
dren: High prevalence confirmed. American autism and other pervasive developmental dis-
Journal of Psychiatry. orders: An update. Journal of Autism and De-
Chen, R., Connelly, R., & Mantel, N. (1993). velopmental Disorders, 33(4), 365–382.
Analysing post-alarm data in a monitoring Fombonne, E. (2003b). The prevalence of autism.
system in order to accept or reject the alarm. Journal of the American Medical Association,
Statistics in Medicine, 12, 1807–1812. 289(1), 1–3.
Chess, S. (1971). Autism in children with congeni- Fombonne, E., & du Mazaubrun, C. (1992). Preva-
tal rubella. Journal of Autism and Childhood lence of infantile autism in 4 French regions.
Schizophrenia, 1, 33–47. Social Psychiatry and Psychiatric Epidemiol-
Cialdella, P., & Mamelle, N. (1989). An epidemio- ogy, 27, 203–210.
logical study of infantile autism in a French Fombonne, E., du Mazaubrun, C., Cans, C., &
department. Journal of Child Psychology and Grandjean, H. (1997). Autism and associated
Psychiatry, 30(1), 165–175. medical disorders in a large French epidemio-
Croen, L. A., Grether, J. K., Hoogstrate, J., & logical sample. Journal of the American Acad-
Selvin, S. (2002). The changing prevalence of emy of Child and Adolescent Psychiatry, 36(11),
autism in California. Journal of Autism and De- 1561–1566.
velopmental Disorders, 32(3), 207–215. Fombonne, E., Simmons, H., Ford, T., Meltzer, H.,
Croen, L. A., Grether, J. K., & Selvin, S. (2002). & Goodman, R. (2001). Prevalence of perva-
Descriptive epidemiology of autism in a Cali- sive developmental disorders in the British na-
fornia population: Who is at risk? Journal of tionwide survey of child mental health.
Autism and Developmental Disorders, 32(3), Journal of the American Academy of Child and
217–224. Adolescent Psychiatry, 40(7), 820–827.
Davidovitch, M., Holtzman, G., & Tirosh, E. (2001, Fombonne, E., & Tidmarsh, L. (2003). Epidemio-
March). Autism in the Haifa area—an epi- logical data on Asperger disorder. Child and
demiological perspective. Israeli Medical As- Adolescent Psychiatric Clinics of North Amer-
sociation Journal, 3, 188–189. ica, 12, 15–21.
Department of Developmental Services. (1999). Ghaziuddin, M., Tsai, L., & Ghaziuddin, N. (1992).
Changes in the population of persons with Autism in Downs’ syndrome: Presentation and
autism and pervasive developmental disorders diagnosis. Journal of Intellectual Disability Re-
in California’s Developmental Services System: search, 35, 449–456.
Epidemiological Studies of Pervasive Developmental Disorders 67

Gillberg, C. (1984). Infantile autism and other Howlin, P., Wing, L., & Gould, J. (1995). The
childhood psychoses in a Swedish region: Epi- recognition of autism in children with Down
demiological aspects. Journal of Child Psy- syndrome: Implications for intervention and
chology and Psychiatry, 25, 35–43. some speculations about pathology. Develop-
Gillberg, C. (1987). Infantile autism in children of mental Medicine and Child Neurology, 37,
immigrant parents. A population-based study 406–414.
from Göteborg, Sweden. British Journal of Psy- Hunt, A., & Lindenbaum, R. H. (1984). Tuberous
chiatry, 150, 856–858. sclerosis: A new estimate of prevalence within
Gillberg, C., & Coleman, M. (1996). Autism and the Oxford region. Journal of Medical Genet-
medical disorders: A review of the literature. ics, 21, 272–277.
Developmental Medicine and Child Neurology, Jick, H., & Kaye, J. A. (2003). Epidemiology and
38, 191–202. possible causes of autism. Pharmacotherapy,
Gillberg, C., & Gillberg, C. (1996). Autism in im- 23(12), 1524–1530.
migrants: A population-based study from Kadesjö, B., Gillberg, C., & Hagberg, B. (1999).
Swedish rural and urban areas. Journal of In- Autism and Asperger syndrome in seven-year-
tellectual Disability Research, 40, 24–31. old children: A total population study. Journal
Gillberg, C., Schaumann, H., & Gillberg, I. C. of Autism and Developmental Disorders, 29(4),
(1995). Autism in immigrants: Children 327–331.
born in Sweden to mothers born in Uganda. Kaye, J., Melero-Montes, M., & Jick, H. (2001).
Journal of Intellectual Disability Research, 39, Mumps, measles, and rubella vaccine and the
141–144. incidence of autism recorded by general practi-
Gillberg, C., Steffenburg, S., & Schaumann, H. tioners: A time trend analysis. British Medical
(1991). Is autism more common now than ten Journal, 322, 0–2.
years ago? British Journal of Psychiatry, 158, Kielinen, M., Linna, S. L., & Moilanen, I. (2000).
403–409. Autism in Northern Finland. European Child
Goodman, R., & Richards, H. (1995). Child and and Adolescent Psychiatry, 9, 162–167.
adolescent psychiatric presentations of second- Knobloch, H., & Pasamanick, B. (1975). Some eti-
generation Afro-Caribbeans in Britain. British ologic and prognostic factors in early infan-
Journal of Psychiatry, 167, 362–369. tile autism and psychosis. Pediatrics, 55,
Gurney, J. G., Fritz, M. S., Ness, K. K., Sievers, P., 182–191.
Newschaffer, C. J., & Shapiro, E. G. (2003). Kulldorff, M. (1999). Statistical evaluation of
Analysis of prevalence trends of autism spec- disease cluster alarms. In A. B. Lawson et al.
trum disorder in Minnesota [Comment]. (Eds.), Disease mapping and risk assessment
Archives of Pediatrics and Adolescent Medi- for public health (pp. 143–149). Chichester,
cine, 157(7), 622–627. West Sussex, England: Wiley.
Hillman, R., Kanafani, N., Takahashi, T., & Miles, Le Couteur, A., Rutter, M., Lord, C., Rios, P.,
J. (2000). Prevalence of autism in Missouri: Robertson, S., Holdgrafer, M., et al. (1989).
Changing trends and the effect of a compre- Autism diagnostic interview: A standardized
hensive state autism project. Missouri Medi- investigator-based instrument. Journal of Autism
cine, 97, 159–163. and Developmental Disorders, 19, 363–387.
Hoffmann, W., & Schlattmann, P. (1999). An analy- Lord, C., Risi, S., Lembrecht, L., Cook, E., Leven-
sis of the geographical distribution of leukemia thal, B., DiLavore, P., et al. (2000). The Autism
incidence in the vicinity of a suspected point Diagnostic Observation Schedule-Generic: A
source: A case study. In A. B. Lawson et al. standard measure of social and communication
(Eds.), Disease mapping and risk assessment deficits associated with the spectrum of
for public health (pp. 396–409). Chichester, autism. Journal of Autism and Developmental
West Sussex, England: Wiley. Disorders, 30, 205–223.
Honda, H., Shimizu, Y., Misumi, K., Niimi, M., & Lord, C., Schopler, E., & Revecki, D. (1982). Sex
Ohashi, Y. (1996). Cumulative incidence and differences in autism. Journal of Autism and
prevalence of childhood autism in children Developmental Disorders, 12, 317–330.
in Japan. British Journal of Psychiatry, 169, Lotter, V. (1966). Epidemiology of autistic condi-
228–235. tions in young children: I. Prevalence. Social
Hoshino, Y., Yashima, Y., Ishige, K., Tachibana, R., Psychiatry, 1, 124–137.
Watanabe, M., Kancki, M., et al. (1982). The Lotter, V. (1967). Epidemiology of autistic con-
epidemiological study of autism in Fukushi- ditions in young children: II. Some charac-
maken. Folia Psychiatrica et Neurologica Japon- teristics of the parents and children. Social
ica, 36, 115–124. Psychiatry, 1(4), 163–173.
68 Diagnosis and Classification

Lowe, T. L., Tanaka, K., Seashore, M. R., Young, Scott, F. J., Baron-Cohen, S., Bolton, P., & Brayne,
J. G., & Cohen, D. J. (1980). Detection of C. (2002). Brief report: Prevalence of autism
phenylketonuria in autistic and psychiatric spectrum conditions in children aged 5–11
children. Journal of the American Medical As- years in Cambridgeshire, UK. Autism, 6(3),
sociation, 243, 126–128. 231–237.
Madsen, K. M., Hviid, A., Vestergaard, M., Schen- Shepherd, C. W., Beard, C. M., Gomez, M. R., Kur-
del, D., Wohlfahrt, J., Thorsen, P., et al. (2002). land, L. T., & Whisnant, J. P. (1991). Tuberous
A population-based study of measles, mumps, sclerosis complex in Olmsted County, Min-
and rubella vaccination and autism. New En- nesota, 1950–1989. Archives of Neurology, 48,
gland Journal of Medicine, 347(19), 1477–1482. 400–401.
Magnusson, P., & Saemundsen, E. (2001). Preva- Smalley, S. (1998). Autism and tuberous sclerosis.
lence of autism in Iceland. Journal of Autism Journal of Autism and Developmental Disor-
and Developmental Disorders, 31, 153–163. ders, 28, 407–414.
Matsuishi, T., Shiotsuki, M., Yoshimura, K., Shoji, Smalley, S. L., Tanguay, P. E., Smith, M., & Gutier-
H., Imuta, F., & Yamashita, F. (1987). High rez, G. (1992). Autism and tuberous sclerosis.
prevalence of infantile autism in Kurume City, Journal of Autism and Developmental Disorders,
Japan. Journal of Child Neurology, 2, 268–271. 22, 339–355.
McCarthy, P., Fitzgerald, M., & Smith, M. A. Smeeth, L., Cook, C., Fombonne, E., Heavey, L.,
(1984). Prevalence of childhood autism in Ire- Rodrigues, L., Smith, P., et al. (2004). Rate of
land. Irish Medical Journal, 77(5), 129–130. first recorded diagnosis of autism and other
MIND Institute. (2002, October 17). Report to the pervasive developmental disorders in United
Legislature on the Principal Findings from the Kingdom general practice, 1988 to 2001. Man-
epidemiology of autism in California: A com- uscript submitted for publication.
prehensive pilot study. Davis: University of Spitzer, R. L., & Siegel, B. (1990). The DSM-III-R
California. field trial of pervasive developmental disor-
Mouridsen, S. E., Bachmann-Andersen, L., ders. Journal of the American Academy of
Sörensen, S. A., Rich, B., & Isager, T. (1992). Child and Adolescent Psychiatry, 6, 855–886.
Neurofibromatosis in infantile autism and Sponheim, E., & Skjeldal, O. (1998). Autism and
other types of childhood psychoses. Acta Pae- related disorders: Epidemiological findings in
dopsychiatrica, 55, 15–18. a Norwegian study using ICD-10 diagnostic
Powell, J., Edwards, A., Edwards, M., Pandit, B. S., criteria. Journal of Autism and Developmental
Sungum-Paliwal, S. R., & Whitehouse, W. Disorders, 28, 217–222.
(2000). Changes in the incidence of childhood Steffenburg, S., & Gillberg, C. (1986). Autism and
autism and other autistic spectrum disorders autistic-like conditions in Swedish rural and
in preschool children from two areas of the urban areas: A population study. British Jour-
West Midlands, UK. Developmental Medicine nal of Psychiatry, 149, 81–87.
and Child Neurology, 42, 624–628. Steinhausen H.-C., Göbel, D., Breinlinger, M., &
Ritvo, E. R., Freeman, B. J., Pingree, C., Mason- Wohlloben, B. (1986). A community survey of
Brothers, A., Jorde, L., Jensen, W. R., et al. infantile autism. Journal of the American Acad-
(1989). The UCLA-University of Utah epidemi- emy of Child Psychiatry, 25(2), 186–189.
ologic survey of autism: Prevalence. American Sturmey, P., & James, V. (2001). Administrative
Journal of Psychiatry, 146(2), 194–199. prevalence of autism in the Texas school sys-
Rumeau-Rouquette, C., du Mazaubrun, C., Verrier, tem. Journal of the American Academy of Child
A., Mlika, A., Bréart, G., Goujard, J., et al. Psychiatry, 40(6), 621.
(1994). Prévalence des handicaps: Évolution Sugiyama, T., & Abe, T. (1989). The prevalence of
dans trois générations d’enfants 1972, 1976, autism in Nagoya, Japan: A total population
1981. Paris: Editions INSERM. study. Journal of Autism and Developmental
Rutter, M. (1970). Autistic children: Infancy to Disorders, 19(1), 87–96.
adulthood. Seminars in Psychiatry, 2, 435–450. Tanoue, Y., Oda, S., Asano, F., & Kawashima, K.
Rutter, M., Bailey, A., Bolton, P., & Le Couteur, A. (1988). Epidemiology of infantile autism in
(1994). Autism and known medical conditions: Southern Ibaraki, Japan: Differences in preva-
Myth and substance. Journal of Child Psychol- lence in birth cohorts. Journal of Autism and
ogy and Psychiatry, 35, 311–322. Developmental Disorders, 18, 155–166.
Schopler, E., Andrews, C. E., & Strupp, K. (1979). Taylor, B., Miller, E., Farrington, C. P., Petropou-
Do autistic children come from upper-middle- los, M. C., Favot-Mayaud, I., Li, J., et al.
class parents? Journal of Autism and Develop- (1999, June 12). Autism and measles, mumps,
mental Disorders, 9(2), 139–151. and rubella vaccine: No epidemiological evi-
Epidemiological Studies of Pervasive Developmental Disorders 69

dence for a causal association. Lancet, 353, authority. Developmental Medicine and Child
2026–2029. Neurology, 45(6), 377–384.
Treffert, D. A. (1970). Epidemiology of infantile Wignyosumarto, S., Mukhlas, M., & Shirataki, S.
autism. Archives of General Psychiatry, 22, (1992). Epidemiological and clinical study
431–438. of autistic children in Yogyakarta, Indonesia.
Tuchman, R. F., Rapin, I., & Shinnar, S. (1991). Kobe Journal of Medical Sciences, 38(1), 1–19.
Autistic and dysphasic children: II. Epilepsy. Wing, L. (1980). Childhood autism and social
Pediatrics, 88, 1219–1225. class: A question of selection? British Journal
Volkmar, F. R. (1992). Childhood disintegrative of Psychiatry, 137, 410–417.
disorder: Issues for DSM-IV. Journal of Autism Wing, L. (1993). The definition and prevalence of
and Developmental Disorders, 22, 625–642. autism: A review. European Child and Adoles-
Volkmar, F. R., Klin, A., Siegel, B., Szatmark, I., cent Psychiatry, 2, 61–74.
Lord, C., Campbell, M., et al. (1994). Field Wing, L., & Gould, J. (1979). Severe impairments
trial for autistic disorder in DSM-IV. American of social interactions and associated abnor-
Journal of Psychiatry, 151(9), 1361–1367. malities in children: Epidemiology and classi-
Volkmar, F. R., & Nelson, D. S. (1990). Seizure dis- fication. Journal of Autism and Developmental
orders in autism. Journal of the American Disorders, 9(1), 11–29.
Academy of Child and Adolescent Psychiatry, 1, Wing, L., Yeates, S. R., Brierly, L. M., & Gould, J.
127–129. (1976). The prevalence of early childhood
Webb, E., Lobo, S., Hervas, A., Scourfield, J., & autism: Comparison of administrative and epi-
Fraser, W. I. (1997). The changing prevalence demiological studies. Psychological Medicine,
of autistic disorder in a Welsh health district. 6, 89–100.
Developmental Medicine and Child Neurology, Yeargin-Allsopp, M., Rice, C., Karapurkar, T.,
39, 150–152. Doernberg, N., Boyle, C., & Murphy, C.
Webb, E., Morey, J., Thompsen, W., Butler, C., Bar- (2003). Prevalence of autism in a U.S. metro-
ber, M., & Fraser, W. I. (2003). Prevalence of politan area. Journal of the American Medical
autistic spectrum disorder in children attend- Association, 289(1), 49–55.
ing mainstream schools in a Welsh education

Childhood Disintegrative Disorder


Nearly 100 years ago, Theodore Heller, a Vi- recognized. Debate initially centered on issues
ennese educator, reported on six children who of continuity with schizophrenia and, more re-
had exhibited severe developmental regression cently, with autism. The early confusion with
at ages 3 to 4 years following a period of ap- schizophrenia reflected the general presump-
parently normal development. After the re- tion that, more or less, all severe psychiatric
gression, recovery was quite limited. Initially, disturbances reflected psychosis that was
Heller (1908) termed this condition dementia equated with schizophrenia (see Chapter 1,
infantilis; subsequently, other terms have been this Handbook, this volume, for a discussion).
used for the concept, for example, Heller’s Only as various lines of evidence began to sug-
syndrome, disintegrative psychosis, and, more gest the importance of making distinctions
recently, childhood disintegrative disorder was it clear, for example, that autism differed
(CDD). Over the past century, more than 100 from schizophrenia of childhood in a host of
cases have been reported in the world litera- ways. The pioneering studies of Kolvin (1971)
ture; not surprisingly, information on the and Rutter (1972) were particularly important
condition is much more limited than that in this regard since they demonstrated that
available on autism. In this chapter, we review within a large group of “psychotic” children,
the development of the diagnostic concept, there was a bimodal pattern of onset. The
current definitions, information on clinical early-onset group had begun to have troubles
features, epidemiology, course, and validity at birth or within a year or so after birth while
of the condition. We also note areas of contro- the late-onset group developed apparently nor-
versy, and issues that remain to be clarified mally for many years. Clinical features of the
are reviewed. early-onset group included marked impair-
ments in social, cognitive, and language devel-
DEVELOPMENT OF THE DIAGNOSTIC opment similar to those described by Kanner
CONCEPT for autism, whereas the late-onset group ex-
hibited delusions, hallucinations, and other
Although only recently accorded official features more similar to schizophrenia. There
recognition in DSM-IV as a diagnostic con- was no higher than expected frequency of
cept, CDD has a long history. schizophrenia among family members of the
Heller’s work occurred shortly after the early-onset group, but there was such an in-
turn of this century, but many decades were to crease among first-degree relatives in the
pass before his diagnostic concept was widely late-onset group. This observation has been

The authors acknowledge the support of NICHD grant 5P01-HD-03008-28. The case report is reprinted
with permission from Psychoses and Pervasive Developmental Disorders in Childhood and Adolescence, Fred
R. Volkmar, MD, editor, American Psychiatric Association Press, 1996. The authors are grateful to Drs.
Kurita, Malhotra, Gillberg, and Deonna for their kind provision of additional information.

Childhood Disintegrative Disorder 71

well replicated (e.g., Makita, 1966; Volkmar, meant that historical information, for example,
Cohen, Hoshino, Rende, & Paul, 1988). It is on the pattern and time of onset, was not par-
also of interest that in these studies a handful ticularly relevant to the diagnosis of autism.
of cases did not seem to fall so simply either Thus, children who would have been recog-
into the early- or late-onset group, for exam- nized as having disintegrative psychosis in
ple, 3 of the 83 cases in Kolvin’s series exhib- ICD-9 would usually have been said to have
ited an intermediate age of onset between the autistic disorder in DSM-III-R. This state of
autistic and schizophrenic groups. affairs became even more complex as the
These and other data on the validity of drafts of the ICD-10 revision began to appear.
autism led to its official recognition in DSM-III
(American Psychiatric Association, 1980), Childhood Disintegrative Disorder in ICD-
where it was placed in a new class of disorder, 10 and DSM-IV
pervasive developmental disorder (PDD; see
Chapter 1, this Handbook, this volume). Infan- In contrast to DSM-III-R, ICD-10 included a
tile autism was defined on the basis of marked draft definition of CDD. This definition was
social, language, and other problems arising by largely consistent with earlier work (e.g.,
30 months of age. Partly in recognition of the Heller, 1930; Zappert, 1921), which had
fact that a few children seemed to develop an generally suggested the following diagnostic
autistic-like condition after that time, DSM- features:
III also included a category, childhood onset
pervasive developmental disorder (COPDD). 1. A distinctive pattern of syndrome onset (a
This category was not meant to be analogous to period of several years of normal develop-
Heller’s diagnostic concept; the implicit ment before a marked deterioration)
presumption in DSM-III was that such cases 2. Progressive deterioration (either gradual or
invariably reflected some progressive neu- abrupt) once the syndrome had its onset
ropathological process. However, the ninth re- with loss of skills in multiple areas
vision of the International Classification of 3. Behavioral and affective symptoms
Diseases (ICD-9; World Health Organization 4. An absence of features of gross neurologi-
[WHO], 1978) had included a category for dis- cal dysfunction
integrative psychosis or Heller’s syndrome de-
fined on the basis of “normal or near normal Draft ICD-10 criteria for the condition in-
development in the first years of life, followed cluded apparently normal development for at
by a loss of social skills and of speech together least 2 years with age-appropriate social, com-
with a severe disorder of emotion, behavior, municative, and other skills; a definite loss of
and relationships.” ICD-9 did not, however, skills in more than one area; development of
prove as influential as DSM-III, primarily be- problems in social interaction, communication,
cause the latter system included explicit guide- and restricted patterns of interest or behavior
lines for diagnosis. of the type observed in autism; and a loss of
Although a very detailed definition of interest in the environment. By definition, the
COPDD was included in DSM-III, it quickly disorder could not coexist with autism or any
became clear that this diagnostic concept was other explicitly defined PDD, schizophrenia,
problematic in many respects (Volkmar, 1987), elective mutism, or the syndrome of acquired
and only a handful of reports appeared in the aphasia with epilepsy.
literature (Burd, Fisher, & Kerbeshian, 1988). Using the draft ICD-10 criteria, Volkmar
When DSM-III was revised (DSM-III-R, Amer- and Cohen (1989) identified a series of 10
ican Psychiatric Association, 1987), the cases of apparent CDD from within a larger
COPDD concept was dropped. Diagnostic cri- sample of individuals with the clinical features
teria for autism were expanded in number and of autism. Characteristics of the CDD were
conceptually, and early onset of autism was then contrasted to cases of autism that had
not an essential diagnostic feature; autistic been identified before or after 24 months. This
disorder was the only operationally defined comparison was particularly appropriate to
disorder within the PDD class. These changes the issue of whether CDD simply represented
72 Diagnosis and Classification

late-onset autism since differences between somewhat less detailed and less truly opera-
the groups on some external measure(s) tionalized. In addition, ICD-10 includes loss of
would tend to support the validity of CDD interest in the environment as a diagnostic cri-
apart from late-onset autism. This was indeed terion and is more explicit in indicating that
the case. Cases with late-onset autism tended the actual behavioral criteria for autism must
to be higher functioning while the CDD cases be met. As a practical matter, it would appear
were more likely to be mute, more likely to be that the diagnosis should probably not be made
in residential placement, and so forth. Thus, if the actual behavioral criteria for autism are
it appeared that CDD was not simply late- not met. In both DSM-IV and ICD-10, the age
onset autism; rather, it appeared to have dis- and pattern of onset are particularly important
tinctive features, clinical course, and even for the definition of the condition; that is, there
worse outcome. must be a marked regression after a period of
The inclusion of CDD in ICD-10 repre- prolonged normal development—arbitrarily set
sented a marked divergence from DSM-III-R at 2 years of age. This regression is associated
and clearly had implications for the definition with the acquisition of behaviors commonly
of autism in DSM-IV (American Psychiatric seen in autism. It is hoped that increased
Association, 1994). As part of the DSM-IV re- awareness of the condition will stimulate
vision process, a review (Volkmar, 1992) of greater identification of cases and more re-
CDD identified 77 cases in the world litera- search and that increasingly better guidelines
tures and suggested that while the condition for diagnosis will be developed (Kurita, 1989).
was apparently relatively rare, it seemed to
merit inclusion in DSM-IV because it appeared CLINICAL FEATURES
to differ from autism in important respects
and because it was not (as had previously been The following section provides a discussion of
assumed) always, or even usually, associated the essential clinical features of CDD.
with an identifiable neurological condition
that might account for the deterioration. Onset of Childhood Disintegrative
Inclusion of CDD in DSM-IV was also sug- Disorder
gested by the results of the DSM-IV field trial
for autism and related conditions (Volkmar As noted previously, the onset of CDD is
et al., 1994; Volkmar & Rutter, 1995). Al- highly distinctive and an essential diagnostic
though the field trial was primarily concerned feature.
with the development and validation of the
Age of Onset
DSM-IV diagnosis of autism, 16 cases of CDD
that had been previously evaluated at partici- Development prior to the regression is rela-
pating centers were included. Of even more in- tively prolonged (several years) and should be
terest, an additional 15 cases that met ICD-10 reasonably normal; for example, the child has
criteria for CDD were identified in the field the capacity to speak in sentences by age 2
trial. In these cases, the clinician had not given (WHO, 1990). Heller’s (1930) impression was
CDD as the clinical diagnosis but had noted that onset was often between ages 3 and 5
the presence of various diagnostic features of years, and this range continues to be the case.
the condition. This is not surprising since, par- Volkmar (1992) reported a mean age at onset
ticularly in the United States, clinicians have of 3.4 years in his review of 77 reported cases.
been much less familiar with the diagnostic The issue of the time of onset is particularly
concept. Relative to cases with autism, in the relevant to distinctions of CDD from autism
field trial, cases with CDD were more likely to because it is clear that, in some cases, autism
be mute and had greater degrees of associated is recognized after 24 months of age (although
mental handicap. almost invariably before age 3; Volkmar, Stier,
As noted by Volkmar and Rutter (1995), the & Cohen, 1985). For example, in the case se-
DSM-IV and ICD-10 criteria for the condition ries collected at Division TEACCH in North
are conceptually very similar. However, for Carolina, over three-fourths of children with
the sake of brevity, the DSM-IV system is autism had been identified by their parents as
Childhood Disintegrative Disorder 73

having difficulties by 2 years of age (Short & rized in Figure 3.1. There is a clear and signif-
Schopler, 1988). Diagnosis of autism is also icant difference in the two distributions of
sometimes delayed by the primary clinicians’ onset of the two conditions.
lack of familiarity with the condition (Siegel,
Characteristics of Onset
Pliner, Eschler, & Elliott, 1989).
Cases with late-onset autism (i.e., whose Several different patterns of onset of CDD
difficulties are apparent after age 2 but before have been observed. Occasionally, the condi-
age 3) tend to be higher functioning, and it tion has a relatively abrupt onset (days to
seems likely that case detection may be de- weeks) but sometimes develops more gradually
layed by the relative preservation of cognitive (weeks to months). There may be a premoni-
abilities (Volkmar & Cohen, 1989). Wohlge- tory phase prior to the marked deterioration;
muth, Kiln, Cohen, and Volkmar (1994) com- during this time, the child may be nonspecifi-
pared aspects of deterioration in autism and cally agitated, anxious, or dysphoric.
CDD. In their report, when deterioration in In several case series, the onset of CDD
autism was reported, it typically involved the has been noted to be associated with some
loss of ability to speak in single words or the psychosocial stress or medical event (Evans-
failure of this ability to progress. In contrast, Jones & Rosenbloom, 1978; Kobayashi & Mu-
in CDD the previously acquired level of lan- rata, 1998; Kurita, 1988; Volkmar, 1992).
guage was much higher, and deterioration was However, the significance of stressful events
always observed in multiple areas, that is, not in syndrome pathogenesis is unclear. The
simply limited to speech. group of stressors reported has been diverse,
Occasional ambiguities are sometimes ob- but all share the feature of being relatively
served, for example, a child with recurrent ear common to preschool children, for example,
infections and delayed speech who then goes birth of a sibling or death of a grandparent,
on to develop a more typical CDD presentation hospitalization for elective surgery, or immu-
at age 3. Kurita (1988) has suggested that nizations. It seems unlikely that such associa-
early development, that is, prior to age 2, may tions have etiological significance (Rutter,
not always be perfectly normal, and there 1985). Davidovitch, Glick, Holtzman, Tirosh,
might be a history of mild delay (see also and Safir (2000) noted that this phenomenon,
Kurita, Kita, & Miyake, 1992). Despite cur- that is, of attributing some etiological signifi-
rent diagnostic criteria, some cases of CDD cance to an associated medical or psychoso-
might develop before age 2 although the diag- cial event, is common in parents of children
nosis of such cases is problematic and may be with autism who regress as compared to those
the source of some confusion, that is, both
with autism and with Rett disorder although
the regression in Rett’s is usually relatively 50
early in life. Similarly, in Landau-Kleffner
Percent of Group

syndrome (of acquired aphasia with epilepsy), 40 Autism

there may be occasional confusion with CDD
or autism, but it appears that the clinical 30
features of this condition and its course are
relatively distinctive (Bishop, 1985, 1994). 20
Further complicating the issue is the problem
of potential early regression in autism, which 10
is variably reported in 20% to 40% of cases
(see Lainhart et al., 2002; Rogers & DiLalla,
1990; Siperstein & Volkmar, 2004; Volkmar 0
<1 1 2 3 4 5 6 7 8 9
et al., 1985, for a discussion). We return to the
Year of Life
issue of differential diagnosis and current con-
troversies later in this chapter. Data related to Figure 3.1 Age of onset in 160 cases with clinical
age of onset of autism and CDD as abstracted diagnoses of CDD and 316 cases with clinical diag-
from the DSM-IV field trial data are summa- noses of autism.
74 Diagnosis and Classification

who do not. Kobayashi and Murata (1998) re- Given that the child typically has been
ported similar results in their study of chil- speaking in full sentences, often quite well,
dren with setback autism. the development of either total mutism or
marked deterioration in verbal language is also
BEHAVIORAL AND CLINICAL very striking and frequent in CDD. Even for
FEATURES those individuals who subsequently regain
speech, it does not typically return to previous
Table 3.1 provides a summary of clinical fea- levels of communicative ability. Rather, com-
tures in a number of reported cases as well as municative abilities are more similar to those
several cases seen by the authors and not previ- observed in autism with a sparsity of commu-
ously reported. nicative acts, limited expressive vocabulary,
Once CDD is established, it resembles and markedly impaired pragmatic skills.
autism in its phenomenological manifesta- Unusual behaviors including stereotyped
tions. Typically, social skills are markedly behaviors, problems with transitions and
impaired. There is, however, some suggestion change, and nonspecific overactivity are typi-
that the degree of impairment may be slightly cally observed (Malhotra & Singh, 1993). As
less than that observed in autism (Kanner, noted previously, various affective responses
1973; Kurita, 1988; Kurita et al., 1992). Par- that appear inexplicable are often observed at
ents usually report that the loss of social in- the time of syndrome onset. As mentioned pre-
teraction skill is dramatic and of great viously, ICD-10 suggests that a general loss of
concern to them. interest in the environment is also usual. Dete-

TABLE 3.1 Characteristics of Disintegrative Disorder Cases

1908–1975 1977–1995 1996–2004
Variable N = 48 N = 58 N = 67
Male/ Female
Sex ratio 35/12 49/9 53/14

Age at onset (years) 3.42 1.12 3.32 1.42 3.21 0.97

Age at follow-up 8.67 4.14 10.88 5.98 10.25 4.81

Symptoms % of N Cases
Speech deterioration / loss 100 47 100 58 100 54
Social disturbance 100 43 98 57 100 54
Stereotypy/resistance to change 100 38 85 54 68 54
Overactivity 100 42 77 37 59 54
Affective symptoms/anxiety 100 17 78 38 55 54
Deterioration self-help skills 94 33 82 49 66 54
Source: Adapted with permission from “Childhood Disintegrative Disorders: Is-
sues for DSM-IV,” by F. R. Volkmar, 1992, Journal of Autistic Developmental Dis-
orders 22, 625–642; and “Childhood Disintegrative Disorder,” by F. R. Volkmar,
A. Klin, W. D. Marans, & D. J. Cohen, in Autism and Pervasive Developmental Dis-
orders, second edition, 1997, New York: Wiley. Additional cases based on case se-
ries reported by Kurita et al., 1994; “Childhood Disintegrative Disorder:
Re-Examination of the Current Concept,” by S. Malhotra and N. Gupta, 2002, Eu-
ropean Child and Adolescent Psychiatry, 11(3), pp. 108–114; Mourdisen et al.,
2000, with additional cases supplied by C. Gillberg and F. R. Volkmar. Note: Re-
sults based on available data.
Childhood Disintegrative Disorder 75

rioration in self-help skills, notably in toileting of CDD cases, the child’s behavior and devel-
skills, is striking (Kurita, 1988; Volkmar, opment deteriorate to the much lower level of
1992) and in contrast to autism where such functioning and remain there. On the one
skills are often acquired somewhat late but are hand, no further deterioration occurs, but sub-
not typically dramatically lost. sequent developmental gains appear to be min-
imal (Volkmar & Cohen, 1989). On the other
EPIDEMIOLOGY hand, the marked developmental regression
seems to be followed by a limited recovery;
Epidemiological data on this condition are for example, a child regains the capacity to
limited. This reflects both (1) the true relative speak although usually only in a limited way
and frequency of the condition apart from (Volkmar & Cohen, 1989). Burd, Ivey, Barth,
autism and (2) the likelihood that cases have and Kerbeshian (1998) provided follow-up
been markedly underdiagnosed. In his review data on two children with CDD after 14 years;
(see Chapter 2, this volume, 2002), Fombonne at the time of follow-up, both were severely
notes that estimates range from about 1 to 9 impaired, exhibited seizure disorder, were
cases per 100,000 children. Data from case se- nonverbal, and were in residential treatment.
ries (e.g., the series of consecutive cases re- With greater awareness of the condition, more
ported by Volkmar & Cohen, 1989) indicate cases are being seen at younger ages and even-
that the disorder was one-tenth as common as tual longer term follow-up of these cases will
autism; however, these data were not based on be needed.
a truly epidemiological sample. It is interest- In a small number of cases, the develop-
ing that a rather similar rate was observed of mental deterioration is progressive and does
children with autism who had regressed after not plateau. This appears to be likely if some
age 3 (Rogers & DiLalla, 1990). A different identifiable, neuropathological process can be
study using a somewhat more epidemiologi- identified. If the process is progressive, death
cally based sample suggested a prevalence rate may be the eventual result (Corbett, 1987),
of 1 in 100,000 (Burd, Fisher, & Kerbeshian, and there may be increased mortality if other
1989). It must be emphasized that the relative medical conditions are present. For example,
lack of familiarity of clinicians with this con- two cases reported by Mouridsen, Rich, and
cept makes interpretation of the available data Isager (1998) died in association with suba-
somewhat suspect. cute sclerosing panencephalitis (death at age 9
Initially, it appeared that the condition years) and with tuberous sclerosis (death at
was equally affecting males and females. age 31 years). One case with whom the authors
However, more recent studies have noted a are familiar died in adolescence apparently
preponderance of males similar to that seen in following a seizure. In a handful of cases, the
autism (Lord, Schopler, & Revicki, 1982). It child has been observed to make a noteworthy
also is possible that some cases of Rett syn- recovery.
drome were originally misdiagnosed as having
Heller’s syndrome; that is, Rett’s condition NEUROBIOLOGICAL FINDINGS
was described only in 1966, and the degree of AND ETIOLOGY
the deterioration in Rett’s cases may be sug-
gestive of CDD (Burd et al., 1989; Hill & Although CDD was originally termed demen-
Rosenbloom, 1986; Millichap, 1987; Rett, tia infantilis, Heller’s impression, and that of
1966; Volkmar, 1992). In cases of CDD ob- others, was that CDD was not associated with
served in the past 20 years, there is a high apparent organic disease; this was also origi-
male predominance. nally Kanner’s impression (1943) about
autism. In both cases, this initial impression
COURSE AND PROGNOSIS has had to be modified. It is now clear that
about 25% of individuals with autism have
Information on course and outcome is an im- seizures, often with an onset later than is typ-
portant factor for evaluating the validity of ical in children, and another 25% have vari-
psychiatric conditions. In approximately 75% ous other EEG abnormalities. In the Volkmar
76 Diagnosis and Classification

(1992) review, EEGs had been obtained in In Volkmar’s (1992) review of published
45 cases of apparent CDD. These data are sup- cases, specific neuropathological conditions
plemented by additional, more recent reports of were only occasionally identified. Late onset,
CDD cases (Malhotra & Singh, 1993; Mourid- for example, after age 6, of CDD appears more
sen, Rich, & Isager, 2000; Volkmar & Rutter, likely to be associated with some specific
1995). Seizures have been noted in various case neuropathological process. Given the child’s
reports, for example, Hill and Rosenbloom marked regression, it is now typical for parents
(1986; 2 of 9 cases), Volkmar and Cohen (1989; to consult with many different specialists and
2 of 10 cases), Malhotra and Singh (1993; 1 for various tests, laboratory studies, and diag-
case), and Kurita, Osada, & Miyake (2004; 3 of nostic procedures to be obtained. Other than
10 cases). In the Malhotra and Singh report EEG abnormalities and occasional seizure dis-
(1993), the onset of developmental deteriora- order, such tests usually are not particularly
tion was associated with seizures. Tuchman and productive although they should be under-
Rapin (1997) reported that regression in chil- taken. However, usually even when very exten-
dren with autism spectrum disorder occurred sive medical investigations are undertaken, it
equally in individuals with and without is not possible to identify a specific general
seizures. Similarly, Shinnar and colleagues medical condition that accounts for the child’s
(2001) reported relatively high rates of seizures deterioration (Volkmar, 1992). Even if such an
and autism spectrum disorder in a large cohort etiology can be identified, the diagnosis of
of children with language regression. Although CDD is made and the presence of the associ-
these data are limited, the rates of seizure dis- ated medical condition noted. This is similar
order and EEG abnormality appear to be simi- to the approach in autism that may be associ-
lar to those observed in autism (e.g., Deykin & ated with various general medical conditions
MacMahon, 1979; Rutter, 1985; Volkmar & (Rutter, Bayley, Boulton, & Le Couter, 1994).
Nelson, 1990) and suggest that an EEG is rou- Except for the EEG information and one
tinely indicated as part of clinical assessment. study by Gillberg, Terenius, Hagberg, Witt-
CDD has been associated with various Engerstrom, and Eriksson (1990) on cere-
conditions such as tuberous sclerosis, neu- brospinal fluid (CSF) beta-endorphins, there is
rolipidoses, metachromatic leukodystrophy, a general absence of information regarding the
Addison-Schilder’s disease, and subacute scle- neurochemistry, neuropsychology, neurophysi-
rosing panencephalitis along with literally ology, or neuroanatomy of CDD. Although the
hundreds of other possible causes including small sample sizes and differences in method
metabolic, infectious, genetic, immunopathic, complicate the interpretation of the limited
environmental, and epileptogenic causes (see available data, in general, striking differences
Dyken & Krawiecki, 1983; Mouridsen et al., in brain morphology and structure have not
1998). As noted previously, the impression, for been observed (Mouridsen et al., 2000). Simi-
example, in DSM-III and DSM-III-R was that larly, information on potential genetic factors
such associated conditions are generally found, is extremely limited. Mouridsen et al. (2000)
but this view has not been supported by the reported one case of a boy with an inversion of
data. Selected disorders associated with loss of chromosome 10 (46xy, inv(10) (p11, 21q21.2),
developmental skills are listed in Table 3.2. but the child’s mother also had the same anom-

TABLE 3.2 Selected Disorders Associated with Loss of Developmental Skills

Infections (HIV, measles, CMV) Mitochondrial defecits (e.g., Leigh disease)
Hypothyroidism Subacute sclerosing panencephalitis
Neurolipidosis Metachromatic Leukodystrophy
Addison-Schilder disease Seizures
Angleman syndrome Gangliosidoses
Lipofuscinosis Aminoacidopathies (e.g., PKU)
Source: For an exhaustive list, see “Neurodegenerative Diseases of Infancy and
Childhood,” by P. Dyken and N. Krawiecki, 1983, “Neurodegenerative Diseases of In-
fancy and Childhood, Annals of Neurology, 13, 351–364.
Childhood Disintegrative Disorder 77

aly and was a successful professional. In other disorders, Down syndrome, Prader-Willi
reports, no unusual genetic findings have been syndrome, and velocardiofacial syndrome are
noted (e.g., Burd et al., 1998; Russo, Perry, familiar examples of chromosomal disorders
Kolodny, & Gillberg, 1996). This is particu- that are demonstrably genetic in etiology, but
larly unfortunate because there is some reason rarely pass through generations within a family.
to think that possibly the etiology (or etiolo- It is conceivable that CDD, as well, could be the
gies) in CDD may be somewhat more homoge- result of rare sporadic chromosomal rearrange-
neous than those more typically seen in ments. The fact that gross chromosomal abnor-
autism, that is, given the pattern of onset and malities have so far not been found in cases of
the relative infrequency of such cases. CDD does not rule out such a mechanism. The
resolution of routine clinical cytogenetic exams
GENETICS OF CHILDHOOD in instances where a specific gene or locus is
DISINTEGRATIVE DISORDER not known is on the order of 4 million base
pairs of DNA. Any abnormality that involved
Given the relevance of genes for other PDDs, it substantially less genetic material than this
is tempting to speculate about their role in this could be missed, even in the face of the exten-
syndrome as well. Despite the paucity of direct sive clinical work-ups that usually accompany
evidence from population genetic or molecular the diagnosis of developmental regression.
studies, even the limited epidemiological and In addition, chromosomal abnormalities are
case report data do provide some basis for de- certainly not the only form of sporadic muta-
veloping and testing hypotheses about disease tion. The most pertinent recent example can be
etiology. In addition, the recent identification found in Rett’s disorder. Mutations of single
of Methyl CpG Binding Protein 2 (MECP2) as molecules in the DNA chain encoding the
the cause of a majority of cases of classic MECP2 gene account for about 80% of cases
Rett’s disorder (Amir et al., 1999) may turn presenting with the classic phenotype (see
out to provide important clues about promising Amir & Zoghbi, 2000, for review). These spo-
approaches to future genetic studies of CDD. radic cases are most often the result of germ-
The literature reviewed in this chapter sug- line mutations in males (Girard et al., 2001). It
gests that the prevalence of CDD is approxi- is certainly possible that an analogous, rarely
mately 1 per 100,000 live births. In addition, occurring mutation in ova or sperm could ac-
case examples of subjects who have any family count for CDD.
members with a PDD of any kind are unusual The absence of reports of multigeneration
(Zwaigenbaum et al., 2000). Indeed, CDD ap- CDD families suggests that it is unlikely that
pears in the majority of reported instances to the syndrome is inherited in dominant fash-
be “sporadic”; that is, it seems to arise “out of ion. Such disorders are transferred from one
nowhere” within a particular family (Malhotra generation to the next, with only one copy of
& Gupta, 2002). These observations that CDD a defective gene necessary to cause the condi-
is rare and does not often cluster in families tion. While not everyone who inherits such a
must be taken into account in formulating hy- dominant gene develops the full-blown disor-
potheses about the role genes could conceiv- der, nonetheless, we would expect to see suc-
ably play in the syndrome. Furthermore, cessive generations in a family having some
Lainhart and colleagues (2002) reported that evidence of the syndrome. To date, no such
rates of the “ broader autism phenotype” did families have been reported. Alternative
not differ in parents of children with autism forms of Mendelian inheritance might better
with and without regression, suggesting no in- fit the available data but are not entirely satis-
crease in the genetic liability—at least in the factory explanations either. Recessive disor-
autism phenotype characterized by regression. ders, in which one defective gene must be
Though the terms familial and genetic are inherited from each of two parents, can often
often used interchangeably, they are not syn- appear to be sporadic, particularly if the of-
onymous. There are multiple mechanisms that fending versions of a gene are found rarely in
can result in the de novo appearance of genetic a population. In this circumstance, multiple
syndromes. In the realm of neuropsychiatric generations in a family would be unaffected
78 Diagnosis and Classification

until the chance mating of two mutation tively homogeneous clinical entity confined
carriers. If this were the case, however, the to girls appears to have been mistaken. In
risk of having the disorder in a sibling of a reality, the phenotype can vary dramatically
child with CDD would be on the order of 25%. in females from the classic syndrome
The fact that cases of CDD are not commonly described in DSM-IV to mild learning disabil-
reported among first-degree relatives would ities. Moreover, a wide variety of phenotypes
argue against this hypothesis but would not have now also been identified in males (Amir
rule it out. A paucity of familial cases could & Zoghbi, 2000; Couvert et al., 2001; Ham-
also result if parents who have one child with mer, Dorrani, Dragich, Kudo, & Schanen,
CDD tend to stop having additional children. 2002; Kleefstra et al., 2002; Meloni et al.,
A similar explanation would have to be in- 2000). The current understanding of CDD
voked if it was hypothesized that CDD was a could similarly turn out to be incomplete
rare X-linked disorder. The observed male when and if a genetic contribution is identi-
preponderance of 41 (Malhotra & Gupta, fied, either because the manifestations of a
2002; Volkmar, 1992; Volkmar & Rutter, 1995) CDD mutation could be quite variable or be-
could be accounted for if girls were provided a cause there could be gender-related difference
measure of protection by a second X chromo- in how such an abnormality is expressed.
some. Again, however, the recurrence risk for Recent studies of Rett’s disorder have also
male offspring in an affected family would be called into question the notion of homogeneity
on the order of 50%, far higher than what is from a genetic as well as clinical perspective.
observed in the available literature. At present, only about 80% of patients with the
Data that are directly relevant to this issue classic Rett’s disorder phenotype have been
are limited. In their study of the broader autis- confirmed to have a MECP2 mutation. Familial
tic phenotype, Lainhart and colleagues (2002) or atypical cases have proven even more diffi-
reported that parents of probands with regres- cult to characterize. These findings could be
sion and nonregressive autism exhibited simi- the result of an incomplete understanding of the
lar rates of the broader autism phenotype. genomic structure and regulation of MECP2.
Zwaigenbaum and colleagues (2000) reported However, it may be that there are other Rett’s
on two half brothers, one with autism and the disorder loci apart from MECP2. The same
other with CDD. could well apply to CDD. More than one genetic
Finally, in the absence of any other com- abnormality might result in similar clinical pre-
pelling explanation, it could be hypothesized sentations, so any single mechanism might
that CDD is a complex disorder resulting from prove an unsatisfactory solution to the problem
either the chance accumulation of a number of of understanding the transmission of CDD.
rare genetic events, from some unknown envi- Finally, the identification of MECP2 as a
ronmental precipitant that alone or in concert major cause of developmental regression in
with a genetic liability results in the appar- girls came about in large part because of the
ently sporadic emergence of the syndrome, or identification of unusual families among the
from a novel genetic mechanism. much more common sporadic cases of Rett’s
disorder (Amir et al., 1999). The search for
Lessons from Rett Syndrome “outlier ” examples of an already unusual dis-
order could prove to be a similarly profitable
Dramatic advances in the understanding of strategy for studying the genetics of CDD.
the genetic basis of Rett’s disorder may turn
out to have considerable relevance for the Future Genetic Studies of Childhood
study of CDD. While preliminary screenings Disintegrative Disorder
of the MECP2 gene in CDD cases have not
identified causative mutations (M. State and J. The combination of being a rare disorder and
Greally, personal communication, 2004), more the absence of clear familial aggregation of
general lessons may be derived from the study CDD cases certainly presents obstacles to the
of Rett’s disorder. For instance, the strongly study of the genetics of the disorder. Nonethe-
held notion that Rett’s disorder was a rela- less, rapid advances in genetic technologies
Childhood Disintegrative Disorder 79

hold promise for the study of CDD. The advent VALIDITY OF CHILDHOOD
of high-resolution chromosomal screening DISINTEGRATIVE DISORDER AS
techniques, the development of extremely sen- A DIAGNOSTIC CATEGORY
sitive proteomic technologies, and the expo-
nential growth in the ability to screen the Diagnostic categories exist for different rea-
genome for mutations will allow investigators sons and purposes (see Rutter, 2003; Rutter
to test hypotheses about the transmission of & Taylor, 2002, for discussion). There are un-
this disorder. As importantly, more concerted derstandable and legitimate tensions between
clinical and epidemiological research can help those who wish to have broader diagnostic
further refine our understanding of the nature categories—for example, to ensure service
of this disorder and provide a more solid foun- provision—and those who advocate for more
dation for future investigations of the potential narrowly defined categories—for example,
role that genes play in the etiology of CDD. for research purposes (see Chapter 1, this
volume). With regard to CDD, these issues
Other Neurobiological Factors and tensions also exist. The validity of the
category has been the object of debate. In this
Despite extensive medical evaluations, the yield section, we summarize some of the informa-
in terms of discovering clearly potentially tion on validity of the condition with respect
causative conditions for the developmental re- to autism and other disorders.
gression is surprisingly low. Russo and col-
leagues (1996) have speculated that a Childhood Disintegrative Disorder
combination of genetic factors and environmen- and Autism
tal stress might be involved and, further, that a
specific immune response might be involved. Although Kanner’s original (1943) impression
Shinnar and colleagues (2001) similarly specu- was that autism was congenital, subsequent
late that the widespread parental report of asso- work has consistently shown that in a minority
ciated stressful events with the onset of the of cases parents report some period of months,
condition might have some etiological signifi- or even a year or more, of normal development
cance. Supporting data for such hypotheses are before symptoms of autism are recognized
limited. In general, the recent presumption has (Kolvin, 1971; Kurita, 1985; Short & Schopler,
been that the stresses reported (e.g., birth of a 1988; Siperstein & Volkmar, 2004; Volkmar
sibling) are not highly unusual for children in et al., 1985, 1988). The cases, sometimes re-
this age group. ferred to as late-onset autism or, particularly
While the cause or causes of CDD remain in the Japanese literature, as setback autism,
unknown, the course of the disorder, the asso- have not, unfortunately, been commonly stud-
ciation with EEG abnormalities and seizure ied, and there is some potential for confusion
disorder, and the occasional association with (or even for overlap) with CDD. That is, al-
known medical conditions suggest that neuro- though, by definition, CDD has its onset after
biological factors are likely central in patho- age 2, with normal previous development, it is
genesis. As with autism, it is possible that possible that this is an artificial distinction,
multiple pathogenic pathways may act to and some cases of onset earlier than age 2 have
produce the condition, and the absence of indeed been reported in the literature (see
clearly identified neuropathological mecha- Figure 3.1). Several other issues further com-
nisms likely reflects more on current research plicate this problem. What is reported as “re-
techniques than the absence of such factors gression” often seems to be more adequately
(Rivinus, Jamison, & Graham, 1975; Wilson, characterized as developmental stagnation; for
1974). Despite the progress in our understand- example, the child says one or two words and
ing of the central nervous system, much re- then speech and subsequent development fail
mains to be discovered. The absence of to progress normally. This state of affairs is
specific neurobiological mechanisms is proba- rather different from that reported in prototyp-
bly more a reflection on the present state of ical development of CDD where a marked loss
science than the absence itself. of functioning in multiple areas occurs. In a
80 Diagnosis and Classification

recent study, Siperstein and Volkmar (2004) any associated neurological or other general
noted that language loss and social skills loss medical condition is clearly indicated in CDD
are often prominent in autism when parents re- although, somewhat surprisingly, even exten-
port loss of skills. Their report also notes that, sive evaluations often fail to reveal an associ-
when regression is strictly defined (e.g., based ated medical condition. In their study of
not only on parental report of regression but dementia using a large cohort of cases in Aus-
also of developmental milestones), demonstra- tralia, Nunn, Williams, and Ouvrier (2002)
ble regression was relatively uncommon. In estimated the prevalence of dementia in child-
their follow-up study of nearly 200 cases of hood to be 5.6 per 100,000 cases. Of the chil-
autism who had lost skills, Kobayashi and Mu- dren identified, in 21% of cases a specific
rata (1998) reported that there were higher etiology could not be determined; they suggest
rates of seizure disorder and lower language the general utility of a broad conceptualiza-
levels associated with reported language set- tion of childhood dementia.
back. Davidovitch and colleagues (2000) ex- In addition to the various metabolic and
amined differences between children with genetic disorders that can cause deteriora-
autism who were reported to have regressed tion, the development of an autistic-like clini-
and those who did not regress; they found little cal picture can follow central nervous system
difference between the groups. Language re- infection or other insult (Weir & Salisbury,
gression is not as commonly reported among 1980). Occasionally, epileptic conditions may
children with overall developmental delay also mimic autism or CDD (Deonna, Ziegler,
(Shinnar et al., 2001). Malin-Ingvar, Ansermet, & Roulet, 1995). This
In a recent study, Kurita et al. (2004) evalu- issue has been most controversial with respect
ated the clinical validity of CDD as defined by to the syndrome of acquired aphasia with
DSM-IV in 10 children with the condition who epilepsy (Landau-Kleffner syndrome) in which
were compared to a larger group of age- and aphasia develops in association with multifo-
gender-matched children with DSM-IV autis- cal spike and spike/wave discharges on EEG
tic disorder with speech loss. Compared to the (Beaumanoir, 1992). This syndrome is rela-
children with autism who had exhibited tively well described but remains poorly un-
speech loss, those with CDD exhibited higher derstood, and, in general, the impression is of
levels of anxiety/dysphoria as well as higher an aphasia rather than a PDD (this condition is
levels of stereotyped behavior. The CDD included in ICD-10 as a developmental lan-
group also had a higher rate of seizure disor- guage disorder but is not recognized in DSM-
der, and there was a suggestion of differences IV). Nonverbal abilities tend to be spared, and
in the profiles on psychological testing. Al- there can be considerable recovery (Bishop,
though differences in levels of retardation 1985). In contrast to CDD where later onset of
were not observed, the authors pointed out that the disorder is usually associated with a worse
the relative youth of the group (mean age of prognosis, later onset in Landau-Kleffner is as-
8.2 years for the CDD cases) limited the value sociated with better outcome (Bishop, 1985).
of such a comparison. Other epileptic conditions may also mimic
autism or CDD (Deonna et al., 1995).
Childhood Disintegrative Disorder and There seems little disagreement that, once
Other Disorders established, CDD resembles autism in terms
of clinical features and course (Militerni,
As noted previously, another area of contro- Bravaccio, & D’Anuono, 1997; Volkmar &
versy has centered on the view of CDD as re- Cohen, 1989). Hendry (2000), however, has
flecting a neuropathological process. There is questioned the utility of CDD as a diagnostic
little disagreement that some brain-based mech- concept, noting the strong overlap in symp-
anism is responsible for the condition; most of toms with autism. Hendry rightly notes the
the debate has centered either on the idea that limitations of the available data, which are
CDD is commonly associated with one or more very small as compared, for example, to that
specific and identifiable other medical condi- on autism. She also raises several relevant
tions. As noted earlier, a thorough search for questions about the difficulties in reliance on
Childhood Disintegrative Disorder 81

parental report and so forth. In contrast, Volk- type seen in CDD occurred. The use of mate-
mar and Rutter (1995), in their review of the rials such as baby books, home videotapes, or
results of CDD in the DSM-IV field trial, came movies may be helpful.
to a different conclusion, observing major dif- In CDD, it is essential that the early devel-
ferences both in the dramatic clinical presen- opment be unequivocally normal with the child
tation and course. As they noted, the major able to speak in sentences prior to the onset of
rationale for including CDD as a recognized the condition. If such a history cannot be docu-
diagnostic entity primarily related to its po- mented, a diagnosis of CDD should not be
tential for clarifying basic mechanisms of made; in such cases, a diagnosis of atypical
pathogenesis. autism may be most appropriate. There are
also the exceptions that tend to prove the rule,
DIFFERENTIAL DIAGNOSIS for example, children who were being reared in
a bilingual context and whose language was
The differential diagnosis of CDD includes the slightly delayed but who otherwise appeared
other PDDs and other conditions. Because of to be developing appropriately prior to the
the very distinctive pattern of onset, historical onset of the regression.
information is critical in making the diagnosis.
Rett’s Disorder
Rett’s disorder (see Chapter 5, this volume) is
Once established, CDD shares the same essen- occasionally confused with CDD; this reflects
tial features on current clinical examination. the fact that some degree of regression is ob-
Thus, historical information is particularly served in both conditions and that the more
important in distinguishing the two conditions. autistic-like phase of Rett’s disorder may be
In about 75% of cases of autism, parents do not most prominent in the preschool years, that is,
report an unequivocally normal period of de- when diagnostic evaluations are first con-
velopment; usually social development is ducted. For individuals familiar with both con-
markedly delayed and deviant, and language ditions, misdiagnosis is relatively unlikely
fails to develop in expected ways. In a smaller given the very different histories and clinical
group of cases, the parents report either that features of the conditions.
the child seemed to develop normally, usually
up to about 18 to 24 months of age, or that sin- Asperger Syndrome
gle words developed but the child did not go on
to develop more complex speech; such cases In this condition, early language and cognitive
should not be diagnosed with CDD. It is not un- development may seem to have been normal or
common for the presenting complaint in autism near normal. Sometimes parents become aware
to center on language loss and then for a care- of the social and other difficulties as the child
ful history to establish earlier preexisting ab- enters a preschool program. This may, incor-
normalities, for example, in terms of unusual rectly, lead some care providers to suspect that
environmental sensitivities or responses (e.g., there was a marked regression in functioning.
Stein, Dixon, & Cowan, 2000). Complexities However, in Asperger, cognitive functions are
may arise if historical information is absent, relatively preserved, and the truly marked and
the report of a truly prolonged period of nor- severe regression of the kind seen in CDD is
mal development is questionable, the clinician absent.
fails to obtain an adequate history, or the
autistic child is somewhat higher functioning. Childhood Schizophrenia
In the last case, it is possible that early lan-
guage seems to develop in a near normal fash- In rare instances, there may be confusion of
ion, but the parent becomes concerned as more CDD with other psychiatric disorders such as
complex social-communicative difficulties be- schizophrenia. Although very early onset
come apparent. Careful history will usually re- schizophrenia (VERS) is very rare, the degree
veal that no actual period of regression of the of regression and deterioration may suggest
82 Diagnosis and Classification

CDD (see Werry, 1992). However, usually the EVALUATION AND MANAGEMENT
characteristic findings of schizophrenia on
clinical examination will clarify the diagnosis. Assessment of the child with CDD may be most
effectively accomplished by a group of profes-
Landau-Kleffner Syndrome sionals who work together as a team or in close
collaboration with one another. Referrals usu-
The syndrome of acquired aphasia with ally come from primary care providers al-
epilepsy (Landau-Kleffner syndrome) has its though occasionally mental health workers,
onset in children and is characterized by ac- educators, and others may question a diagnosis,
quired aphasia in association with multifocal for example, of autism, because of some fea-
spike and spike/wave discharges on EEG tures of the case that they know to be unusual.
(Beaumanoir, 1992). This syndrome is rela- Given the potential for multiple assessments, it
tively well described but remains poorly is important that the professionals involved in
understood (S. Wilson, Djukic, Shinnar, the child’s care work effectively to avoid frag-
Dharmani, & Rapin, 2003). Children with the mentation and duplication of effort. Various
Landau-Kleffner syndrome usually exhibit professionals may be involved, for example,
marked interest in communication using child psychiatrists, psychologists, speech
nonverbal modalities, and typically the re- pathologists, pediatric neurologists, occupa-
gression is largely confined to the area of lan- tional and physical therapists, and others. It is
guage. Nonverbal abilities tend to be spared, appropriate and desirable to engage parents ac-
and there is often considerable recovery tively in the assessment process.
(Bishop, 1985). In contrast to CDD where Given the unusual history of the onset of
later onset of the disorder is usually associ- CDD, a careful history is particularly critical.
ated with a worse prognosis, later onset in This history should include information re-
Landau-Kleffner is associated with better lated to the pregnancy and neonatal period,
outcome (Bishop, 1985). Other epileptic con- early developmental history, as well as med-
ditions may also mimic autism or CDD (De- ical and family history. As noted previously,
onna et al., 1995), and, as noted previously, a the examination of videotapes may clearly
history of language regression should prompt document the child’s early normal develop-
a careful assessment for possible seizure dis- ment. Information on the pattern and age of
order (Shinnar et al., 2001). onset of the condition is central.
Although extensive medical investigations
Other Associated Medical Conditions usually fail to reveal the presence of another
specific medical condition or specific neu-
As noted earlier, a thorough search for any as- rodegenerative disorder, a careful search for
sociated neurological or other general medical such conditions is indicated. This is particu-
condition is indicated in CDD. Specific find- larly true if aspects of the case are unusual; for
ings on examination or in the history may help example, if the onset is rather later (after age
guide the process of evaluation. If any such 6), or if the deterioration is progressive and
condition is identified, it is then specified does not plateau. Initial consultation with a
on Axis III. Disorders with onset after some pediatric neurologist is always indicated.
period of normal development include gan- Given the severity of these conditions, an EEG
gliosidosis, metachromatic leukodystrophy, and CT or MRI scan are also usually obtained.
Niemman-Pick disease, and so forth. Develop- Tests of communication and cognitive abil-
mental deterioration and the development of ity should be chosen with consideration of the
an autistic-like clinical picture can follow child’s current levels of functioning to obtain
central nervous system infection or other in- estimates of functioning useful in document-
sult (Weir & Salisbury, 1980). Rarely clini- ing subsequent developmental change as well
cally significant regression occur in the as for educational and rehabilitative program-
context of overt seizures (Tuchman & Rapin, ming. The use of developmental and other
1997; S. Wilson et al., 2003). tests typically given to younger children may
Childhood Disintegrative Disorder 83

be appropriate, for example (Bayley, 1969; specific pharmacological treatments for CDD
Dunst, 1980; Uzgiris & Hunt, 1975). For although it is likely that many of the same
somewhat higher functioning but nonverbal agents used with some benefit in autism will
children, the Leiter International Perfor- sometimes also be helpful. The use of any
mance Scale (Leiter, 1948) may be useful. As pharmacological agent should include a careful
with children with autism, modifications in assessment of potential risks and benefits.
usual assessment procedures may have to be The families (parents, sibs, and extended
made. Communication scales that may be ap- family members) of patients should be sup-
propriate include the Receptive-Expressive ported. This includes provision of appropriate
Emergent Language Scale (REEL; Bzoch & information about the condition, helping fami-
League, 1971); the Sequenced Inventory lies make use of available local and other re-
of Communicative Development (SICD; sources, and helping family members receive
Hedrick, Prather, & Tobin, 1975); and the mutual support, for example, through support
Reynell Developmental Language Scales groups. Given that the prognosis of CDD ap-
(Reynell & Gruber, 1990). pears, in general, to be somewhat worse than
The Vineland Adaptive Behavior Scales that of autism, the stresses experienced by
(expanded form; Sparrow, Balla, & Cicchetti, parents and siblings may also be greater.
1984) should be administered to document lev-
els of adaptive behaviors. This instrument pro- CASE REPORT
vides useful information both for diagnostic
and programming purposes. Donald was the youngest of three children
As part of the psychiatric or psychological born to college-educated parents. The preg-
assessment, the child should be observed in nancy, labor and delivery, and early develop-
more and less structured activities, for exam- ment were unremarkable. He appeared to be a
ple, during developmental assessment or while normally active and sociable baby. He was
interacting with parents. As part of the history, smiling at 6 weeks, sitting at 7 months, crawl-
the examiner should specifically inquire about ing at 9 months, and walking without support
the child’s current and past social skills (e.g., at 15 months. He had several ear infections in
deferential attachments, interest in parents and the first year of life but said his first words by
peers, use of gaze), communication (receptive 12 months and was speaking in full sentences
and expressive, articulation problems, typical shortly after his second birthday. Videotapes
utterances, level of language organization prior provided by his parents confirmed his appar-
to the regression, unusual features such as ently normal developmental status.
echolalia, and nature of language loss), and re- Shortly after his third birthday, Donald’s
sponses to the environment and motor behaviors parents became concerned about his develop-
(e.g., self-stimulatory behaviors, difficulties ment as, over the course of several weeks, he
with change or transitions). The acquisition and lost both receptive and expressive language
any loss of adaptive skills, for example, toilet- and became progressively less interested
ing, self-care, or related skills, should also be in interaction and the inanimate environ-
reviewed. Observation of the child’s play is ment. He developed various self-stimulatory
helpful, for example, in documenting levels of behaviors and lost the ability to use the toilet
language, cognitive, and social organization independently.
and in observing gross and fine motor skills. No apparent reason for the regression was
Problematic or unusual behaviors relevant to identified. Extensive medical investigations
the diagnosis or likely to present obstacles for were undertaken. Although he was noted to
intervention should be noted. have a borderline abnormal EEG, no specific
Treatment of CDD is essentially the same medical condition that might account for his
as for autism. Methods such as behavior modi- developmental deterioration was identified.
fication and special education should be used Subsequent evaluations at other centers and
to help encourage the acquisition, or reacqui- by other specialists (including pediatric neu-
sition, of basic adaptive skills. There are no rologists and geneticists) similarly failed to
84 Diagnosis and Classification

identify such a condition. There was no family REFERENCES

history of similar problems or developmental
difficulties in members of the immediate or American Psychiatric Association. (1980). Diag-
extended family. nostic and statistical manual of mental disor-
Although he had previously been in a regu- ders (3rd ed.). Washington, DC: Author.
American Psychiatric Association. (1987). Diag-
lar nursery school setting, his behavior and de-
nostic and statistical manual of mental disor-
velopmental deterioration warranted his ders (3rd ed., rev.). Washington, DC: Author.
placement in a special educational setting. At American Psychiatric Association. (1994). Diag-
4 years old, comprehensive evaluation re- nostic and statistical manual of mental disor-
vealed that his cognitive skills were at about ders (4th ed.). Washington, DC: Author.
the 18-month level with language and social Amir, R. E., Van den Veyver, I. B., Wan, M., Tran,
skills at an even lower level. Over the course of C. Q., Francke, U., & Zoghbi, H. Y. (1999).
many months, some, although highly limited, Rett syndrome is caused by mutations in X-
expressive language skills returned as he was linked MECP2, encoding methyl-CpG-binding
able to say an occasional single word. His so- protein 2 [see comment]. Nature Genetics,
23(2), 185–188.
cial unrelatedness, lack of interest in the envi-
Amir, R. E., & Zoghbi, H. Y. (2000). Rett syn-
ronment, and unusual behaviors continued. At
drome: Methyl-CpG-binding protein 2 muta-
the time of follow-up, at age 16, he continues tions and phenotype-genotype correlations.
to be severely impaired. American Journal of Medical Genetics, 97(2),
CONCLUSION Bayley, N. (1969). Bayley scales of infant develop-
ment. New York: Psychological Corporation.
Although described nearly 100 years ago, Beaumanoir, A. (1992). The Landau-Kleffner syn-
CDD remains infrequently reported and very drome. In J. Roger, M. Bureau, C. Dravet, F. E.
Dreifuss, A. Perret, & P. Wolf (Eds.), Epileptic
uncommonly studied. As noted by Rapin (in
seizures in infancy, childhood, and adolescence
press), the classification of children who de-
(2nd ed., pp. 231–243). London: John Libbey.
velop an autistic-like condition after several Bishop, D. V. M. (1985). Age on onset and outcome
years of normal development remains very in “acquired aphasia with convulsive disorder ”
poorly understood. It is clear that, once the (Landau-Kleffner syndrome). Developmental
condition is established, it behaviorally is very Medicine and Child Neurology, 27, 705–712.
similar to autism. The condition differs from Bishop, D. V. M. (1994). Developmental disorders
autism in the nature and pattern of its onset of speech and language. In M. Rutter, E. Tay-
and in course and certain clinical features. lor, & L. Hersov (Eds.), Child and adolescent
While it appears very likely that some under- psychiatry: Modern approaches (3rd ed.,
lying neurobiological basis, or bases, of the pp. 546–568). London: Blackwell.
Burd, L., Fisher, W., & Kerbeshian, J. (1988).
condition exists, precise etiological mecha-
Childhood onset pervasive developmental dis-
nisms have not been identified. Past its obvi- order. Journal of Child Psychology and Psychi-
ous importance for affected children and their atry and Allied Disciplines, 29(2), 155–163.
families, studies of such mechanisms may be Burd, L., Fisher, W., & Kerbeshian, J. (1989).
important in providing a better understanding Pervasive disintegrative disorder: Are Rett
of mechanisms of syndrome pathogenesis that syndrome and Heller dementia infantilis sub-
may have relevance to autism (Darby & Clark, types? Developmental Medicine and Child
1992). Neurology, 31(5), 609–616.
Burd, L., Ivey, M., Barth, A., & Kerbeshian, J.
(1998). Two males with childhood disintegra-
Cross-References tive disorder: A prospective 14-year outcome
study. Developmental Medicine and Child Neu-
Issues of diagnosis and classification are dis- rology, 40(10), 702–707.
cussed in Chapters 1 through 7; medical and Bzoch, K., & League, R. (1971). Receptive Expres-
neurological conditions potentially relevant sive Emergent Language Scale. Gainesville,
to CDD are reviewed in Chapters 16, 18, 19, FL: Language Educational Division, Computer
and 20. Management Corporation.
Childhood Disintegrative Disorder 85

Corbett, J. (1987). Development, disintegration Hedrick, D., Prather, F., & Tobin A. (1975). Se-
and dementia. Journal of Mental Deficiency quenced inventory of communicative develop-
Research, 31(4), 349–356. ment. Seattle: University of Washington Press.
Couvert, P., Bienvenu, T., Aquaviva, C., Poirier, Heller, T. (1908). Dementia infantilis. Zeitschrift
K., Moraine, C., Gendrot, C., et al. (2001). fur die Erforschung und Behandlung des Ju-
MECP2 is highly mutated in X-linked mental genlichen Schwachsinns, 2, 141–165.
retardation. Human Molecular Genetics, 10, Heller, T. (1930). Uber Dementia infantalis.
941–946. Zeitschrift fur Kinderforschung, 37, 661–667.
Darby, J. K., & Clark, L. (1992). Autism syndrome Hendry, S. N. (2000). Childhood disintegrative dis-
as a final common pathway of behavioral ex- order. Clinical Psychology Review, 20, 77–90.
pression for many organic disorders. American Hill, A. E., & Rosenbloom, L. (1986). Disintegra-
Journal of Psychiatry, 149(1), 146. tive psychosis of childhood: Teenage follow-
Davidovitch, M., Glick, L., Holtzman, G., Tirosh, up. Developmental Medicine and Child
E., & Safir, M. P. (2000). Developmental re- Neurology, 28(1), 34–40.
gression in autism: Maternal perception. Jour- Kanner, L. (1943). Autistic disturbances of affec-
nal of Autism and Developmental Disorders, tive contact. Nervous Child, 2, 217–250.
30(2), 113–119. Kleefstra, T., Yntema, H. G., Oudakker, A. R., de
Deonna, T., Ziegler, A., Malin-Ingvar, M., Anser- Vries, B. B., van Bokhoven, H., Hamel, B. C.,
met, F., & Roulet, E. (1995). Reversible behav- et al. (2002). Localization of a gene for non-
ioral autistic-like regression: A manifestation specific X-linked mental retardation (MRX
of a special (new?) epileptic syndrome in a 28 76) to Xp22.3-Xp21.3. American Journal of
month old child: A 2 year longitudinal study. Medical Genetics, 110, 410–411.
Neurocase, 1, 1–9. Kobayashi, R., & Murata, T. (1998). Setback
Deykin, E. Y., & MacMahon, B. (1979). The inci- phenomenon in autism and long term progno-
dence of seizures among children with autistic sis. Archives Psychiatrica Scandinavia, 98,
symptoms. American Journal of Psychiatry, 296–303.
136(10), 1310–1312. Kolvin, I. (1971). Studies in childhood psychoses: I.
Dunst, C. (1980). A clinical and educational man- Diagnostic criteria and classification. British
ual for use with the Uzgiris and Hunt Scales Journal of Psychiatry, 118, 381–384.
Baltimore: University Park Press. Kurita, H. (1985). Infantile autism with speech
Dyken, P., & Krawiecki, N. (1983). Neurodegenera- loss before the age of thirty months. Journal of
tive diseases of infancy and childhood. Annals the American Academy of Child Psychiatry,
of Neurology, 13, 351–364. 24(2), 191–196.
Evans-Jones, L. G., & Rosenbloom, L. (1978, Au- Kurita, H. (1988). The concept and nosology of
gust). Disintegrative psychosis in childhood. Heller’s syndrome: Review of articles and re-
Developmental Medicine and Child Neurology, port of two cases. Japanese Journal of Psychia-
20(4), 462–470. try and Neurology, 42, 785–793.
Fombonne, E. (2002). Prevalence of childhood dis- Kurita, H. (1989). Heller’s syndrome as a type of
integrative disorder. Autism: Journal of Re- pervasive developmental disorder. Journal of
search and Practice, 6(2), 149–157. Mental Health, 35, 71–81.
Gillberg, C., Terenius, L., Hagberg, B., Witt- Kurita, H., Kita, M., & Miyake, Y. (1992). A com-
Engerstrom, I., & Eriksson, I. (1990). CSF parative study of development and symptoms
beta-endorphins in childhood neuropsychi- among disintegrative psychosis and infantile
atric disorders. Brain Development, 12(1), autism with and without speech loss. Journal
88–92. of Autism and Developmental Disorders, 22(2),
Girard, M., Couvert, P., Carrie, A., Tardieu, M., 175–188.
Chelly, J., Beldjord, C., et al. (2001). Parental Kurita, H., Osada, H., & Miyake, Y. (2004). Exter-
origin of de novo MECP2 mutations in Rett nal validity of childhood disintegrative disor-
syndrome. European Journal of Human Genet- der in comparison with autistic disorder.
ics, 9(3), 231–236. Journal of Autism and Developmental Disor-
Hammer, S., Dorrani, N., Dragich, J., Kudo, S., & ders, 34, 355–362.
Schanen, C. (2002). The phenotypic conse- Lainhart, J. E., Ozonoff, S., Coon, H., Krasny, L.,
quences of MECP2 mutations extend beyond Dinh, E., Nie, J., et al. (2002). Autism, re-
Rett syndrome. Mental Retardation and Devel- gression, and the broader autism phenotype.
opmental Disabilities Research Reviews, 8, American Journal of Medical Genetics, 113,
94–98. 231–237.
86 Diagnosis and Classification

Leiter, R. G. (1948). Leiter International Perfor- Rogers, S. J., & DiLalla, D. L. (1990). Age of
mance Scale. Chicago: Stoelting. symptom onset in young children with Perva-
Lord, C., Schopler, E., & Revicki, D. (1982). Sex sive Developmental Disorder. Journal of the
differences in autism. Journal of Autism and American Academy of Child and Adolescent
Developmental Disorders, 12, 317–330. Psychiatry, 29, 863–872.
Makita, K. (1966). The age of onset of childhood Russo, M., Perry, R., Kolodny, E., & Gillberg, C.
schizophrenia. Folia Psychiatric Neurologica (1996). Heller syndrome in a pre-school boy.
Japonica, 20, 111–112. Proposed medical evaluation and hypothe-
Malhotra, S., & Gupta, N. (1999). Childhood disin- sized pathogenesis. European Child and Ado-
tegrative disorder. Journal of Autism and De- lescent Psychiatry, 5(3), 172–177.
velopmental Disorders, 29, 491–498. Rutter, M. (1972, October). Childhood schizophre-
Malhotra, S., & Gupta, N. (2002). Childhood disin- nia reconsidered. Journal of Autism and Child-
tegrative disorder: Re-examination of the cur- hood Schizophrenia, 2(4), 315–337.
rent concept. European Child and Adolescent Rutter, M. (1985). Infantile autism and other per-
Psychiatry, 11(3), 108–114. vasive developmental disorders. In M. Rutter
Malhotra, S., & Singh, S. (1993). Disintegrative & L. Hersov (Eds.), Child and adolescent psy-
psychosis of childhood: An appraisal and case chiatry—Modern approaches (pp. 545–566).
study. Acta Paedopsychiatrica, 56, 37–40. London: Blackwell.
Meloni, I., Bruttini, M., Longo, I., Mari, F., Rizzo- Rutter, M. (2003). Categories, Dimensions, and the
lio, F., D’Adamo, P., et al. (2000). A mutation Mental Health of Children and Adolescents.
in the Rett syndrome gene, MECP2, causes Annals of the New York Academy of Science,
X-linked mental retardation and progressive 1008, 1–12.
spasticity in males. American Journal of Rutter, M., Bailey, A., Bolton, P., & Le Couter, A.
Human Genetics, 67, 982–985. (1994). Autism and known medical conditions:
Militerni, R., Bravaccio, C., & D’Anuono, P. Myth and substance. Journal of Child Psychol-
(1997). Childhood disintegrative disorder: ogy and Psychiatry and Allied Disciplines,
Review of cases and pathological considera- 35(2), 311–322.
tion. Developmental Brain Dysfunction, 10, Rutter, M., & Taylor, E. (2002). Classification:
67–74. Conceptual issues and substantive findings. In
Millichap, J. G. (1987). Rett’s syndrome: A variant E. Taylor & M. Rutter (Eds.), Child and ado-
of Heller’s dementia? [Letter]. Lancet, 1, 440. lescent psychiatry (4th ed., pp. 3–17). Oxford,
Mouridsen, S. E., Rich, B., & Isager, T. (1998). Va- UK: Blackwell Publishing.
lidity of childhood disintegrative psychosis. Shinnar, S., Rapin, I., Arnold, S., Tuchman, R.,
General findings of a long-term follow-up Shulman, L., Ballabasn-Gil, K., et al. (2001).
study. British Journal of Psychiatry, 172, Language regression in childhood. Pediatric
263–267. Neurology, 24, 183–194.
Mouridsen, S. E., Rich, B., & Isager, T. (2000). A Short, A., & Schopler, E. (1988). Factors relating
comparative study of genetic and neurobiolog- to age of onset in autism. Journal of Autism
ical findings in disintegrative psychosis and and Developmental Disorders, 18, 207–216.
infantile autism. Psychiatry and Clinical Neu- Siegel, B., Pliner, C., Eschler, J., & Elliott, G. R.
rosciences, 54(4), 441–446. (1989). How children with autism are diag-
Nunn, K., Williams, K., & Ouvrier, R. (2002). The nosed: Difficulties in identification of chil-
Australian Childhood Dementia Study. Euro- dren with multiple developmental delays.
pean Child and Adolescent Psychiatry, 11, Developmental and Behavioral Pediatrics, 9,
63–70. 199–204.
Rett, A. (1966). Uber ein eigenartiges hirn- Sparrow, S. S., Balla, D., & Cicchetti, D. V. (1984).
tophisces Syndroem bei hyperammonie im Vineland adaptive behavior scales (Survey
Kindersalter. Wein Medizinische Wochen- Form). Circle Pines, MN: American Guidance
schrift, 118, 723–726. Service.
Reynell, J., & Gruber, C. (1990). Reynell Develop- Stein, M. T., Dixon, S. D., & Cowan, C. (2000). A
mental Language Scales—US Edition. Los An- two-year-old boy with language regression and
geles, CA: Western Psychological Services. unusual social interactions. Journal of Devel-
Rivinus, T. M., Jamison, D. L., & Graham, P. J. opmental and Behavioral Pediatrics, 21(4),
(1975). Childhood organic neurological dis- 285–290.
ease presenting as a psychiatric disorder. Tuchman, R. F., & Rapin, I. (1997). Regression in
Archives of Disease in Childhood, 50, 115–119. pervasive developmental disorders: Seizures
Childhood Disintegrative Disorder 87

and epileptiform electroencephalogram corre- disorder. American Journal of Psychiatry,

lates. Pediatrics, 99(4), 560–566. 142(12), 1450–1452.
Uzgiris, I. C., & Hunt, J. M. (1975). Assessment in Weir, K., & Salisbury, D. M. (1980). Acute onset of
infancy: Ordinal scales of psychological devel- autistic features following brain damage in a
opment. Urbana: University of Illinois Press. ten-year-old. Journal of Autism and Develop-
Volkmar, F. R. (1987). Diagnostic issues in the per- mental Disorders, 10(2), 185–191.
vasive developmental disorders. Journal of Werry, J. S. (1992). Child and adolescent (early
Child Psychology and Psychiatry and Allied onset) schizophrenia: A review in light of
Disciplines, 28(3), 365–369. DSM-III-R. Journal of Autism and Develop-
Volkmar, F. R. (1992). Childhood disintegrative mental Disorders, 22(4), 601–624.
disorder: Issues for DSM-IV. Journal of Autism Wilson, J. (1974). Investigation of degenerative
Developmental Disorders, 22, 625–642. disease of the central nervous system. Archives
Volkmar, F. R. (1996). The “disintegrative” disor- of Disease in Childhood, 47, 163–170.
ders. In F. Volkmar (Ed.), Childhood psy- Wilson, S., Djukic, A., Shinnar, S., Dharmani, C.,
choses and pervasive developmental disorders & Rapin, I. (2003). Clinical characteristics of
(pp. 223–248). Washington, DC: American language regression in children. Developmen-
Psychiatric Association Press. tal Medicine and Child Neurology, 45(8),
Volkmar, F. R., & Cohen, D. J. (1989). Disintegra- 508–514.
tive disorder or “late onset ” autism. Journal of Wohlgemuth, D., Kiln, A., Cohen, D. J., & Volkmar,
Child Psychology and Psychiatry and Allied F. R. (1994, October). Childhood disintegrative
Disciplines, 30(5), 717–724. disorder: Diagnosis and phenomenology. Paper
Volkmar, F. R., Cohen, D. J., Hoshino, Y., Rende, presented at the annual meeting of the Ameri-
R. D., & Paul, R. (1988). Phenomenology and can Academy of Child and Adolescent Psychia-
classification of the childhood psychoses. Psy- try, New York.
chological Medicine, 18(1), 191–201. World Health Organization. (1978). Mental disor-
Volkmar, F. R., Klin, A., Marans, W. D., & Cohen, ders: Glossary and guide to their classification
D. J. (1997). Childhood disintegrative disor- in accordance with the ninth revision of the in-
der. In D. J. Cohen & F. R. Volkmar (Eds.), ternational classification of diseases. Geneva,
Autism and pervasive developmental disorders Switzerland: Author.
(2nd ed., pp. 47–59). New York: Wiley. World Health Organization. (1990). International
Volkmar, F. R., Klin, A., Siegel, B., Szatmari, P., Classification of Diseases (Draft version, 10th
Lord, C., Campbell, M., et al. (1994). Field ed.). Geneva, Switzerland: Author.
trial for autistic disorder in DSM-IV. American Zappert, J. (1921). Dementia infatilies Heller.
Journal of Psychiatry, 151(9), 1361–1367. Monatsschrift fur Kinderheilkunde, 22, 389–391.
Volkmar, F. R., & Rutter M. (1995). Childhood dis- Zwaigenbaum, L., Szatmari, P., Mahoney, W.,
integrative disorder: Results of the DSM-IV Bryson, S., Bartolucci, G., & MacLean, J.
autism field trial. Journal of the American (2000). High functioning autism and child-
Academy of Child and Adolescent Psychiatry, hood disintegrative disorder in half brothers.
34, 1092–1095. Journal of Autism and Developmental Disor-
Volkmar, F. R., Stier, D. M., & Cohen, D. J. (1985). ders, 30(2), 121–126.
Age of recognition of pervasive developmental

Asperger Syndrome


Asperger syndrome (AS) is a severe and general population in an attempt to define a

chronic developmental disorder closely related continuum of social function ranging from se-
to autistic disorder and pervasive developmen- vere autism to “normalcy.”
tal disorder-not otherwise specified (PDD- The confusions and controversies surround-
NOS), and, together, these disorders comprise ing the definition and validity of AS mean that
a continuum referred to as the autism spectrum the nosological status of AS as a discrete con-
disorders (ASDs). Having autism as the para- dition separate from autism is still uncertain
digmatic and anchoring disorder in this diag- (Frith, 2004; Volkmar & Klin, 2000). The
nostic category, the ASDs more generally are usage of the term, however, has been justified
characterized by marked and enduring impair- in pragmatic terms to bring more attention to
ments within the domains of social inter- individuals with autism with higher verbal
action, communication, play and imagination, abilities (Wing, 2000) or to less disabled indi-
and a restricted range of behaviors or inter- viduals but with social vulnerabilities accom-
ests. Although the diagnostic criteria for AS panied by a learning style conducive of
are still evolving, the “official” definitions technical knowledge or skills (Baron-Cohen,
provided in the International Classification of Wheelwright, Stone, & Rutherford, 1999) and
Diseases, 10th edition (ICD-10; World Health as a model illustrating the need for research on
Organization [WHO], 1993) and the Diagnos- factors mediating specific manifestations and
tic and Statistical Manual of Mental Disorders, outcome in the ASDs (Szatmari, 2000). It is
fourth edition (DSM-IV; American Psychiatric important to keep these various uses of the
Association, 2000) distinguish it from autism term AS as the backdrop to any discussion of
primarily on the basis of a relative preserva- its nosologic and validity status to avoid less
tion of linguistic and cognitive capacities in productive lines of inquiry meant to reconcile
the first 3 years of life. Both in practice and in clinical research data yielded by studies using
research studies, however, the term AS has different definitions, criteria, and approaches
been used to refer to individuals with variedly to this condition. The prevailing attitude
defined manifestations of the ASDs, including among researchers, clinicians, and advocates
autism without mental retardation (or higher alike, however, is that regardless of these vari-
functioning autism [HFA]), “milder ” forms of ous uncertainties, individuals with AS have an
autism marked by higher cognitive and lin- early-onset social disability that impairs their
guistic abilities, and more socially motivated capacity for meeting the demands of everyday
but socially vulnerable adolescents and adults life and that this is an uncontroversial clinical
with unusual and socially interfering circum- fact necessitating, therefore, systematic re-
scribed interests. AS has also been used as a search involving all tools available to clinical
conceptual bridge between autism and the science.

Asperger Syndrome 89

In this chapter, we describe the historical in understanding nonverbal cues conveyed

background of AS, its clinical features, and by others.
special considerations for assessment and in- • Idiosyncrasies in verbal communication:
tervention. We review current diagnostic is- Spontaneous communication is character-
sues including validity studies and suggest ized by highly circumstantial utterances
potentially fruitful guidelines for future (e.g., failing to distinguish abstractions
research. from autobiographical narration), long-
winded and incoherent verbal accounts fail-
ASPERGER’S ORIGINAL CLINICAL ing to convey a clear message or thought
CONCEPT (e.g., tangential speech prompted by a se-
ries of associations), and one-sidedness
Hans Asperger (1906–1980) was an Austrian (e.g., failing to demarcate changes of topic
pediatrician with interest in special education or to introduce new material). Both the
who in 1944 described four children ages 6 to style of language and tone of delivery are
11 years who had difficulty integrating so- pedantic, like a “little professor.”
cially into groups despite seemingly adequate • Social adaptation and special interests:
cognitive and verbal skills (1944; Frith pro- There are egocentric preoccupations with
vided an English translation of the original ar- unusual and circumscribed interests that
ticle in 1991, adding an account of Asperger’s absorb most of the person’s attention and
professional background). He originally termed energy, thus precluding the acquisition of
the condition he described Autistischen Psy- practical skills necessary for self-help and
chopathen im Kindesalter, or autistic personal- social integration. Some of these interests,
ity disorders in childhood, echoing Bleuler’s for example, in letters and numbers, are
(1916) use of the term autism in schizophrenia often quite precocious. Later in life, spe-
to signify extreme egocentrism, or the shutting cial interests tend to evolve into specific,
off of relations between the person affected topic-related collections of information
and other people. He contrasted this condition abundant with encyclopedic knowledge,
from schizophrenia, however, by emphasizing such as astronomy or geography.
the stable and enduring nature of the social • Intellectualization of af fect: Emotional pre-
impairments in autistic psychopathy and by sentation is marked by poor empathy, the
voicing the optimistic view that unlike in tendency to intellectualize feelings, and an
schizophrenia, his patients were able to even- accompanying absence of intuitive under-
tually develop some relationships. The term standing of other people’s affective experi-
psychopathy is best translated as personality ences and communication.
disorder (i.e., a group of stable personality • Clumsiness and poor body awareness: Mo-
traits). In Asperger’s time and cultural ethos, toric presentation is characterized by odd
this term did not have the negative connotation posture and gait, poor body awareness, and
of its present usage. The choice of the label clumsiness. Asperger emphasized his pa-
autistic psychopathy also reflected Asperger’s tients’ inability to participate in group
belief that difficulties in socialization repre- sports or other self-help activities involv-
sented the defining feature of the condition. ing motor coordination and integration and
However, he also detailed various clusters of described in detail their poor graphomotor
behavioral symptoms and clinical features that skills.
appeared in conjunction with the social dis- • Conduct problems: The most common
ability. His observations continue to be incor- reason for clinical referral of Asperger’s
porated into current definitions of the disorder patients involved failure at school and
and include: associated behavioral problems including
aggressiveness, noncompliance, and nega-
• Impairment in nonverbal communication: tivism, which were accounted for in terms
There is a reduction in the quantity and di- of their deficits in social understanding
versity of facial expressions and limitations and extreme pursuit of highly circum-
in the use of gesture, as well as difficulties scribed interests. Asperger was particularly
90 Diagnosis and Classification

concerned about his patients’ poor social EVOLVING CONCEPTUALIZATION

adjustment and how they were mercilessly OF ASPERGER SYNDROME
bullied and teased by peers.
• Onset: Asperger thought that the condition Discussions of Asperger’s work were not avail-
could not be recognized in early childhood. able in English until the 1970s. Prior to that,
Speech and language skills as well as cu- however, a handful of related publications in
riosity about the environment in general, English with some relevance to his work had
including people, were thought not to be appeared. For example, Robinson and Vitale
conspicuously deviant. (1954) described three cases of children ages 8
• Familial and gender patterns: Some 30 to 11 who showed a pattern of circumscribed
years prior to the first publication revealing interests reminiscent of Asperger’s patients.
genetic contributions to the etiology of the The children were interested in developmen-
ASDs, Asperger highlighted the familial tally precocious topics such as chemistry,
nature of the condition affecting his pa- nuclear fission, transportation systems, as-
tients, suggesting that similar traits in par- tronomy, electricity, and mortgages, about
ents or relatives were found in almost every which they talked incessantly in one-sided
single case. His patients were almost exclu- conversations with adults and peers. These
sively boys. children were generally socially isolated, play
with other children had to revolve around their
The year before Asperger’s article on autis- interests, and they had overly blunt manners
tic psychopathy appeared in the German child and no concern about their grooming and other
psychiatry literature, Leo Kanner (1943) had social niceties. Kanner (1954) was the invited
published his classic description of 11 children discussant of this paper. The descriptions re-
with “autistic disturbances of affective con- minded him of infantile autism, but he felt that
tact.” Though both authors were unaware of the these children were less socially withdrawn
other’s work, there were many commonalities and more affectively engaged with others, al-
between their patients, including problems with though they lacked friends and did not partici-
social interaction, affect, and communication, pate in group activities. The circumscribed
as well as unusual and idiosyncratic patterns of interests dominated the lives of these children,
interest. The main differences related to As- monopolizing their learning and interfering
perger’s observations that his patients’ speech with their ability to engage others in recipro-
and language acquisition were less commonly cal relationships. Yet, the interests were often
delayed, motor deficits were more likely, onset unusual, consisting of collections of facts
was later in childhood, and all the initial cases rather than conceptual inquiries, making them
occurred in boys. More importantly, however, even more difficult to share with others as a
there were significant differences in terms of hobby or to build on as a possible vocational
aspects and severity of symptomatology in the possibility. It is of interest, therefore, that
areas of social-emotional functioning, speech, Kanner’s incisive clinical observations greatly
language and cognitive skills, motor manner- resembled those of Asperger’s when he was
isms, and circumscribed interests, which to a called to describe children similar to those en-
great extent may have been a function of the countered by Asperger in his original case
specific patients seen by Kanner (who were studies.
primarily preschoolers, less verbal, and more The concept of autistic psychopathy was in-
cognitively disabled) and by Asperger (who troduced to a wider English-speaking reader-
were primarily school-age children, highly ver- ship by Van Krevelen (1963), who made a
bal, and cognitively able). Consequently, Kan- deliberate attempt to distinguish it from Kan-
ner’s description became associated with the ner’s autism. In Van Krevelen’s view (1971),
“classically” cognitively impaired or “lower the conditions are sharply different. Kanner’s
functioning” child with autism, whereas As- autism is manifested from the first months of
perger’s description lends itself more readily life, the child walks before he or she talks,
to an association with the more cognitively speech is delayed or absent, language never at-
able and highly verbal older child with autism. tains the function of communication, there is a
Asperger Syndrome 91

lack of interest in others, and prognosis is poor. outnumbered females, around 20% were girls.
In contrast, Asperger’s condition is manifested These modifications blurred the distinctions
from the third year of life or later, the child originally suggested by Van Krevelen (1971)
talks before he or she walks, language aims at and reiterated by Asperger (1979), bringing
communication but remains one-sided, the the syndrome into an autism spectrum of dis-
child seeks interaction but in an awkward fash- abilities defined by Wing in terms of the triad
ion, and prognosis is rather good. In many of impairments involving social, communica-
ways, van Krevelen’s distinctions remain the tion, and imaginative activities (Wing, 2000).
core argument for those supporting a separa- Wing’s (1981) publication sparked a great
tion between the two conditions. Neither Kan- deal of interest, and the quantity of research
ner (1954) nor Asperger (in at least most of his publications and clinical studies addressing
early writings) ever appeared to make a strong AS has continued to grow steadily (Klin,
argument for a more categorical separation of Volkmar, & Sparrow, 2000). Much of this lit-
the two conditions (though Kanner never pub- erature concerns an attempt to define the dis-
lished an explicit account of his views of As- tinction (or lack thereof ) between AS and
perger’s work, and Asperger both limited the HFA. Historically, researchers have used the
use of autistic psychopathy to cognitively able label idiosyncratically, taking liberties to
and highly verbal individuals and appeared to modify or to emphasize elements of the syn-
support a distinction later in life; see Asperger, drome depending on their clinical experiences
1979; and Hippler & Klicpera, 2003). or theoretical stance. This variability renders
An influential review and series of case re- it virtually impossible to compare findings of
ports by Lorna Wing (1981) finally popular- studies adopting different definitions known
ized Asperger’s work among English readers. to yield, not surprisingly, different groups for
She reported 34 cases, ages 5 to 35, of whom comparison (e.g., Ghaziuddin, Tsai, & Ghazi-
19 had a clinical presentation similar to As- uddin, 1992; Klin, Pauls, Schultz, & Volkmar,
perger’s account, whereas 19 had a consistent in press). Some progress has been made to-
current presentation but did not have the char- ward this end through the establishment of
acteristic onset patterns and early history. “official,” though tentative, diagnostic crite-
Concerned that the term psychopathy might ria now incorporated in ICD-10 (WHO, 1993)
connote sociopathic behavior rather than the and DSM-IV (American Psychiatric Associa-
intended personality disorder and hoping to tion, 1994), although this definition has itself
ground the condition in developmental terms, been very controversial.
she renamed the disorder Asperger’s syn-
drome. Wing summarized Asperger’s descrip- Issues in Definition of Asperger Syndrome
tion and proposed some modifications for the
syndrome based on her case studies. Though Although Wing (1981) did not provide a spe-
Asperger thought the condition was unrecog- cific set of categorical diagnostic criteria,
nizable prior to 3 years of age, Wing suggested many case reports and research studies de-
the following difficulties were present in the rived such criteria from her publication and
first 2 years of life: (1) a lack of normal inter- used these to characterize their respective pa-
est and pleasure in other people, (2) babbling tient or subject population. However, accounts
that is limited in quantity and quality, (3) re- often included different sets of criteria, re-
duced sharing of interests and activities, (4) flecting the author’s decision as to which be-
absence of an intense drive to communicate, haviors were necessary, only suggestive, or
both verbally and nonverbally, with others, (5) altogether irrelevant. As noted, such variabil-
speech that is abnormal in terms of delayed ac- ity in diagnostic assignment suggested that
quisition or impoverished content consisting comparisons among studies would be difficult
mainly of stereotyped utterances, and (6) fail- and pointed to the need for a uniform nosology
ure to develop a full repertoire of imaginative of AS (Rutter & Gould, 1985). Despite Wing’s
pretend play. Wing also suggested that AS can (2000) repeated attempts to slow the process
be found in individuals with mild mental retar- that she unwittingly started and echoing Van
dation and that, although in her sample males Krevelen’s (1971) initial approach, the stage
92 Diagnosis and Classification

was set for the field to test the validity status The ICD-10 (WHO, 1993) and DSM-IV def-
of AS relative to HFA. In this context, there initions attempted to bridge over the differ-
was a need for consensual criteria for the defi- ences among previous definitions and to
nition of AS to be universally adopted by clin- contrast AS with autism, without which valida-
ical researchers. Following the framework of tion work would be impossible. To do so, it fo-
current diagnostic systems such as ICD-10 cused primarily on the number of overall
(WHO, 1993) and DSM-IV (1994), this defini- symptoms (greater in autism and lower in AS,
tion had to be categorical (i.e., AS and autism which was intended to capture variable degrees
should be defined in terms of mutually exclu- of severity), excluded the language cluster in-
sive criteria) rather than dimensional (i.e., AS cluded in the diagnosis of autism (indicating
and autism should not differ simply in terms of the absence of severe language impairment in
degree of symptomatology). This task was AS while sidestepping an attempt to define the
complicated by the fact that Wing’s (1981) ac- peculiarities typical of verbal communication
count, from which most subsequent definitions in AS), and specified onset criteria that con-
were derived (Gillberg & Gillberg, 1989; Szat- trasted with those for autism (the absence of
mari, Bartolucci, & Bremner, 1989; Tantam, clinically significant delays in speech, lan-
1988a), implied that the two conditions could guage, and cognitive development in AS). In
not be separated and that AS was a subtype of many respects, this definition was both stricter
autism with only minor differences in clinical than some (particularly relative to the onset
manifestation. Many case reports and research criteria) and less specific than others (as it sub-
studies avoided this issue altogether, particu- sumed the more unique symptoms of AS under
larly by failing to state whether subjects also the autism clusters of social impairment and re-
fulfilled criteria for autism. While the desig- stricted areas of interest). In essence, AS was
nation of Asperger syndrome as a “ variant ” or defined in terms of the criteria for autism
“subtype” of autism (e.g., as highly verbal in- (either presence or absence). The ICD-10 crite-
dividuals with autism with normative or supe- ria for AS are given in Table 4.1.
rior IQs) is justifiable from several practical In practice, the ICD-10 and DSM-IV defin-
perspectives (e.g., to bring attention to a sub- itions make a distinction between autism and
group of people with autism thus facilitating AS solely on the basis of the onset criteria. In
advocacy, to provide a concept around which autism, any concerns prior to the age of 3 years
clients and their families can more easily coa- involving social interaction, social communi-
lesce and share experiences; Wing, 2000; cation, or symbolic/imaginative play are suffi-
Szatmari, 2000), this is not so in formal diag- cient for the criteria to be met. In contrast, any
nostic systems, which are meant, among other concern involving cognitive development (in
things, to provide mutually exclusive cate- essence, typical exploration of and curiosity
gories to facilitate research. Thus, the central about the environment given that the majority
issue in recent years has been whether AS and of children are not developmentally assessed
HFA (and, by necessity, PDD-NOS) are dis- prior to age 3), self-help skills, or more
tinctive in ways that are independent of the broadly defined adaptive behavior (other than
definition used to assign group membership. In social interaction but including social commu-
other words, research has focused on whether nication) would rule out the diagnosis of AS.
these concepts are associated with a different The possibly overinclusive nature of onset cri-
developmental course; different neurocogni- teria for autism and overexclusive nature of
tive, neurobiological, or genetic underpin- onset criteria for AS (and any ambiguities left
nings; or a different outcome (Klin et al., in in the definition, e.g., how to distinguish so-
press; Volkmar, Lord, Bailey, Schultz, & Klin, cial interaction from social communication)
2004). The end result of this process was the are resolved in terms of the precedence rule—
tentative inclusion of AS in ICD-10 (WHO, if an individual meets criteria for autism, he or
1993). And because ICD-10 and DSM-IV were she cannot be assigned the diagnosis of AS
intended to be equivalent, AS was also in- (Volkmar & Klin, 2000).
cluded in the DSM-IV autism / PDD field trials Although the advent of the DSM-IV defi-
(Volkmar, Klin, Seigel, et al., 1994). nition was intended to create a consensual
Asperger Syndrome 93

TABLE 4.1 ICD-10 Research Diagnostic Heavey, & Le Couteur, 1998; Rutter, 1999;
Guidelines for Asperger Syndrome Volkmar, Klin, & Cohen, 1997). For the pur-
1. There is no clinically significant general delay in pose of this chapter, however, this process has
spoken or receptive language or cognitive highlighted the differences among some influ-
development. Diagnosis requires that single words ential definitions of AS. Table 4.2 presents a
should have developed by 2 years of age or earlier comparison of key diagnostic features among
and that communicative phrases be used by 3 years the most widely used definitions, which in-
of age or earlier. Self-help skills, adaptive
behavior, and curiosity about the environment
clude Asperger’s (1944) original account and
during the first 3 years should be at a level later emphases and changes (Asperger, 1979;
consistent with normal intellectual development. Van Krevelen, 1971), as well as those of Wing
However, motor milestones may be somewhat (1981), Gillberg and Gillberg (1989), Tantam
delayed and motor clumsiness is usual (although (1988a), and Szatmari, Bremner, and Nagy
not a necessary diagnostic feature). Isolated (1989a, 1989b). These definitions are com-
special skills, often related to abnormal
pared with the one formalized in DSM-IV
preoccupations, are common, but are not required
for the diagnosis.
(American Psychiatric Association, 1994),
2. There are qualitative abnormalities in reciprocal which, as noted, is conceptually equivalent to
social interaction (criteria for autism). the one included in ICD-10 (WHO, 1993;
3. The individual exhibits an unusually intense, Volkmar et al., 1994). Items in bold are those
circumscribed interest or restricted, repetitive, features deemed necessary for the diagnosis of
and stereotyped patterns of behavior interests and AS according to the given diagnostic system.
activities (criteria for autism; however, it would
It is evident from Table 4.2 that a direct com-
be less usual for these to include either motor
mannerisms or preoccupations with part-objects parison among the various diagnostic systems is
or nonfunctional elements of play materials). not straightforward. Ghaziuddin and colleagues
4. The disorder is not attributable to other varieties (1992) attempted to compare some of these sys-
of pervasive developmental disorder; simple tems. Having simplified the criteria for compar-
schizophrenia schizotypal disorder, obsessive- ison, the study compared Asperger’s (1944)
compulsive disorder, anakastic personality definition with Wing’s (1981)—which was
disorder; reactive and disinhibited attachment
disorders of childhood.
made to be equivalent to Gillberg and Gillberg’s
(1989) and Tantam’s (1988)—and the defini-
Source: Reprinted with permission from World Health tions of Szatmari and colleagues’ (1989a,
Organization (1993). Disorders of psychological de-
1989b) and ICD-10 (WHO, 1993). Of 15 pa-
velopment (Criteria for Research) (Geneva: WHO),
pp. 154–155. tients identified as having AS according to
Wing’s criteria, only 10 patients met Szatmari
diagnostic starting point for research, it has et al.’s criteria, and only 8 patients met the de-
been consistently criticized as overly narrow scription by Asperger as well as ICD-10 crite-
(Eisenmajer et al., 1996; Szatmari, Archer, ria. The primary reason that seven patients did
Fisman, Streiner, & Wilson, 1995), rendering not fulfill ICD-10 (and Asperger’s) criteria was
the diagnostic assignment of AS improbable or their failure to meet the onset criteria, specifi-
even “ virtually impossible” according to some cally the lack of clinically significant delay
authors (Leekam, Libby, Wing, Gould, & Gill- in speech and language acquisition. Therefore,
berg, 2000; Mayes, Calhoun, & Crites, 2001; despite the lack of required specific social, lan-
Miller & Ozonoff, 2000). While this critique guage and communication, and absorbing inter-
has reinforced the approach of some re- ests criteria (which differ form Asperger’s
searchers to cluster all ASDs together in account), the ICD-10 definition is, in practice,
search of common underlying factors, other re- more restrictive than other systems because of
searchers are pursuing subgrouping efforts be- its emphasis on specific onset patterns (which is
cause of their recognition that autism is a consistent with Asperger’s account).
clinically heterogeneous disorder and that the This state of affairs raises several issues for
characterization of subtypes of PDD might nosology research:
help behavioral and biological research by
allowing the identification of clinically more 1. By simultaneously emphasizing the lack of
homogeneous groups (Bailey, Palferman, communication and cognitive delays in the
94 Diagnosis and Classification

TABLE 4.2 Comparison of Six Sets of Clinical Criteria Defining Asperger Syndrome
Gillberg & Szatmari
Asperger Wing Gillberg Tantam et al. DSM-IV
Clinical Feature* (1944, 1979) (1981) (1989) (1988) (1989) (APA, 1994)

Social Impairment
Poor nonverbal
communication Yes Yes Yes Yes Yes Yes
Poor empathy Yes Yes Yes Yes Yes Yes
Failure to develop
friendship Yes Yes Yes Yes Yes Yes
Poor prosody and
pragmatics Yes Yes Yes Yes Yes Not stated
Idiosyncratic language Yes Yes Not stated Not stated Yes Not stated
Impoverished imagi-
native play Yes Yes Not stated Not stated Not stated Not stated
All Absorbing Interest Yes Yes Yes Yes Not stated Often
Motor Clumsiness Yes Yes Yes Yes Not stated Often
Onset (0–3 years)
Speech delays/deviance No May be present May be present Not stated Not stated No
Cognitive delays No May be present Not stated Not stated Not stated No
Motor delays Yes Sometimes Not stated Not stated Not stated May be present
Exclusion of Autism Yes (1979) No No No Yes Yes
Mental Retardation No May be present Not stated Not stated Not stated Not stated

* Symptoms that are defined as necessary for the presence of the condition are given in bold.

first years of life while failing to specify re- with Asperger’s description. Also, if pursu-
quired social, communication, motor, and ing this strategy, it is unlikely that such
absorbing interest features that are thought poorly detailed, “all or nothing” description
to be typical in AS, the ICD-10 definition of onset characteristics will capture mean-
differentiates autism from AS solely based ingful variability of developmental concepts
on the onset criteria, in fact, irrespective of of interest (e.g., social motivation, joint at-
the nature of the patient’s social impairment tention, nonverbal communication), which
later in life. Whether individuals diagnosed are likely to mediate outcome.
in this manner will have a later presentation 2. By failing to include social, communicative,
consistent with Asperger’s description re- or restricted interest patterns more specific
mains to be documented. Various authors to AS ( by subsuming those under general
(Gillberg & Gillberg, 1989; Leekam, Libby, clusters defining autism), the ICD-10 defi-
Wing, Gould, & Gillberg, 2000; Wing, nition of AS disregards features that could
1981) have reported cases that would not serve as discriminative factors and may be
meet the ICD-10 onset criteria and yet pre- unique among AS patients. For example,
sented with what they thought was AS. Tantam (1988a) suggests that in contrast
However, even higher functioning individu- to the prevailing view of individuals with
als with autism appear to have the onset of autism, individuals with AS may wish to
their condition before age 3 years (Volkmar be sociable and yet fail to establish rela-
& Cohen, 1989). In summary, while it is in- tionships. Although this may also be true
teresting to study eventual clinical presenta- for older adolescents and adults with
tions on the basis of early profiles of autism, it would be unusual for a younger
development, there is no assurance that later child with autism. This was an important
onset will lead to individuals conforming point raised by Van Krevelen (1971), who
Asperger Syndrome 95

thought children with autism disregarded dimensions creating the autism spectrum
others while children with AS approached (e.g., Baron-Cohen, 2002; Constantino &
others in an eccentric fashion. Both were so- Todd, 2003). The second approach—here
cially impaired, but qualitatively different. called the early language approach, makes a
3. Just as autistic social dysfunction can be distinction between AS and HFA by dividing
defined only in the context of the child’s children with ASD and apparently normal cog-
overall developmental level (Rutter, 1978), nitive development into two groups in terms of
there may be a host of developmental fac- language development in the first 3 years of
tors that should be considered in attempts life. A diagnosis of AS is given if the child
to differentiate the social, affective, and achieved single words by age 2 years and
communication presentation of individuals phrase speech (typically defined as non-
with HFA from those with AS. These in- echolalic three-word combinations used mean-
clude chronological age (e.g., less pro- ingfully for communication) by the age of 3
nounced contrast in older adolescents and years. A diagnosis of autism is given if the
adults), IQ and language level (e.g., less child does not meet these criteria (Gilchrist
pronounced contrast in individuals with et al., 2001; Szatmari et al., 1995, 2000). The
higher IQ and language abilities), and the third approach criticizes the first two and of-
presence of any medical condition disrupt- fers an alternative (Klin et al., in press). As to
ing speech and language acquisition early the spectrum approach, it is unlikely, though
in life independently from psychiatric di- not entirely impossible, that single underlying
agnosis (e.g., cleft palate, intermittent factors or dimensions could generate the
hearing loss resulting from chronic ear highly complex and heterogeneous spectrum
infection). of autism-related disorders in a predictable,
quantified manner. As to the early language
In summary, the different diagnostic sys- approach, there is concern that it greatly nar-
tems have provided different sets of criteria rows the potential lines of distinction between
but have not resolved the key issue dominating autism and AS in that other aspects of onset
research on AS since the early 1970s, namely, (e.g., social engagement and social cognitive
its validity status relative to autism. That some patterns), as well as any unique features in
influential authors had not meant to suggest current presentation (e.g., social motivation,
that AS was a distinguishable diagnostic entity verbosity), are disregarded (Volkmar & Klin,
is unlikely to quell this debate. There is a need 2000). More practically, there is concern that
to go beyond the current sterile impasse fuel- because individuals with HFA may not present
ing so much of the current classification de- with speech delays as defined in this diagnos-
bate, which is still based on rather arbitrary tic scheme, there is a potential for the resultant
decisions and vague semantics rather than samples (of individuals with HFA and AS) to
empirical validation of operationalized defini- overlap considerably in terms of other sympto-
tions that reflect more sophisticated develop- matology, thus increasing the potential for
mental constructs. type II errors (i.e., not finding differences).
Three approaches appear to be emerging To address these potential limitations, the
from this discussion. The first, here called the third approach—here called the unique fea-
spectrum approach, disregards the nosologic tures approach (possibly better described as
discussion entirely by equating AS with “ bell-ringers” approach)—emphasizes the fea-
higher functioning ASDs or with some social tures more specifically associated with AS
vulnerability of the kind seen in ASDs (some highlighted by Asperger (1944) and several
individuals with this characterization may not other authors (e.g., Klin & Volkmar, 1997;
meet clinical criteria for a PDD). Proponents Tantam, 1988a). These features are included in
of this approach make the assumptions that all the narratives accompanying the definition of
early-onset and chronic social vulnerabilities AS in ICD-10 (WHO, 1993) and in DSM-IV-TR
share some factors in common and that re- (American Psychiatric Association, 2000) but
search is best invested in better examining and are not included in the respective sets of diag-
possibly quantifying these factors seen as the nostic criteria. In this approach, onset criteria
96 Diagnosis and Classification

are given in more detail. For example, distinc- and 1980. Only an estimated 1.15% was diag-
tions are made between children who isolate nosed with AP. Thus, the condition that he de-
themselves (more typical of autism) and those scribed was likely to be quite specific. The
who seek others, sometimes incessantly, but in estimate of cases falling within the autism
a socially insensitive manner (more typical of spectrum, but without an explicit diagnosis of
AS); and between children whose language is AP, was double that number. Of the AP cases,
delayed, echolalic, or otherwise stereotyped 95% were boys, a manifold increase in the typ-
(more typical of autism) and those whose lan- ical autism gender ratio (4 boys for every girl),
guage is adequate or even precocious and larger but closer to the ratio reported for
whose difficulties in this area are limited to higher functioning ASDs, and closer still to
the communicative use of language (i.e., prag- the gender ratio reported for individuals with
matics; more typical of AS). The added details AS (Volkmar et al., 2004). The clinical con-
are meant to facilitate research into develop- cepts used to describe patients suggested the
mental paths of social disabilities with a notion that these children had social motiva-
greater degree of specificity. Additional mod- tion but were very awkward in approaching
ifications proposed involve the inclusion of others, leading to failure to establish relation-
one-sided verbosity as a necessary communi- ships. For example, “contact disorder ” or “in-
cation criterion in AS and the presence of fac- stinct disorder ” were used to refer to a lack of
tual, circumscribed interests that interfere common sense, impaired “practical intelli-
with both general learning and reciprocal gence” in everyday situations, and a need to
social conversation. These communication learn these skills through their intellect (Hip-
symptoms were introduced with a view to cap- pler & Klicpera, 2003, p. 294). Fifty-four per-
ture the observed greater social motivation cent of cases showed excellent verbal abilities
seen in individuals with AS (relative to HFA). (e.g., fund of information) but impaired non-
Research systematically assessing the util- verbal abilities (e.g., visual-spatial skills),
ity of these approaches is still limited. One re- echoing a learning profile thought to be more
cent study (Klin et al., in press), however, has typical of individuals with AS than of those
shown that three diagnostic schemes for AS with HFA (Klin, Volkmar, Sparrow, Cicchetti,
used simultaneously—DSM-IV, the early & Rourke, 1995). On IQ tests, higher verbal IQ
language approach, and the unique features (VIQ) than performance IQ (PIQ) was more
approach—have low agreement in case assign- than twice as common as the reverse (44%
ment and lead to different results in compar- versus 18%), and almost the totality of cases
isons of IQ profiles, patterns of comorbidity, had full-scale IQs within or above the norma-
and familial aggregation of psychiatric symp- tive range. Some resemblance between the
toms across the approach-specific resultant child with AP and one or more family mem-
groups of HFA, AS, and PDD-NOS. bers was observed in 53% of the sample, with
52% of fathers being reported as being simi-
ASPERGER’S WORK REVISITED lar to their child in aspects of social dysfunc-
tion, a much higher estimate of familiality
Hippler and Klicpera (2003) recently con- than observed in autism but similar to some
ducted a retrospective analysis of the clinical recent family genetic data using the more
case records of individuals with autistic psy- specific feature of AS to define the proband
chopathy (AP) diagnosed by Hans Asperger (Klin et al., in press). Eighty-two percent of
and his team in the various clinics that they AP cases were reported to have circum-
practiced. Although there are obvious limita- scribed interests of the kinds subsequently
tions in using current concepts to reanalyze described as particularly specific to AS (e.g.,
data generated through the lens of a different topic-based obsessions). An attempt to apply
era (with its unique system of diagnostic con- the ICD-10 criteria for AS resulted in a diag-
cepts and debates), this paper helps us better nosis of AS in 68% of cases; 25% did not meet
understand Asperger’s work and development the requirement of absence of clinical symp-
of his diagnostic concept. For example, he saw toms before the age of 3; of these, half had
approximately 9,800 children between 1951 some language delays.
Asperger Syndrome 97

Hippler and Klicpera’s article (2003) is of difference (e.g., level and pervasiveness of
very helpful in showing the very specific pro- social disability, more severe in AS), outcome
file of children singled out by Asperger over ( less positive in AS), and relatedness to the
his lifetime, which contrasts markedly with schizophrenia spectrum of disorders (the asso-
some of the current definitions of AS. Still, ciation is stronger in schizoid personality in
there is little to gain from an exegetic analysis childhood).
of this nature, if the validity and utility of the The concept of nonverbal learning disabili-
construct cannot be assessed relative to pre- ties (NLD) was originally proposed by John-
dictions independent of definitions for case son and Myklebust (1971; Myklebust, 1975)
assignment. The development of our under- and was subsequently elaborated and thor-
standing of autism offers an interesting con- oughly researched by Rourke (1989). NLD
trast. Despite the remarkably enduring quality refers to a profile of neuropsychological assets
of Kanner’s (1943) original description, many and deficits that have a significant negative
key issues have been modified as a result of impact on a person’s social and communica-
subsequent research. The fact that Rutter’s tion skills (see Chapter 13, this Handbook, this
(1978) codification of Kanner’s prose proved volume). Deficits in neuropsychological skills
to be an effective and reliable diagnostic tool such as tactile perception, psychomotor coor-
is partially thanks to Kanner’s brilliance, but dination, visual-spatial organization, and non-
it is also a function of the severity of the con- verbal problem solving occur in the presence
dition he described. Given that the syndrome of preserved rote verbal abilities. The particu-
described by Asperger is more equivocal than lar profile of strengths and deficits results in a
Kanner’s autism—overlapping with autism to characteristic learning style such as the ten-
some extent but also maybe shading into ec- dency to overly rely on overlearned behaviors
centric normalcy, it is appropriate that addi- when dealing with novel or complex situations.
tional work is needed to create an effective Poor pragmatics and prosody in speech are
and reliable definition. seen in the presence of relatively preserved
formal language skills such as vocabulary and
ALTERNATIVE DIAGNOSTIC syntax and single-word reading abilities. Dif-
CONCEPTS ficulties in appreciating the subtle, and some-
times obvious, nonverbal aspects of social
Professionals from diverse disciplines such as interaction lead to major deficits in social per-
adult psychiatry, neuropsychology, and neurol- ception and judgment, which often result in
ogy have dealt with individuals with signifi- social isolation and rejection and increased
cant problems in social interaction who did not risk for social withdrawal and serious mood
seem to precisely fit Kanner’s (1943) concept disorders (Rourke, Young, & Leenaars, 1989).
of infantile autism (Klin & Volkmar, 1997). Although potentially associated with a number
Wolff and colleagues (Wolff & Barlow, 1979; of different conditions (Rourke, 1995), it is
Wolff & Chick, 1980) proposed the term often seen in individuals with AS but less so in
schizoid personality in childhood in their de- individuals with autism (Gunter, Ghaziuddin,
scription of children with social isolation and & Ellis, 2002; Klin, Sparrow, Volkmar, Cic-
emotional detachment, unusual communicative chetti, & Rourke, 1995; Lincoln, Courchesne,
style, and rigidity of thought and behavior. Al- Allen, Hanson, & Ene, 1998; Siegel, Minshew,
though the initial emphasis on the condition as a & Goldstein, 1996).
personality disorder resulted in more sketchy Another influential concept was proposed
accounts of developmental course, attempts to by Rapin and Allen (1983), who used the term
reconstruct the developmental history of these semantic-pragmatic disorder to describe cases
children were made (e.g., Wolff, 1991, 1995). in which speech and language skills were ade-
Several attempts have been made to compare quate in form (syntax and phonology) but im-
this concept with AS (e.g., Nagy & Szatmari, poverished in content and use (semantics and
1986; Tantam, 1988b), showing some areas of pragmatics). Bishop (1989, 1998) refined the
similarity (e.g., abnormalities in empathy and descriptions and assessment instrumentation
nonverbal communication) but also some areas used in the characterization of communicative
98 Diagnosis and Classification

difficulties exhibited by these children. More child began to talk before learning to walk.
recent research (Bishop, 2000) has shown that Vocabulary acquisition, however, may be un-
the association between semantic and prag- usual as children may learn complex or adult-
matic deficits is not a necessary one and that a like words, typically associated with a special
term such as pragmatic language impairment interest prior to learning more typical, child-
might be preferable to refer to these children. like vocabulary associated with social play
Other concepts, such as developmental and experiences. Parents may report that once
learning disability of the right hemisphere the child began to talk, there was a pedantic
(Denckla, 1983; Weintraub & Mesulam, quality to their speech, both in terms of their
1983), have been proposed. Although the choice of words and sentence construction
number of labels available and the diversity of (more formal than seen in typical peers) and
disciplines from which they arose are a testa- in terms of the tone of voice and phrasing
ment to the robustness of the underlying clini- (sometimes assuming a “ teaching” quality).
cal phenomena, these various labels have the Attachment patterns to family members are
potential for introducing confusion in noso- often seen as unremarkable, and in contrast to
logic discussion of AS. Their discipline- autism, there are few visible signs of social
specific nature may also inhibit the kinds of disability in the highly familiar environment of
interdisciplinary research that could provide the child’s home. In many ways, these children
an integrated and comprehensive understand- appear to orient to others although they may
ing of individuals with AS. Confusion is par- use them more instrumentally than recipro-
ticularly an issue for parents whose child cally, for example, to speak to them rather than
might receive any of several different diagnos- to truly engage them in a shared pursuit. The
tic labels depending on the training and disci- social disability becomes more apparent when
pline of the clinician with whom they consult. the child is outside the home environment, par-
Given that the various conditions described ticularly in group situations involving same-
reflect primarily differences in discipline- age children. In such situations, they may
specific emphasis ( language and communica- approach other children in inappropriate or
tion versus neuropsychological profiles versus awkward ways, for example, speaking loudly
neurological profiles), thus mapping on differ- when in close proximity, or otherwise becom-
ent levels of discourse, it is clear that these are ing very upset when other children are not
not mutually exclusive concepts. In other willing to play with them following their
words, they describe an overlapping group of agenda, which typically involves the pursuit of
children with social vulnerabilities. Therefore, play or games restricted to a narrow and often
it is likely more helpful if they are not used as developmentally atypical interest. The con-
competing diagnostic concepts but as method- trast in presentation in the home and family
ological approaches to better understand setting relative to social group settings is
aspects of the social, communicative, neuro- likely explained by the nature of the social
psychological, and neurologically based dis- partners. At home, interactions with parents
abilities evidenced in individuals with AS. are highly scaffolded by the adults, who mold
their behavior to adjust to the child’s ap-
CLINICAL FEATURES OF ASPERGER proaches, guiding the interaction and prevent-
SYNDROME ing communication breakdown. In social group
settings involving peers, there is a need for
Onset Patterns more conventional and socially appropriate be-
havior because peers show little tolerance for
Children with AS do not present with clini- deviation from usual expectations. It is typi-
cally significant delays in language acquisi- cal, therefore, for parents to first become con-
tion, cognitive development, or self-help cerned about their children’s behavior at the
skills in the first years of life. In fact, lan- time of transitions to school-like settings (e.g.,
guage acquisition, in terms of vocabulary and informal play groups, nursery, or preschool).
sentence construction, may be precocious in During preschool years, some children may
some cases and parents may report that the develop intense interests about which they
Asperger Syndrome 99

begin to learn a great deal. Some of these may and implied communications (e.g., signs of
be unusual for their age (e.g., geography, ar- boredom, haste to leave, and need for privacy).
rows, storm drains). They may acquire a Chronically frustrated by their repeated experi-
wealth of facts and information related to a ences of failure to engage others and form
given interest, choosing to talk a great deal to friendships, some individuals with AS develop
other children, who fail to show the same in- symptoms of a mood disorder that may require
terest and, consequently, distance themselves treatment, including medication. They also may
from such approaches. Typical experiences an- react inappropriately to, or fail to interpret the
choring a child’s understanding of his or her valence of, the context of the affective inter-
day (e.g., routines at preschool, transition action, often conveying a sense of insensitivity,
times, social rituals) may not convey a sense of formality, or disregard for the other person’s
familiarity, prompting the child to rely (often emotional expressions. They may be able to de-
precociously) on more formal definitions of scribe correctly, in a cognitive and often for-
experiences such as time (e.g., a 3-year-old malistic fashion, other people’s emotions,
may repeatedly ask, “What time is it?”), rigid expected intentions, and social conventions; yet,
rules (an explicitly and verbally defined they are unable to act on this knowledge in an
schedule), or repeated questions of adults. To intuitive and spontaneous fashion, thus losing
some children, these behaviors appear to be re- the tempo of the interaction. Their poor intu-
lated to incapacity to make intuitive sense of ition and lack of spontaneous adaptation are ac-
their own experiences, of nonexplicit nonver- companied by marked reliance on formalistic
bal cues conveyed by others, and of the rules of behavior and rigid social conventions.
changeable behavior of people who make sense This presentation is largely responsible for the
only within the overall context of a social situ- impression of social naiveté and behavioral
ation, yet another construct adding to the expe- rigidity that is so forcefully conveyed by these
rience of confusion in these children. The individuals (Klin & Volkmar, 1997).
pursuit of unchangeable phenomena (that re-
main the same and do not depend on context) Communication Patterns
may have earlier led the child to exhibit a fas-
cination with letters and numbers, which sub- Although significant abnormalities of speech
sequently may evolve into precocious reading and language are not typical of individuals
abilities and fascination with facts, all of with AS, there are at least three aspects of
which may serve to ground the child in what these individuals’ communication patterns that
otherwise would be an overwhelmingly per- are of clinical interest (Klin, 1994). First,
plexing, constantly shifting, and primarily im- speech may be marked by poor prosody, al-
plicit social environment. though inflection and intonation may not be as
rigid and monotonic as in autism (Ghaziuddin
Social Functioning & Gerstein, 1996). They often exhibit a con-
stricted range of intonation patterns that is
In some contrast to the social presentation in used with little regard to the communicative
autism, individuals with AS find themselves so- functioning of the utterance (e.g., assertions
cially isolated but are not usually withdrawn in of fact, humorous remarks). Rate of speech
the presence of other people, typically ap- may be unusual (e.g., too fast) or may lack in
proaching others but in an inappropriate or ec- fluency (e.g., jerky speech), and there is often
centric fashion. For example, they may engage poor volume modulation (e.g., voice is too loud
the interlocutor, usually an adult, in one-sided despite physical proximity to the conversa-
conversation characterized by long-winded, tional partner). The latter feature may be par-
pedantic speech about a favorite and often un- ticularly noticeable in the context of a lack of
usual and narrow topic. They may express inter- adjustment to the given social setting (e.g., in
est in friendships and in meeting people, but a library or a noisy crowd). Second, speech
their wishes are often thwarted by their awk- may often be tangential and circumstantial,
ward approaches and insensitivity to the other conveying a sense of looseness of associations
person’s feelings, intentions, and nonliteral and incoherence. Even though in a very small
100 Diagnosis and Classification

number of cases this symptom may be an indi- Motoric Difficulties

cator of a possible thought disorder, the lack of
contingency in speech is a result of the one- Individuals with AS often present with a his-
sided, egocentric conversational style (e.g., tory of motor difficulties and delayed acquisi-
unrelenting monologues about the names, tion of skills requiring sophisticated motor
codes, and attributes of innumerable TV sta- coordination, such as pedaling a bike, catching
tions in the country), failure to provide the a ball, opening jars, and climbing outdoor play
background for comments and to clearly de- equipment. They are often visibly awkward
marcate changes in topic, and failure to sup- and poorly coordinated and may exhibit un-
press the vocal output accompanying internal usual gait patterns, odd posture, poor manipu-
thoughts. Third, the communication style of lative and handwriting skills, and significant
individuals with AS is often characterized by deficits in visual-motor skills (Gillberg, 1990;
marked verbosity. The child may talk inces- Tantam, 1988a). When tested, children with
santly, usually about a favorite subject, often AS have been shown to experience difficulties
in complete disregard to whether the listener on measures of apraxia, balance, tandem gait,
might be interested, engaged, or attempting to and finger-thumb apposition, suggesting diffi-
interject a comment, or change the subject of culties in proprioception (Weimer, Schatz,
conversation. Despite such long-winded mono- Lincoln, Ballantyne, & Trauner, 2001). It is not
logues, the individual may never come to a clear, however, whether “clumsiness” or poor
point or conclusion. Attempts by the interlocu- motor coordination skills differentiate indi-
tor to elaborate on issues of content or logic or viduals with AS from those with HFA (Green,
to shift the interchange to related topics are Gilchrist, Burton, & Cox, 2002; Smith, 2000),
often unsuccessful. who may show similar difficulties.

Circumscribed Interests Comorbid Features

Individuals with AS typically amass a large From a clinical standpoint, the most common
amount of factual information about a topic in comorbid conditions impacting on individuals
a very intense fashion. The actual topic may with AS are anxiety and depression (Klin &
change from time to time but often dominates Volkmar, 1997). In many cases, anxiety seems
the content of social exchange. Frequently, the to be secondary to a sense of being over-
entire family may be immersed in the subject whelmed by the fast pace and competitive so-
for long periods of time. This behavior is pecu- cial demands of typical peer interactions
liar in the sense that oftentimes extraordinary coupled with a sense of lack of control over the
amounts of factual information are learned outcome of such social events. Similarly, de-
about very circumscribed topics (e.g., snakes, pression appears to emerge as a result of in-
names of stars, TV guides, deep fat fryers, creased awareness of repeated experiences of
weather information, personal information on failure despite an intent, which can be quite in-
members of Congress) without a genuine un- tense, to establish relationships, make friends,
derstanding of the broader phenomena in- or have a romantic experience. Research esti-
volved. This symptom may not always be easily mates of comorbid anxiety and/or depression
recognized in childhood because strong inter- in individuals with AS are as high as 65%
ests in certain topics, such as dinosaurs or (Ellis, Ellis, Fraser, & Deb, 1994; Fujikawa,
fashionable fictional characters, are so ubiqui- Kobayashi, Koga, & Murata, 1987; Ghaziud-
tous. However, in both younger and older chil- din, 2002; Ghaziuddin, Ghaziuddin, & Greden,
dren the special interests typically interfere 2002; Ghaziuddin, Weidmer-Mikhail, & Ghaz-
with learning in general because they absorb iuddin, 1998; Green et al., 2000; Howlin &
so much of the child’s attention and motiva- Goode, 1998). Some data suggest, however,
tion, and they have a very negative impact on that individuals with AS and HFA are equally
their ability to engage others in reciprocal so- at increased risk for problems with anxiety
cial interaction because the interest intrudes and depression, with no differences between
and often dominates conversation with others. the groups (Kim, Szatmari, Bryson, Streiner,
Asperger Syndrome 101

& Wilson, 2000). In some respects, these re- may lead them, for example, to make blunt re-
sults are difficult to reconcile with our experi- quests of a sexual nature; or their intense and
ence, according to which individuals with all-absorbing circumscribed interests may
autism are likely to suffer less anxiety and de- lead them to commit eccentric acts associated
pression because of their lack of involvement with those interests (e.g., accumulate clutter
with others, which in turn, makes them (1) less in the backyard of a multifamily home). More
vulnerable to the challenges inherent in trying typically, however, these individuals are too
to navigate a complex social world and (2) less naïve to become “competent criminals.” In
vulnerable to experiences of inadequacy re- fact, individuals with AS are much more likely
sulting from failure to establish relationships to be victims than victimizers due to their in-
despite repeated (and painful) attempts. ability to read the intents of others, which in a
Several other conditions have been described typical school environment may involve re-
in association with AS, although these descrip- peated mocking and cruel singling out of the
tions are more typical of early published reports child because of his or her eccentricities, poor
that tended to be case studies rather than case- grooming, or social clumsiness. Adults may
control series. These early reports emphasized eventually grow despondent in social situa-
possible associations with Tourette’s syndrome tions, eventually gravitating toward the pe-
(e.g., Gillberg & Rastam, 1992; Kerbeshian & riphery of society.
Burd, 1986; Littlejohns, Clarke, & Corbett, Another association suggested by early re-
1990; Marriage, Miles, Stokes, & Davey, 1993) ports was the presence of schizophrenia in
and obsessive-compulsive disorder (Thomsen, some individuals with AS. In fact, there was
1994). Reports of associations with psychotic some suggestion that AS might, in some sense,
conditions have included reports of psychotic be a “ bridging” condition between autism and
depression and bipolar disorder (manic depres- schizophrenia (Wolff, 1995). Though several
sive psychosis; Gillberg, 1985). research studies have suggested an association
Some reports, though primarily case stud- between the disorders (e.g., Nagy & Szatmari,
ies, have focused on conduct problems, more 1986; Tantam, 1991), the preponderance of
specifically on violent and criminal behavior empirical evidence fails to support this spec-
(e.g., Baron-Cohen, 1988; Everall & Le Cou- ulation. Ghaziuddin, Leininger, and Tsai
teur, 1990; Mawson, Grounds, & Tantam, (1995) found that patients with AS had
1985; Scragg & Shah, 1994; Tantam, 1988c; greater levels of disorganized thinking than an
Wing, 1986). However, the hypothesis that the HFA comparison group, but they were not
combination of high intelligence and verbal more likely to exhibit a thought disorder. Re-
skills with poor empathy and social cognition views of large case series reveal that schizo-
fosters violent or criminal behavior has not phrenia is no more common among individuals
been empirically borne out. Ghaziuddin and with AS than in the general population (Volk-
colleagues (1991) reviewed the literature and mar & Cohen, 1991). And in family genetic
found a lack of support for this speculation. studies, there is no evidence of larger aggrega-
Our own experience suggests that individuals tion of schizophrenia-related symptoms in
with AS do not present with antisocial or so- families of probands with ASD or of ASD-re-
ciopathic characteristics; the absence of em- lated symptoms in families of probands with
pathy connotes poor insight into the social and schizophrenia (see Chapter 16, Handbook,
emotional nature of other people, not an ab- this volume). Though research has not sup-
sence of compassion for their welfare. ported a relationship between AS and schizo-
Individuals with AS often transgress rules at phrenia, several factors have contributed to
school, with people, and in the community at this enduring misperception. The combina-
large, and their behavior may indeed lead to tion of excessive verbalization and poor social
formal encounters with school authorities or judgment and social monitoring in AS or the
law enforcement officers. However, they tendency of persons with AS to speak exces-
typically do not engage in these acts willfully sively about special interests may result in
or maliciously. Their social ineptitude and highly inappropriate and bizarre behavior that
unawareness of social rules and expectations mimics psychotic behavior (e.g., incoherent or
102 Diagnosis and Classification

accelerated speech). Additionally, there has working prevalence rate of 2 per 10,000 pend-
been an overreliance on single-case studies as ing further research. Prevalence estimates
well as some ascertainment biases (Bejerot & will always depend on the definition of AS
Duvner, 1995; Ryan, 1992; Taiminen, 1994). and the adoption of operationalized concepts
that can be applied by epidemiologists with a
PREVALENCE degree of reliability. Hence there is a need for
progress in defining consensual approaches to
Studies of prevalence in general, and of psy- this condition because, from a purely practi-
chiatric disorders in particular, are greatly cal standpoint, prevalence estimates have
affected by the quality and nature of the defi- profound implications for allocation of re-
nition used to single out cases from the gen- sources and service provision.
eral population (e.g., precise versus vague,
discrete versus overlapping with other condi- VALIDITY OF ASPERGER SYNDROME
tions). As discussed earlier, varying criteria
continue to be employed in the diagnosis of As amply documented in the previous sections,
AS, constituting a major confound in any at- the validity of AS as distinct from other condi-
tempt to interpret epidemiological studies of tions, notably HFA, but also PDD-NOS, re-
AS (Ehlers & Gillberg, 1993). Nevertheless, mains controversial (Ozonoff & Griffith, 2000;
consistent with the trend observed in ASDs in Volkmar & Klin, 2000). A lack of uniformity in
general (Fombonne & Tidmarsh, 2003) rates usage of the term also characterizes definitions
of AS appear to be on the rise. Twenty years adopted for the purpose of research, as differ-
ago, an epidemiological study in an area of ent sets of diagnostic criteria have been used by
London (Wing, 1981; Wing & Gould, 1979) different researchers with resulting complica-
reported a prevalence rate of 0.6 to 1.1 per tions for interpretation of research (Ghaziuddin
10,000. However, this study assessed only et al., 1992; Klin et al., in press). There is little
children under the age of 15 with mild mental disagreement about the fact that AS is on a phe-
retardation, clearly a minority among indi- nomenological continuum with autism, particu-
viduals assigned the label AS, thereby very larly in relation to the problems in the areas of
likely underestimating the true prevalence. social and communication functioning (Wing,
Gillberg and Gillberg (1989) reported rates of 1991). For example, within the DSM-III-R
10 to 26 children per 10,000 among children (American Psychiatric Association, 1987) di-
with normal intelligence, with an additional agnostic system, persons with AS would either
0.4 per 10,000 showing the combination of meet criteria for autistic disorder or would be
AS and mental retardation. A subsequent said to exhibit PDD-NOS (Tsai, 1992). What is
study (Ehlers & Gillberg, 1993) calculated a less clear is whether the condition is qualita-
minimum prevalence rate of 3.6 per 1,000 tively different from, rather than just a milder
children (7 to 16 years of age) for a diagnosis form of, autism.
of AS and 7.1 per 1,000 children when includ- Several studies investigating different as-
ing suspected cases. Fombonne and Tidmarsh pects of the disorder have attempted to
(2003) reviewed a number of epidemiological identify discriminating criteria between the
studies that included the diagnostic category two conditions with only mixed results to
of AS and reported a wide range of prevalence date. These studies have involved the neu-
rates, from 0.3 to 48.4 per 10,000, which ropsychological, social-cognitive, neurobio-
clearly reflected methodologic differences logical, genetic, and prognostic aspects of the
across studies. Illustrating a common problem disorders.
in epidemiological research, the highest esti-
mate derived from the study with the smallest Neuropsychological Studies
sample size. Making allowances for the
methodological shortcomings of the available A number of studies have attempted to dis-
studies and for the great ambiguity associ- tinguish AS from other ASDs, and HFA in
ated with the definition of the phenomenon it- particular, by demonstrating a pattern of neu-
self, Fombonne and Tidmarsh suggested a ropsychological performance that uniquely
Asperger Syndrome 103

characterizes the disorders (e.g., Wing, 1998). Coding Subtests). In contrast, individuals with
Though findings are not wholly consistent, autism displayed an isolated skill on a measure
these studies suggest that individuals with of visual parts-to-whole reasoning (Block De-
HFA tend to have greater impairment within sign subtest). Lincoln and colleagues (1998)
the domains of language and verbal compre- carried out a meta-analysis of several studies
hension (Lincoln, Allen, & Kilman, 1995; addressing neuropsychological profiles differ-
Siegel et al., 1996) while exhibiting relative entiating AS from HFA. Results indicated that
strength in nonverbal areas (Klin, Carter, & individuals with AS showed a pattern of
Sparrow, 1997). Several studies have empiri- stronger VIQ relative to PIQ. In contrast, indi-
cally demonstrated significant distinctions be- viduals with HFA exhibited the reverse pat-
tween groups. In one study (Klin, Volkmar, tern, with stronger nonverbal abilities and
Sparrow, Cicchetti, & Rourke, 1995), a group weaker verbal skills. The authors concluded
of individuals with AS and HFA of compara- that autism is characterized by impaired ver-
ble chronological age and full-scale IQ were bal skills, which are intact in AS. Miller and
compared in multiple domains of neuropsy- Ozonoff (2000) also compared groups with
chological function. On average, individuals HFA and AS with similar results but with an
with AS exhibited a significant VIQ-PIQ dif- important caveat. The AS group had higher
ferential with stronger verbal abilities, while VIQ and full-scale IQ and exhibited a greater
individuals in the HFA group tended to have difference between verbal and nonverbal abili-
comparable verbal and performance abilities. ties (with verbal abilities being stronger).
With respect to other domains of neuropsy- However, this study also observed better vi-
chological function, 11 areas were shown to sual-perceptual skills in individuals with AS,
discriminate between the two groups. Some suggesting a more general effect of higher per-
neuropsychological skills represented areas of formance or IQ in the group with AS rather
strength in AS and weakness in HFA, whereas than reverse profiles, suggesting that “AS is
the converse was true for other domains of simply high-IQ autism.” Evidence was sugges-
ability. Six areas of psychological deficit were tive but inconclusive regarding a trend toward
associated with a diagnosis of AS: fine and poorer motor performance in the group with
gross motor skills, visual motor integration, AS. Considering the fact that many studies
visual-spatial perception, nonverbal concept demonstrating this pattern of verbal strengths
formation, and visual memory. Five areas of specific to AS had used varying diagnostic
psychological deficits were negatively corre- criteria, Ghaziuddin and Mountain-Kimchi
lated with a diagnosis of AS: articulation, ver- (2004) compared AS with HFA using unmodi-
bal output, auditory perception, vocabulary, fied, official diagnostic criteria. The study
and verbal memory. Finally, all but three sub- found that, overall, individuals with AS had
jects with AS presented with a neuropsycho- higher VIQ with greater strengths in fund of
logical profile consistent with an NLD knowledge and vocabulary relative to individu-
(Rourke, 1989). In contrast, only one subject als with HFA. A recent review of the literature
in the HFA group did so. These results indi- (Reitzel & Szatmari, 2003) confirmed this pat-
cated an overlap between AS, but not HFA, tern of differential neuropsychological pro-
and NLD, suggesting an empirical distinction files, with individuals with HFA consistently
between AS and HFA based on neurocognitive demonstrating lower scores on measures of
profiles. verbal functioning. Importantly, however, it
Since this study was published, a number of was noted that studies have not consistently
publications have replicated, or partially repli- demonstrated that individuals with AS possess
cated, its findings. Ehlers et al. (1997) com- nonverbal weaknesses or increased spatial or
pared results on an intelligence test across motor problems relative to individuals with
groups with AS and autism. The authors found HFA, echoing Miller and Ozonoff ’s conclu-
that individuals with AS exhibited stronger sions (2000). In other words, differences in
verbal ability with a weakness in subtests mea- neuropsychological profiles were explained in
suring visual-spatial organization and grapho- terms of an overriding effect of higher IQ in in-
motor skills (WISC-R Object Assembly, dividuals with AS relative to those with HFA,
104 Diagnosis and Classification

rather than a double dissociation of strengths creased cognitive demands. Ozonoff and col-
and deficits suggested by the NLD model. leagues (1991a) found that executive function
Several studies have failed to detect signif- did not differ between individuals with AS and
icant differences in neuropsychological pro- HFA; however, it was the most reliable distinc-
files between individuals with AS and HFA. tion between both groups and typical controls,
One of the earliest investigations in this area suggesting the primacy of deficits in this area
was conducted by Szatmari, Tuff, Finlayson, underlying a more broadly defined PDD group.
and Bartolucci (1990), who administered a The differences in results and conclusions
comprehensive test battery including intelli- reported in these studies of neurocognitive
gence, achievement, and neuropsychological functioning in AS and HFA most likely reflect
measures to groups with AS and HFA matched different diagnostic approaches and ascertain-
on full-scale IQ. Few differences were de- ment procedures adopted by the various
tected; the AS group tended to perform better groups of investigators. In general, studies
on a test of verbal concept formation (WISC- adopting the most stringent diagnostic proce-
R Similarities subtest), while the HFA group dures, that is, selecting the most phenomeno-
showed higher performance on a test of motor logically prototypical cases, found maximal
speed and coordination (Grooved Pegboard, distinctions between groups (e.g., Klin et al.,
nondominant hand). The authors concluded 1995; Lincoln et al., 1998). Studies that failed
that the lack of differentiation on the neu- to find distinctions between groups (e.g.,
rocognitive battery suggested that the AS and Ozonoff, Rogers, & Pennington, 1991; Szat-
HFA groups could be combined in a more gen- mari et al., 1990) tended to adopt broadened or
eral PDD category. A second study was con- otherwise modified definitions, thereby possi-
ducted by Ozonoff, Rogers, and Pennington bly reducing power to detect results ( however,
(1991). A neuropsychological battery was ad- see Miller & Ozonoff, 2000, for an exception).
ministered to groups with HFA and AS, This assumption, however, cannot be systemat-
matched on full-scale IQ and PIQ, as well as ically tested until investigators from different
chronological age. Although both groups dif- research groups begin to adopt consensual def-
fered on some measures from typical control initions and standardized procedures for case
subjects, they differed from each other only in assignment. For example, one study (Klin
regard to verbal memory, likely an artifact of et al., in press) applying three diagnostic
the significantly higher verbal skills in indi- schemes simultaneously to the same pool of
viduals with AS (Ozonoff, Rogers, & Penning- subjects with AS and HFA found significant
ton, 1991). The authors concluded that the two differences in IQ differential (VIQ relative to
groups could not be reliably differentiated in PIQ) when case assignment followed two of
terms of their neuropsychological profiles. the three diagnostic systems (the DSM-IV ap-
Though much of the research into neuropsy- proach and the unique features approach, as
chological distinctions has focused on verbal discussed previously), but not when case as-
versus nonverbal abilities, concurrent work has signment followed the third system (the early
addressed a particular aspect of neuropsycho- language approach).
logical skills, namely executive functions. One
marginally significant finding in Szatmari and Social-Cognitive Studies
colleagues’ (1990) study was that individuals
with autism performed less well than those A number of research studies have explored
with AS on a single measure of executive func- autistic children’s ability to impute mental
tions (Wisconsin Card Sort Test, perseverative states such as beliefs, desires, and intentions to
errors). Rinehart, Bradshaw, Tonge, Brereton, others and to themselves or to have a theory of
and Bellgrove (2002) assessed executive func- other people’s (and their own) subjectivity—a
tions using visual-spatial tasks subserved by theory of mind (ToM; Baron-Cohen, Tager-
fronto-striatal brain regions. On this measure, Flusberg, & Cohen, 1999). The major hypothe-
individuals with AS were found to have intact sis emerging from this work is that individuals
performance, while those with HFA exhibited with autism lack this capacity and that this
inhibitory deficits that became evident with in- fundamental deficit may explain Wing’s
Asperger Syndrome 105

(Wing & Gould, 1979) triad of symptoms emission tomography to examine brain activ-
defining autism, namely, impairment of social ity in individuals with AS during performance
and communicative functioning and imagina- of a ToM task. Patients with AS, like typical
tive activities (see Chapter 47, this Handbook, controls, showed selective activation in the
Volume 2, for a review). Problems in these medial prefrontal cortex; however, in subjects
abilities, as measured by joint attention and with AS the activity was localized in a
pretend play, are evident in children with slightly different region of the PFC ( lower
autism as early as 18 months (Baron-Cohen, and more anterior). This finding is consistent
Cox, Baird, Sweettenham, & Nightingale, with behavioral data suggesting existing, but
1996; Charman et al., 1997). abnormal, ToM abilities in AS.
Several studies have found evidence sug- Though much research suggests differen-
gesting that differential abilities in ToM may tial ToM abilities in individuals with AS and
help to distinguish AS from autism. Ozonoff, autism, several studies have obtained results
Pennington, and Rogers (1991) and Ozonoff, inconsistent with this conclusion. Dahlgren
Rogers, and Pennington (1991) investigated and Trillingsgaard (1996) found that individ-
“ first- and second-order theory of mind.” uals with AS demonstrated intact ToM skills
First-order attributions refer to a situation in relative to typical individuals, but so did the
which a subject must attribute a mental HFA group; ToM performance was not useful
state—for example, a false belief—to another in differentiating between the two diagnoses.
person, whereas second-order attributions re- A third pattern of results was obtained by
quire recursive thinking about mental states, Baron-Cohen, Wheelwright, and Jolliffee
in which a subject is required to predict one (1997), who found that individuals with AS
person’s thought about another person’s and HFA were comparably impaired on ad-
thoughts. First-order tasks begin to be mas- vanced tasks of ToM.
tered by age 4 in normally developing children, Although at least some individuals with AS
whereas the ability to make second-order at- demonstrate intact abilities in many of these
tributions develops later, at around the age of 7 laboratory studies, their ability to take the per-
(Perner & Wimmer, 1985). In this study, the spective of another person—to adequately
autistic group exhibited significant impair- evaluate other people’s interests, beliefs, in-
ment in relation to both the AS and an age- and tentions and feelings—is typically impaired in
IQ-matched control group, and the AS group real-life contexts. This observation was origi-
showed intact ToM abilities. Bowler (1992) nally made by Asperger himself (1944) and
also found that, on similar tasks, individuals was reiterated by Van Krevelen (1971), who
with AS showed no impairment in ToM abili- captured this phenomenon in terms of a lack of
ties relative to both IQ-matched typical indi- an intuitive understanding of other people’s
viduals and a group of schizophrenic patients. social behavior or to avail themselves from
Ziatas, Durkin, and Pratt (2003) reported that empathic feelings and from negotiating social
children with AS showed lesser impairments interaction by means of quick-paced, nonver-
in mentalizing abilities than children with bal means. In this context, individuals with AS
autism. Several publications indicate that were said to mediate their social and emo-
these patterns may be evident early in develop- tional exchange through explicit verbal and
ment; assessments at 18 months failed to de- logical means, cognitively, rigidly, and in a
tect impairments in joint attention and pretend rule-governed fashion, rather than affectively,
play in children subsequently diagnosed with intuitively, and in a self-adjusting fashion (see
AS (Baird et al., 2000; Cox et al., 1999). Stud- also Klin, Jones, Schultz, & Volkmar, 2003).
ies of adults have also indicated that patients Studies to date offer provocative intima-
with AS exhibit more subtle impairments in tions of ToM as a useful differentiator between
ToM abilities, as demonstrated by inconsistent AS and autism, but results are not fully consis-
or delayed responses on more sophisticated tent with clinical observations of social experi-
and naturalistic assessments (Royers, Buysse, ences of people with AS. Furthermore, these
Ponnet, & Pichal, 2001; Kaland et al., 2002). findings are vulnerable to several methodologi-
One study (Happé et al., 1996) used positron cal criticisms. Frith (2004) suggested that
106 Diagnosis and Classification

findings pointing to a milder impairment in fast, simultaneous, holistic, and intuitive reac-
ToM among individuals with AS may simply tions (e.g., highly synchronized and constantly
reflect high verbal ability and intelligence; changing facial expressions, posture, and
these cognitive characteristics provide indi- prosodic cues among a host of other dynamic
viduals with AS with an advantage in that they elements). Thus, traditional ToM tasks may
may be able to succeed on ToM tasks by means oversimplify social problem solving, creating
of well-reasoned responses based on logical an illusion of competence not commensurate
inference, rather than true social intuition. In- with demonstrated abilities in real-life social
deed, several studies of ToM abilities have contexts. Aware of this problem, Tager-Flusberg
highlighted the important role played by verbal (2001) contrasted two aspects of ToM compe-
skills on subjects’ performance (e.g., Happé, tence: first, the ability to successfully acquire
1995; Sparrevohn & Howie, 1995). To avoid social information from the environment
the formulation of an explicit task and possible (social perception) and, second, the ability to
verbal mediation, several studies have exam- reason about it (social cognition). She hypothe-
ined the tendency to attribute social meaning sized that autism entails impairment in both
(including ToM) to ambiguous visual displays aspects, but AS involves only problems with
(geometric shapes moving and interacting in a social perception. This conceptualization of-
“social manner ”; e.g., Klin, 2000), showing fers a potential explanation for the problems
significant deficits in ToM skills in individu- experienced by individuals with AS in real life
als with AS, which were similar to those with that are not replicated in lab studies; real-life
HFA. One study (Grossman, Klin, Carter, & situations entail fluid, ambiguous, contextually
Volkmar, 2000) used a paradigm deliberately defined information, while laboratory studies
created to test the impact of verbal mediation tend to provide more explicit information.
in processing social stimuli. In this case, the These criticisms indicate that to truly as-
task involved the processing of facial expres- sess ToM ability as a differentiating factor be-
sions, rather than ToM, although its results are tween AS and HFA, there is a need for more
relevant to the present discussion. Individuals naturalistic and ecologically valid studies. In-
with AS were found to be particularly suscep- vestigators interested in ToM capacities must
tible to language interference in their perfor- devise experimental situations less amenable
mance when facial expressions were presented to logical and verbal solutions. Optimally,
simultaneously with mismatching written ver- these should involve processing of more visual
bal labels. While typical controls’ accuracy rather than verbal stimuli and more naturalis-
and reaction time when identifying facial ex- tic and socially contextualized rather than ab-
pressions were by and large unaffected by this stract and logical situations, and processing
“strooplike” phenomenon, this was not so for time (i.e., latency of response) should be an
the individuals with AS, for whom the faces important parameter to be considered given
appeared less salient than the verbal labels. that, clinically, individuals with AS typically
In this context, Klin, Jones, Schultz, and cannot avail themselves of their formal social
Volkmar (2002) proposed that the success evi- knowledge in quick-paced, simultaneously
denced by individuals with AS in laboratory shifting, social situations.
measures may be a function of the poor eco-
logical validity of the paradigms used. In other Medical and Neurobiological Studies
words, these individuals may be able to “solve”
social problems if they are explicitly formu- A number of studies have attempted to differ-
lated to them, whereas in real life there is a entiate AS from HFA based on neurobiological
need to naturally seek salient social cues and measures. Several reports of neurobiological
generate the problem to be solved, and many abnormalities associated with AS have ap-
such “problems” are not necessarily amenable peared, although no consensus has been
to logical translation or verbal mediation be- reached because many of these reports have
cause this process would imply very slow, se- been case studies or involved very small sam-
quential, piecemeal, and laborious processing, ples and because of the ubiquitous nosologic
whereas real-life social situations demand issues discussed earlier. Wing (1981) noted a
Asperger Syndrome 107

higher than expected frequency of perinatal same area, although it was larger on the right
problems (nearly half of her original sample) side; his images also showed decreased tissue
among this population. However, Gillberg and in the anterior-inferior right temporal lobe,
Gillberg (1989) concluded that reduced obstet- suggesting an atrophic process or a regional
ric optimality was more common in autism neurodevelopmental growth failure. Both fa-
than in AS. Other reports have associated cer- ther and son exhibited a similar neuropsycho-
tain medical conditions with individual cases logical profile, and the similarity of brain
of AS—for example, aminoaciduria (Miles & anomalies and symptoms is suggestive of po-
Capelle, 1987) and ligamentous laxity (Tan- tential familial transmission.
tam, Evered, & Hersov, 1990). Casanova, Bux- A more limited set of studies has employed
hoeveden, Switala, and Roy (2002) found functional imaging and spectroscopy to
abnormalities in the organization of mini- demonstrate differences in brain function and
columns in both AS and autism, suggesting neurotransmitter concentrations between indi-
common underlying neural pathology. A case viduals with AS and typical controls. Murphy
series by Gillberg (1989) reported high fre- and colleagues (2002) used proton magnetic
quencies of medical anomalies, also common resonance spectroscopy to examine the rela-
to both diagnostic groups; however, these re- tionship between abnormalities in frontal and
sults have been challenged (Rutter, Bailey, parietal lobes and clinical symptomatology.
Bolton, & Le Couteur, 1994). Compared to typical controls, patients with
Neuroimaging studies of AS and autism AS had higher levels of N-acetylaspartate
have become one of the most prolific areas of (NAA), creatine and phosphocreatine, and
research in the field. Nevertheless, there is choline, neurotransmitters that serve as indi-
scant data to differentiate neuroanatomical cators of neuronal density, mitochondrial
structures and functions between AS and metabolism, phosphate metabolism, and mem-
autism. Most studies have sampled one group brane turnover. Furthermore, there were posi-
or the other or collapsed them into an ASD tive correlations between neurotransmitter
group. Some studies have, however, focused on levels and clinical presentation: Higher levels
individuals with AS as compared to typical of prefrontal NAA were associated with obses-
controls and found discrepancies in brain sional behavior, and increased prefrontal
structure. Berthier, Starkstein, and Leiguarda choline was associated with social function.
(1990) reported MRI results indicating left The authors concluded that individuals with
frontal macrogyria and bilateral opercular AS exhibit abnormalities in neuronal integrity
polymicrogyria in patients with AS. These of the prefrontal cortex, which relates to their
findings were linked to cortical migration ab- unique pattern of symptoms. McKelvey, Lam-
normalities in a first-degree relative with bert, Mottron, and Shevell (1995) reported on
bipolar disorder. Other studies have reported three patients with abnormal right hemisphere
gray tissue anomalies in subjects with AS as functioning on SPECT imaging, a finding
compared to typical controls (McAlonan et al., consistent with the NLD neurocognitive
2002). Several case studies have revealed left model of AS.
temporal lobe damage (Jones & Kerwin, 1990; A handful of studies have attempted to study
CT scan) and left occipital hypoperfusion differential brain structure in individuals with
(Ozbayrak, Kapucu, Erdem, & Aras, 1991; AS and autism. Studies have suggested that,
SPECT) in patients with AS. In a case study across diagnosis, higher functioning individuals
by Volkmar and colleagues (1996), a father on the autism spectrum tend not to exhibit hy-
and son with AS showed virtually identical ab- poplasia of the neocerebellar vermis, a finding
normalities on their MRIs. The father’s im- more common in lower functioning people on
ages showed a large, bilateral, V-shaped wedge the spectrum that may be associated more with
of missing tissue just superior to the ascending mental retardation than with autism (e.g., Piven
ramus of the Sylvian fissure, at about the level et al., 1992; Piven, Palmer, Jacobi, Childress, &
where the middle frontal gyrus normally inter- Arndt, 1997). In some contrast, Lincoln and
sects with the precentral sulcus. The son’s im- colleagues (1998) reported that individuals
ages showed similar dysmorphology in the with autism (relative to AS) exhibited more
108 Diagnosis and Classification

consistent pathology of the cerebellar vermis, categorical subgrouping, dimensional analyses

thinner posterior corpus callosum, and (e.g., association between quantified brain ac-
smaller anterior corpus callosum. One recent tivation and measurable aspects of the pheno-
study has provided exciting information about type) may prove more powerful in the long
potential neuroanatomical distinctions be- run. Equally important are the initial efforts to
tween subgroups of ASDs. Lotspeich et al. move to measures of interconnectivity of key
(2004) compared individuals with HFA and brain systems subserving cognitive and social
low-functioning autism (LFA) and AS. The processing (see Chapter 19, this Handbook,
study indicated enlarged cerebral gray matter this volume).
volumes in HFA and LFA relative to controls.
The AS group showed enlargement that was in- Genetics
termediate with respect to patients with
autism and typical controls but was not signif- From its original description by Asperger
icantly different from either. The authors in- (1944), there has been evidence for familial
ferred that cerebral gray matter may correlate vulnerabilities in AS. As noted, Asperger ob-
with symptom severity, suggesting that AS served similar characteristics in family mem-
represents the mild end of the continuum. This bers, particularly in fathers. This genetic
study also detected differential relationships predisposition is in keeping with more recent
between brain structure and neurocognitive findings demonstrating that the ASDs are
function; cerebral white matter volume was among the most heritable of psychiatric con-
positively correlated with PIQ only in subjects ditions (Rutter, Bailey, Simonoff, & Pickles,
with AS, and gray matter correlated with PIQ 1997), although knowledge about patterns of
in both HFA and AS, but in different direc- inheritability and genetic etiologies is still
tions (positive correlation in AS, negative cor- limited (see Chapter 16, this Handbook, this
relation in HFA). Another study added a volume). Given the heterogeneity of the
comorbid component, namely Tourette’s syn- ASDs, it is of little surprise, but of great im-
drome (TS) to this discussion. Berthier, Bayes, portance, to consider the fact that the ASDs
and Tolosa (1993) used MRI to study brain are the more severe manifestations of a
structure in patients with comorbid AS and TS broader and possibly more prevalent pheno-
and with patients with TS alone. Results re- type of social-communicative difficulties and
vealed cortical and subcortical abnormalities behavioral rigidities (e.g., Bailey et al., 1995;
in 5 of the 7 patients with AS and TS, but only Le Couteur et al., 1996).
in 1 of the 9 patients with TS only. Consistent with Asperger’s original obser-
Findings to date do not present a clear pic- vation, a number of case reports have reported
ture of the neurobiological underpinnings of similar traits in family members, especially
AS or replicated distinctions among the ASDs. among fathers (Bowman, 1988; DeLong &
In many ways, issues of definition of the ASD Dwyer, 1988; Gillberg, Gillberg, & Steffen-
subgroups make it impossible to derive firm burg, 1992; Volkmar et al., 1996). Others have
conclusions from the available data. Addition- reported data on social vulnerabilities in other
ally, there are methodological issues related to relatives. Multiple births concordant for AS
intersite differences in imaging techniques have been observed (Burgoine & Wing, 1983).
(e.g., Lotspeich et al., 2004). Nevertheless, Volkmar and colleagues (1997) examined fam-
this growing volume of literature holds great ily history data and found that nearly half of
promise, particularly in those studies in which the families surveyed reported information
a more specific brain-behavior relationship is positive to AS, ASDs, and the broader autism
found. In this context, neuroimaging studies, phenotype in first-degree relatives, most com-
and particularly functional MRI research, monly among males. The probands with AS
should be guided by neuropsychological and were more likely (relative to general popula-
social-cognitive models that hold the potential tion rates) to have a sibling diagnosed with
to explain core aspects of the presentation of autism or an ASD; this was twice as likely in
individuals with ASDs. And although analyses male siblings as in female siblings. Rates of
have often been performed on the basis of autism and ASD were also increased in first
Asperger Syndrome 109

cousins. Klin et al. (in press) reported similar and hypothesized neuropsychological, social-
results. When using the unique features diag- cognitive, or neurobiological patterns.
nostic scheme (as described earlier), it was of
interest that the rates of ASDs or the broader Outcome
autism phenotype in parents and grandparents
(males and females) of probands with AS was Asperger’s (1944) initial description pre-
over triple the rate found for the same relatives dicted a positive outcome for many of his pa-
of probands with HFA (17% versus 5%). This tients, largely based on his hope that they
finding suggests even stronger genetic contri- would be able to use their special talents for
butions in families of probands of AS relative the purpose of obtaining employment and lead-
to those of probands with autism (Volkmar & ing self-supporting lives. His observation of
Klin, 2000). similar traits in family members encouraged
Though still limited, there are some reports the notion that patients could marry and raise
of autism and AS co-occurring in different families. Although his account was tempered
members of the same family (e.g., Wing, somewhat by experience (Asperger, 1979),
1981). In our clinical experience and in a pre- Asperger continued to believe that a more pos-
liminary report (Klin et al., in press), we have itive outcome was a central criterion differen-
observed some sibships in which one brother tiating individuals with his syndrome from
has autism and the other has AS. The fact that those with Kanner’s autism. In considering
the triplets reported by Burgoine and Wing autism inclusive of low-functioning cases, this
(1983) had some aspects of their history more would certainly be the case (Gillberg, 1991).
similar to patterns observed in autism also When compared with individuals with HFA,
suggests a common familial link between outcome of individuals with AS is more likely
autism and AS. Collectively, these findings to be more positive on the basis of case re-
point to genetic mechanisms in common not ports, anecdotal evidence, and some prelimi-
only between autism and AS but also among nary studies, but data supporting this
all ASDs (Frith, 2004). hypothesis are still scant. More positive out-
In addition to evidence of familial patterns comes for individuals with AS could simply re-
in AS, direct evidence of a genetic component flect the overall effects of the preserved
has been published. Specific genetic abnor- language and cognitive skills required for the
malities have been reported in case studies of diagnosis (for at least the first years of life).
patients with AS—for example, one case with Nevertheless, some studies support the notion
a balanced de novo translocation (Anneren, of differential outcomes in individuals with
Dahl, Uddenfeldt, & Janols, 1995), one case AS and HFA. Several studies examined out-
with an autosomal fragile site (Saliba & Grif- come in young adults with AS (Newson, Daw-
fiths, 1990), and a possible association with son, & Everaard, 1984/1985; Tantam, 1991)
fragile X syndrome (Bartolucci & Szatmari, and reported that although most remained at
1987). Tentler and colleagues (2003) found home despite high cognitive potential, a mi-
two patients with AS with balanced transloca- nority of them had married and/or held regular
tions on chromosome 17, perhaps suggesting a employment. Though a comparable sample
candidate sequence. Although these findings with autism was not assessed, this outcome ap-
are only preliminary, they provide guidelines pears better than typically expected for indi-
for future research studies and concrete viduals with HFA. Gillberg (1998) suggested
sources for explicating the shared genetic that patients with AS experience more positive
components of AS and other ASDs. outcomes, especially in the domains of acade-
Fast-accruing genetic data have the promise mics and self-help skills. In contrast, Szat-
of moving research on ASD subtypes from the mari, Bartolucci, et al. (1989b) compared
rather imprecise phenotypic comparisons to outcomes in children with AS and HFA and
date to potential endophenotypes mediating noted minimal differences. Others have re-
syndrome expression. For this to happen, how- ported that older individuals and adults with
ever, there is a need for much more refined phe- AS and HFA are so similar that they are diffi-
notypic characterization of both core symptoms cult or impossible to distinguish on the basis of
110 Diagnosis and Classification

outcome (Gilchrist et al., 2001; Howlin, 2003; Summary of Validation Studies

Szatmari et al., 2000). All of these three stud-
ies, however, used the early language approach A common thread throughout the reviewed val-
as a diagnostic scheme for case assignment, idation studies is the sore need for consensual
which, at least in our experience, yields the and reliable diagnostic criteria for AS and
least contrast between the two groups (Klin demonstration of the fidelity of the concept
et al., in press). A recent study (Szatmari, across sites. Future research should attempt to:
Bryson, Boyle, et al., 2003) looked into predic-
tors of outcome by ASD subtype by following 1. Ensure that dif ferences between AS and
children with autism and AS between the ages other conditions are independent of the cri-
of 4 and 6 years through 10 to 13 years using teria defining disorders. Historically, stud-
adaptive behavior (i.e., levels of ability, which ies have employed circular reasoning, with
are norm-based) and measures of autistic results reflecting the diagnostic criteria
symptoms (i.e., levels of disability) and found rather than providing additional informa-
that the power of prediction was stable over tion. Dependent variables should be inde-
time but different for the two groups. The au- pendent of (i.e., not a function of, not
thors found that the association between lan- highly correlated with, not associated with,
guage skills and outcome was stronger in the and not a result of ) diagnostic criteria.
autism group than in the AS group, suggesting 2. Adopt an agreed-on definition, and opera-
that for the individuals with AS, higher lan- tionalize it in much greater detail (including
guage skills did not confer any advantage. Of the process used to obtain data used in diag-
great practical importance, however, was that nostic decision making). One possibility is
the explanatory power of the predictor vari- to modify well-proven diagnostic instru-
ables was greater for communication and so- mentation such as the Autism Diagnostic
cial skills than for autistic symptoms, Interview-Revised (Rutter, Le Couteur, &
suggesting the possibility that adaptive compe- Lord, 2003) and the Autism Diagnostic
tence ( levels of self-sufficiency) and autistic Observation Schedule (Lord, Rutter,
symptomatology ( levels of disability) may DiLavore, & Risi, 1999; see Chapter 28,
constitute rather independent facets of these this Handbook, Volume 2) to include an “AS
children’s development. This finding was cor- module” covering the more unique features
roborated in a recent study (Klin, Saulnier, of AS in every domain of characterization
Sparrow, Cicchetti, Lord, & Volkmar, in (i.e., onset, social, communication, and
press). From a practical perspective, this point narrow interests and behaviors), as well as
suggests that intervention programs should differential diagnostic algorithms keyed to
emphasize both reduction of symptoms and the consensually adopted defining criteria.
improvement of self-sufficiency skills given It is indeed very unlikely that any of the is-
that the focus on the former does not necessar- sues related to the nosologic status of AS
ily imply advancement of the latter. discussed in this chapter will be resolved in
Despite these interesting preliminary data, the absence of an effort of this kind.
it is important to recognize that the status of 3. Include larger samples and multisite com-
research into differential outcomes is still parisons. Much of our understanding of the
quite limited (Howlin, 1997). This is particu- putative differences between AS and HFA
larly regrettable given that of all the possible stems from reports involving case studies
reasons to separate these two conditions, dif- or small case series. Larger studies will not
ferential outcome would probably be the one only improve power to detect differences
with the greatest practical relevance. There is but also increase the likelihood that results
a need to ensure, however, that well-known are meaningful and replicable. And there
mediators of outcome, and particularly IQ and have been striking differences in findings
language levels past the first years of life, are across research groups, which may reflect
not components of the diagnostic definitions, not only different methodologies but also
which would create circularity in the study de- different biases in subject recruitment. To
sign and would likely result in spurious differ- overcome this critical difficulty, there is a
ences between the two groups. need for cross-site studies using tools of the
Asperger Syndrome 111

kinds described earlier within a common individuals with AS. For a more comprehen-
effort to carry out a well-controlled, multi- sive discussion of the issues outlined here, the
center field trial. reader is referred to a more complete text
4. Avoid oversimplifications of key clinical (Klin, Sparrow, Marans, Carter, & Volkmar,
phenomena. Historically, research on AS 2000) and Chapter 29, Volume 2, of this
has failed to acknowledge, describe, opera- Handbook.
tionalize, or quantify the complex set of AS, like the other PDDs, involves delays
clinical phenomena involved in key con- and deviant patterns of behavior in multiple
cepts defining AS such as circumscribed areas of functioning. Thorough evaluation of
interests, social awkwardness in the pres- all relevant domains requires different areas of
ence of social motivation, pedantic style, expertise, including overall developmental
and verbosity as a strategy for social adap- functioning, neuropsychological features, lan-
tation among others. Many studies gloss guage and communication skills, adaptive
over these phenomena, providing limited or functioning, and behavioral status. Therefore,
no description of the sample beyond rather the clinical assessment of individuals with this
vague statements. disorder is most effectively conducted by an
5. Ensure the achievement of, and report data experienced, interdisciplinary team.
on, interrater reliability of procedures used In the majority of cases, a comprehensive
in diagnostic process (e.g., semistructured interdisciplinary assessment will involve the
interviews, best-estimate diagnoses) and, if following components: a thorough develop-
possible, document the entire process to mental and health history, psychological and
make possible the replication of the various communication assessments, and a diagnostic
procedures by members of a dif ferent re- work-up including differential diagnosis.
search team. Further consultation regarding behavioral
management, motor disabilities, possible neu-
Additionally, it may be useful for investiga- rological concerns, comorbidities and associ-
tors to depart from the prevalent model of vali- ated psychopharmacology, and assessment
dation research in which data are collected and related to advanced studies or vocational
compared on the basis of clinical groups training may also be needed. Given the pre-
formed by a priori diagnostic assignments. vailing difficulties in the definition of AS and
Because the validity of ASD subtypes is still the great heterogeneity of the condition, it is
uncertain (see Chapter 1, this Handbook, this crucial that the aim of the clinical assessment
volume), much could be learned by reversing be a comprehensive and detailed profile of the
the process. Researchers could assess meaning- individual’s assets, deficits, and challenges,
ful and interesting patterns (e.g., detailed onset rather than simply a diagnostic label. Effective
and early development patterns, specific neu- educational and treatment programs can be de-
ropsychological, social-cognitive, or neurobio- vised only on the basis of such a profile, given
logical profiles) and then carefully chart the the need to address specific deficits while cap-
range of phenotypic expressions associated with italizing on the person’s various resources and
them. This approach would redefine phenotype strengths.
as a dependent variable and open the door for The psychological assessment aims at es-
the use of statistical techniques to empirically tablishing the overall level of intellectual func-
define the existence of true categories. tioning, profiles of psychomotor functioning,
verbal and nonverbal cognitive strengths and
CLINICAL ASSESSMENT OF weaknesses, style of learning, and adaptive be-
INDIVIDUALS WITH ASPERGER haviors (or independent living skills). At a
SYNDROME minimum, the psychological assessment should
include assessments of intelligence and adap-
The topic of clinical assessment of children tive functioning, although the assessment of
with ASD is reviewed in much greater detail more detailed neuropsychological skills can be
in Chapters 28 through 32, Volume 2, of this of great help to further delineate the child’s
Handbook. In this section, we briefly discuss profiles of strengths and deficits (e.g., in re-
assessment procedures uniquely relevant to gard to organizational skills). A description of
112 Diagnosis and Classification

results should include not only quantified in- skills (e.g., understanding of the language of
formation but also a judgment as to how repre- mental states including intentions, emotions,
sentative the child’s performance was during and beliefs), reciprocity, and rules of conver-
the assessment procedure and a description of sation.
the conditions that are likely to foster optimal The diagnostic assessment should integrate
and diminished performance. For example, the information obtained in all components of the
child’s responses to the amount of structure comprehensive evaluation, with a special em-
imposed by the adult, the optimal pace for phasis on developmental history and current
presentation of tasks, successful strategies to symptomatology. It should include observations
facilitate learning from modeling and demon- of the child during more and less structured pe-
strations, and effective ways of containing off- riods. This effort should take advantage of ob-
task and maladaptive behaviors such as servations in all settings, including the clinic’s
cognitive and behavioral rigidity (e.g., verbal reception area (e.g., contacts with other chil-
perseverations, perfectionism), distractibility dren or with family members), the halls (e.g.,
(e.g., difficulty inhibiting irrelevant re- how the child interacts initially with the exam-
sponses, tangentiality), and anxiety are all im- iners), as well as in the testing room during
portant observations that can be extremely breaks, periods of silence, or otherwise un-
useful for designing and optimizing interven- structured situations. Often, the child’s dis-
tion programs. Within the psychological as- ability is much more apparent during such
sessment, particular emphasis should be periods in which the child is not given any in-
placed on the assessment of adaptive function- struction and has no adult-imposed expecta-
ing, which refers to capacities for personal and tion as to how to behave. Specific areas for
social self-sufficiency in real-life situations. observation and inquiry include the patient’s
The importance of this component of the clini- patterns of special interest and leisure time,
cal assessment cannot be overemphasized. Its social and affective presentation, quality of at-
aim is to obtain a measure of the child’s typi- tachment to family members, development of
cal patterns of functioning in familiar and rep- peer relationships and friendships, capacities
resentative environments such as the home and for self-awareness, perspective-taking and
the school, which may contrast markedly with level of insight into social and behavioral prob-
the demonstrated level of performance and lems, typical reactions in novel situations, and
presentation in the clinic. It provides the clini- ability to intuit another person’s feelings and
cian with an essential indicator of the extent to infer another person’s intentions and beliefs.
which the child is able to use his or her poten- Problem behaviors that are likely to interfere
tial (as measured in the assessment) in the pro- with remedial programming should be noted
cess of adaptation to environmental demands. (e.g., anxiety, temper tantrums). The child’s
A large discrepancy between intellectual level ability to understand ambiguous, nonliteral
and adaptive level signifies that a priority communications (particularly teasing and sar-
should be made of instruction within the con- casm) should be further examined, particu-
text of naturally occurring situations to foster larly in regard to the child’s patterns of
and facilitate the use of skills to enhance qual- response (e.g., misunderstandings of such
ity of life. communications may elicit aggressive behav-
The communication assessment should ex- iors). Other areas of observation involve the
amine nonverbal forms of communication presence of obsessions or compulsions, ritual-
(e.g., gaze, gestures), nonliteral language (e.g., ized behaviors, depression and panic attacks,
metaphor, irony, absurdities, and humor), integrity of thought, and reality testing.
suprasegmental aspects of speech (e.g., patterns One important aspect of the diagnostic
of inflection, stress, and volume modulation), characterization of individuals with AS relates
pragmatics (e.g., turn taking, sensitivity to cues to the need to differentiate diagnostic instru-
provided by the interlocutor), and content, ments from diagnostic screeners. The former
coherence, and contingency of conversation. typically involve comprehensive, semistruc-
Particular attention should be given to persever- tured procedures including both information
ation on circumscribed topics, metalinguistic provided by parents or caregivers and direct
Asperger Syndrome 113

examination and aim at producing a detailed more optimistic about individuals with AS
profile of developmental patterns and current given the common finding of relatively pre-
behavioral presentation needed for diagnostic served social motivation. Unfortunately, there
assignment. The latter typically involve the is currently a paucity of systematic data
completion of a brief checklist by parents, demonstrating the effectiveness of particular
caregivers, teachers, or other professionals, or interventions, although some progress has
by self-report, and aim at identifying those been made in this area (see Klin & Volkmar,
with a higher probability of having the condi- 2003, for a summary of treatment studies and
tion from the general population. It is critical of treatment approaches). Medications may be
to emphasize that screeners are not meant to effective in treating associated features of the
replace the diagnostic process. Rather, they disorder, such as inattention, anxiety, or de-
are meant to provide the basis for further refer- pression, although psychopharmacology can-
ral. Currently, the only diagnostic instrument not yet target core impairments in AS or, for
available that considers the distinction of that matter, other ASDs. In recent years, clini-
autism from AS is the Asperger Syndrome cians and educators experienced with this
(and high-functioning autism) Diagnostic In- population have published a number of vol-
terview (ASDI; Gillberg, Gillberg, Rastam, & umes proffering useful education and inter-
Wentz, 2001), although there are no data sup- vention strategies (e.g., Attwood, 1998; Bashe
porting this instrument’s validity insofar as & Kirby, 2001; Howlin, 1999; Myles & Simp-
the distinction between AS and HFA is con- son, 1997; Ozonoff, Dawson, & McPartland,
cerned. There are no other diagnostic instru- 2002; Powers & Poland, 2002; Wing, 2001). In
ments developed for the specific purpose of general, recommended interventions focus on:
subtyping the ASDs into autism, AS, and (1) devising strategies to take advantage of the
PDD-NOS. As mentioned earlier, there seems unique constellation of strengths characteris-
to be a great need for modification of current tic of AS to compensate for areas of difficulty
instruments to make possible this differentia- and (2) modifying contexts (i.e., environ-
tion, although this recommendation is closely ments) to maximally support the learning and
tied to other developments in validity studies behavioral styles of this population. The rec-
of AS. In contrast, there are a large number of ommended content or treatments should be
diagnostic screeners purported to identify in- tailored, based on a thorough assessment, to
dividuals with AS, some of which appear to be the individual needs of the patient. However,
of help in identifying vulnerabilities within nearly universally applicable foci include ac-
the autism spectrum, but none has been shown quisition of basic social and communication
to differentiate (nor has that been a stated ob- skills, adaptive functioning, organizational
jective) individuals with AS from those with skills, and, depending on what is developmen-
autism or PDD-NOS. tally appropriate, academic or vocational
skills. It is also crucial that any intervention
TREATMENT AND INTERVENTIONS program incorporate techniques to encourage
generalization of acquired skills to ensure the
The topic of treatment and interventions is use of learned abilities in novel contexts.
briefly summarized here, with an emphasis on
issues that are uniquely relevant to individuals Securing Services
with AS. More comprehensive guidelines are
reviewed in Chapters 33 through 44, Volume 2, Despite the great progress in dissemination of
of this Handbook. awareness and understanding of AS, the au-
As in autism, treatment of AS is essen- thorities deciding on eligibility for services
tially supportive and symptomatic, and to a may fail to appreciate the severity of the vari-
great extent, overlaps with the treatment ous disabilities exhibited by individuals with
guidelines applicable to individuals with HFA AS. Proficient verbal skills, overall IQ usually
(Mesibov, 1992). Though a similar set of rec- within the normal range, and a solitary
ommendations applies to individuals with lifestyle often mask outstanding deficiencies
both conditions, interventionists are often observed primarily in novel or otherwise
114 Diagnosis and Classification

socially demanding situations, thus decreasing ical profile of assets and deficits; specific in-
the perception of the very salient needs for tervention techniques should be similar to
supportive intervention. Hence there is a need those usually employed for many subtypes of
for active participation on the part of the clini- learning disabilities, with an effort to circum-
cian, together with parents and possibly an ad- vent the identified difficulties by means of
vocate, to forcefully pursue the student’s compensatory strategies relying on areas of
access to appropriately supportive programs. strength, in this case, usually verbal abilities.
The formalization of the diagnosis in ICD-10 If significant motor and visual-motor deficits
and DSM-IV, together with the directives from exist, the individual should receive physical
the National Research Council report (2001) and occupational therapies and, where avail-
that individuals with all forms of ASDs (in- able, assistive technologies (e.g., using a lap-
cluding AS) should be provided with intensive top to type assignments rather than writing).
and comprehensive educational programs of Occupational therapy should not only focus on
the kinds prescribed for individuals with traditional techniques designed to remediate
autism, has facilitated this process somewhat. motor deficits but also reflect an effort to inte-
In the past, many individuals with AS were di- grate these activities with learning of visual-
agnosed as having learning disabilities with spatial concepts, visual-spatial orientation,
some eccentric features, a nonpsychiatric di- and body awareness.
agnostic label that is much less effective in se-
curing services. Others, who were given the Adaptive Functioning
diagnosis of autism or PDD-NOS, had often to
contend with educational programs designed The acquisition of self-sufficiency skills in all
for much lower functioning children, thus fail- areas of functioning should also be a priority.
ing to have their relative strengths and unique Because individuals with AS tend to rely on
disabilities properly addressed. Yet another rigid rules and routines, these can be used to
group of individuals with AS is sometimes foster positive habits and enhance the person’s
characterized as exhibiting social-emotional quality of life and that of family members. The
maladjustment (SEM), an educational label teaching approach should follow closely the
that is often associated with conduct problems guidelines set earlier (see preceding Learning
and volitional maladaptive behaviors. These subsection) and should be practiced routinely
individuals are often placed in educational set- in naturally occurring situations and across
tings for individuals with conduct disorders, different settings to maximize generalization
thus allowing for the worse mismatch possible, of acquired skills.
namely, of individuals with a very naïve un-
derstanding of social situations with those Behavioral Management
who can and do manipulate social situations to
their advantage without the benefit of self- Challenging behaviors are common among in-
restraint. dividuals with AS. As noted, their motivations
are rarely malicious and are more likely to
Learning stem from difficulties with arousal regulation
and poor emotional insight into self and others.
In educating students with AS, it is crucial to Specific problem-solving strategies, usually
take advantage of their typically strong lan- following a verbal algorithm, may be taught for
guage skills and concrete style of thinking. handling the requirements of frequently occur-
Skills, concepts, appropriate procedures, cog- ring, troublesome situations (e.g., involving
nitive strategies, and behavioral norms may be novelty, intense social demands, or frustra-
most effectively taught in an explicit and rote tion). Training is usually necessary for recog-
fashion using a parts-to-whole verbal instruc- nizing situations as troublesome and for
tion approach, where the verbal steps are in selecting the best available learned strategy to
the correct sequence for the behavior to be ef- use in such situations. Cognitive and behav-
fective. The educational program should be ioral strategies for anxiety management (e.g.,
tailored to the child, not to the diagnosis, and breathing exercises) are often helpful in teach-
derived from the individual’s neuropsycholog- ing students to control negative emotions. In
Asperger Syndrome 115

designing any intervention to control problem- include typical peer models, but it is important
atic behaviors, it is important to collect data to that typical peers receive prior training) work-
both understand the function that they are ing in concert with one or more educators.
serving to the child (i.e., using functional be- They are uniquely beneficial in allowing for
havior analysis) and to ascertain a true esti- didactic training and practice with peers in a
mate of the efficacy of treatment. single setting. Furthermore, they may foster
development of supportive relationships (for
Communication and Social Skills Training both children and their parents) that endure
after group has concluded. Teaching may in-
Because individuals with AS experience com- clude the following:
munication deficits predominantly within the
domain of pragmatics, these skills are best 1. Appropriate nonverbal behavior (e.g., the
taught by or with the support of a communica- use of gaze for social interaction, monitor-
tion specialist with expertise in this area. Con- ing and patterning inflection of voice).
versation skills should be taught directly This may involve imitative drills, working
including self-monitoring level of diction and with a mirror, and so forth.
volume, taking turns in conversation, topic in- 2. Verbal decoding of nonverbal behaviors of
troduction, and maintenance. others (e.g., the meaning of social and af-
The interrelationship between language and fective cues, posture, voice patterns).
social ability makes it helpful to interweave 3. Processing of visual information simulta-
communicative training with interventions to neously with auditory information (to fos-
specifically address social skills. A general ter integration of competing stimuli and
principle for teaching social skills is to explic- facilitate the creation of the appropriate so-
itly define rules for social behavior; individu- cial context of the interaction).
als with AS lack the intuition and social 4. Social and self-awareness.
wherewithal to detect and respond to social in- 5. Social norms and expectations.
formation effectively, and this strategy helps 6. Perspective-taking skills.
them to compensate for these deficits. On an 7. Correct interpretation of ambiguous com-
individual basis, teaching explicit scripts that munications (e.g., nonliteral language).
can be used in social situations (e.g., ap-
proaching a peer) capitalizes on strong rote All of these should be taught, practiced, and
memory abilities; these scripts can then be re- fostered in multiple environments.
hearsed in role play. For children requiring
more basic social interventions, practicing ex- Assistive Technology
pression and recognition of facial expressions
is useful; in students with higher levels of abil- Recent advances in computer technology and
ity, this may involve teaching them to read the expansion of the Internet offer ever-
more subtle social cues. Many aids, including expanding opportunities for individuals with
software programs, have been developed to fa- ASDs. The intellectual capabilities of people
cilitate such goals; in this domain and, more with AS render them viable computer users,
generally, it is helpful to move from static to and, for many individuals, computer-based
dynamic. For example, children may practice work is intrinsically very rewarding. Comput-
recognizing affect in pictures or cartoons, ers are now considered useful in numerous as-
subsequently moving on to videotapes, then fi- pects of intervention. Given the frequency of
nally during in vivo role modeling. A number graphomotor problems, typing academic as-
of excellent guidelines have been published to signments or employment-related tasks allevi-
guide parents and interventionists in teaching ates much of the difficulty associated with
social skills. manual penmanship. Organizational software
Social skills training groups are an effec- and personal data assistants can be immensely
tive and increasingly common intervention useful in helping individuals with schedules,
for working on pragmatics and social abili- to-do lists, and a variety of other organiza-
ties. Such groups involve a small number of tional tasks that are very challenging for those
children with social impairments (they may with executive dysfunction (Ozonoff, 1998).
116 Diagnosis and Classification

The Internet is an excellent medium for par- empathy, social difficulties, and depressive
ents and educators to communicate regarding symptoms, and a more direct, problem-solving
academic goals and behavior management, focus can at times be more beneficial than an
thereby increasing consistency across people insight-oriented approach.
and settings, which is so vital to maximize the
effectiveness of interventions for individuals Pharmacotherapy
with AS. Finally, the computer skills acquired
through these activities may, and often do, be- No medications have been shown to decrease
come a valuable asset in the context of voca- the core social and communicative symptoma-
tional opportunities later in life. tology of individuals with ASDs. However,
pharmacological approaches, typically com-
Vocational Training bined with behavioral interventions and appro-
priate adaptations in the child’s environment,
Adults with AS often fail to meet entry re- can be very helpful in allaying comorbid
quirements for jobs in their area of training or symptoms involving anxiety, depression, inat-
fail to maintain a job because of their poor tention, and others, thus making the child
interview skills, social disabilities, eccentrici- more available to educational and therapeutic
ties, anxiety attacks, and, at times, poor activities. Physicians prescribing medications
grooming. All of these areas may require ap- should be knowledgeable not only of the co-
propriate interventions and preparation prior morbid conditions that they are treating but
to challenging events. To ensure success, it is also, and especially, of the nature of the ASDs.
also important that they are trained for and are Multifaceted symptoms should be treated with
placed in jobs for which they are not neuropsy- a minimal number of agents, and planning of
chologically impaired and in which they will pharmacological intervention should include
enjoy a certain degree of support and shelter. clear definitions of targeted symptoms and
It is also preferable that the job does not in- evaluative procedures to ensure that the med-
volve intensive social demands, time pres- ication is having a positive impact on the
sures, or the need to quickly improvise or child’s functioning and well-being (Towbin,
generate solutions to novel situations. 2003). Continued follow-up is required to ad-
just treatment (e.g., to slowly reach therapeu-
Self-Support tic levels), evaluate effects, and possibly
discontinue or change the intervention if there
Because individuals with AS are usually self- are prohibitive side effects or if there is no
described loners despite an often intense wish demonstrated benefit.
to make friends and have a more active social
life, there is a need to facilitate social contact CONCLUSION
within the context of an activity-oriented
group (e.g., church communities, hobby clubs, Since the previous edition of the Handbook in
and self-support groups). The little experience 1997, the major positive changes involving the
available with the latter suggests that individu- diagnostic construct of AS occurred in areas
als with AS enjoy the opportunity to meet oth- other than progress in clinical research. First,
ers with similar problems and may develop there has been an ever-increasing awareness by
relationships around an activity or subject of the general public and authorities deciding on
shared interest. Internet support groups and supportive services of the needs and strengths
chat rooms have proved to be excellent forums of individuals with this condition. Lorna Wing’s
for individuals with AS and their families to influential introduction of the term in 1981 has,
make contact with others who may share an in- 2 decades later, accomplished its stated goal of
terest in AS or may simply partake in a com- bringing individuals with severe social disabili-
mon enjoyed activity. For many individuals ties accompanied by preserved cognitive and
with AS, a strategy that may help in the man- language skills to the attention of the clinical
agement of an individual’s everyday needs is to community and educational systems. And al-
maintain long-term supportive contact with a though misperceptions remain, there is a gen-
counselor. Therapy can focus on problems of eral consensus that these individuals require
Asperger Syndrome 117

special services without which they cannot Two general lines of research may be adopted
fulfill their potential and live more indepen- to elucidate this issue.
dent and fulfilling lives. Second, with the for- First, if we consider the possibility that the
malization of the disorder and the entitlements two conditions are the same or simply variable
associated with it, there has been a great need manifestations of the same underlying con-
for practical resources addressing these indi- struct—the ASDs—then there is a need to
viduals’ needs for social and communication clarify the dimensions creating this spectrum.
skills training and other areas of intervention With the exception of IQ and language skills,
throughout their life span. In response to this there has not been much progress in elucidat-
need, there has been an exponential growth in ing the potential factors mediating expression
the past few years of publications meant to set of the syndrome and determining eventual
guidelines and strategies for teaching skills outcomes. To this goal, there is a need for re-
necessary for social adaptation, learning, self- search relating, in a quantified manner, spe-
regulation, and vocational success. Most of cific neuropsychological, social-cognitive, or
these strategies have not been systematically neurobiological models to phenotypic expres-
examined, and indeed, this is a great priority sion including measures of outcome such as
for the next decade. levels of symptomatology and of ability or
In contrast to these developments, clinical adaptation. All ASDs have their onset in the
research has been hindered by lack of progress early years of a child’s life. Longitudinal re-
as to the validity status of AS. Much of this search is needed in which key developmental
research has been flawed due in great degree processes hypothesized to mediate phenotypic
to a lack of a consensual, validated, and well- expression are evaluated relative to the range
operationalized definition of the condition. of phenotypic variability later in life. The
While the question of whether AS can be dif- rather vague onset criteria established for the
ferentiated from autism without intellectual ASDs could hardly be expected to elucidate
disabilities may have originated from the the thorny diagnostic issues presented by these
chance referrals to Leo Kanner and Hans conditions later in life. And there are still very
Asperger, it has endured for more substantial few studies that have attempted to examine
reasons. The ASDs refer to a highly heteroge- prospectively the early years of individuals
neous phenotype, which nevertheless is pri- later diagnosed as having AS. As these individ-
marily defined by the pronounced and early uals typically come to the attention of special-
emerging social disability. Given the complex- ized clinics later in childhood (given the
ity of social behavior, it is not unreasonable to absence of the more common reasons for refer-
expect that there may be more than one devel- ral such as language delays and social isola-
opmental pathway generating the variable tion), most of what we know about the onset of
manifestations of these conditions. Therefore, this condition is based on retrospective infor-
the successful identification of more homoge- mation. Yet, the formal definitions of AS de-
neous subtypes can potentially be of great pend entirely on onset patterns as a way of
benefit to behavioral and neurobiological re- differentiating it from autism. In summary,
search seeking to yield more effective treat- there is a need for much greater knowledge
ment strategies and to elucidate the causative about the eventual impact on phenopic expres-
processes underlying the etiologies of these sion of variables ranging from social motiva-
conditions. In the face of this challenge, how- tion to joint attention skills, from sensory and
ever, research studies have failed to produce psychomotor adaptation and self-regulation to
the level of substantive detail in phenotypic executive abilities, and from different learning
characterization and of documentation of lon- styles to different language profiles to gain
gitudinal outcomes required for true progress more knowledge of different pathways to so-
in the understanding of developmental mecha- cial disabilities.
nisms mediating syndrome expression. It is Second, if we consider the possibility that
premature to close the door on this debate be- AS and autism are different conditions, then
cause the data are still limited as to whether there is a need to clarify what is meant by the
the ASDs, and autism and AS in particular, word dif ferent. There are various practical
should be considered the same or different. reasons that a person may justifiably choose
118 Diagnosis and Classification

to see two conditions as different (Szatmari, Cross-References

2000), but the most important ones concern
the fact that they may have different etiologies Other diagnostic concepts are discussed in
or pathogenetic pathways (e.g., genetic mech- Chapters 1, 3, 5, and 6; outcome studies in
anisms), they may respond differently to dif- Chapter 7; and assessment issues in Chapters
ferent kinds of treatment (e.g., because of 27 through 32.
different neurocognitive, social-cognitive, or
neurobiological profiles), or they may imply REFERENCES
differential outcomes (which would signify
more detailed guidelines for intervention to American Psychiatric Association. (1987). Diag-
better address the eventual challenges and nostic and statistical manual of mental disor-
ders (3rd ed., rev.). Washington, DC: Author.
maximize factors promoting eventual success).
American Psychiatric Association. (1994). Diag-
None of these important areas of research can, nostic and statistical manual of mental disor-
however, be dramatically advanced without a ders (4th ed.). Washington, DC: Author.
much more detailed, standardized, and quanti- American Psychiatric Association. (2000). Diagnos-
fied description of the two conditions. There tic and statistical manual of mental disorders
is a great deal of knowledge about autism in (4th ed., text rev.). Washington, DC: Author.
general, but much less about the group of Anneren, G., Dahl, N., Uddenfeldt, U., & Janols,
higher functioning individuals, most of whom L. O. (1995). Asperger syndrome in a boy
are seen in clinics at a later age for the same with a balanced de novo translocation:
reasons that individuals with AS are also not T(17;19) (p13.3;p11) [letter]. American Jour-
seen in early childhood. As suggested earlier, nal of Medical Genetics, 56(3), 330–331.
Asperger, H. (1944). Die “Autistischen Psychopa-
given the current status of the diagnostic pro-
then” im Kindesalter. Archiv für Psychiatrie
cess feeding into research studies of AS, it is
und Nervenkrankheiten, 117, 76–136.
unlikely that progress will be made without a Asperger, H. (1979). Problems of infantile autism.
concerted multisite effort adopting rigorous Communication, 13, 45–52.
diagnostic procedures. These should include Attwood, T. (1998). Asperger’s syndrome: A guide
new and especially developed diagnostic in- for parents and professionals. Philadelphia:
strumentation operationalized around a con- Jessica Kingsley.
sensual set of defining criteria. Bailey, A., Le Couteur, A., Gottesman, I., Bolton,
Finally, interest in AS has opened the door P., Simonoff, E., Yuzda, E., et al. (1995).
to debates that had been until recently very Autism as a strongly genetic disorder: Evi-
limited. Broader attention to individuals with dence from a British twin study. Psychological
Medicine, 25, 63–77.
AS who are rather eccentric and socially awk-
Bailey, A., Palferman, S., Heavey, L., & Le Cou-
ward but who may also have special talents in
teur, A. (1998). Autism: The phenotype in rel-
domains of factual knowledge or technological atives. Journal of Autism and Developmental
skills raises questions about the borderlands of Disorders, 28, 369–392.
this condition. For example, to what extent Baird, G., Charman, T., Baron-Cohen, S., Cox, A.,
should we stretch the spectrum to encompass a Swettenham, J., Wheelwright, S., et al. (2000).
much larger group of more successful and not A screening instrument for autism at 18
necessarily disabled people whose social in- months of age: A 6 year follow-up study. Jour-
eptness has had, nevertheless, great impact nal of the American Academy of Child and
over their internal, social, and affective lives Adolescent Psychiatry, 39(6), 694–702.
(Asperger, 1954, as cited in Bosch, 1970; Baron-Cohen, S. (1988). An assessment of violence
Baron-Cohen, 2002)? Whether an approach in a young man with Asperger’s syndrome.
Journal of Child Psychology and Psychiatry,
taking this larger view of the phenotype, larger
29, 351–360.
in fact than the commonly accepted broader
Baron-Cohen, S. (2002). The extreme male brain
autism phenotype, might contribute to a better theory of autism. Trends in Cognitive Sciences,
understanding of genetic mechanisms needs to 6, 248–254.
be answered. But to do so, there will be a need Baron-Cohen, S., Cox, A., Baird, G., Swettenham,
for increasingly more refined phenotypic mea- J., & Nightingale, N. (1996). Psychological
sures of sociability. markers in the detection of autism in infancy
Asperger Syndrome 119

in a large population. British Journal of Psy- atry. New York: Dover. (Original work pub-
chiatry, 168(2), 158–163. lished 1916)
Baron-Cohen, S., Tager-Flusberg, H., & Cohen, Bosch, G. (1970). Infantile autism. New York:
D. J. (Eds.). (1999). Understanding other Springer-Verlag.
minds: Perspectives from autism (2nd ed.). Ox- Bowler, D. M. (1992). “ Theory of Mind” in As-
ford, England: Oxford University Press. perger’s syndrome. Journal of Child Psychol-
Baron-Cohen, S., Wheelwright, S., & Jolliffee, T. ogy and Psychiatry, 33(5), 877–893.
(1997). Is there a “language of the eyes”? Evi- Bowman, E. P. (1988). Asperger’s syndrome and
dence from normal adults and adults with autism: The case for a connection. British
autism or Asperger syndrome. Visual Cogni- Journal of Psychiatry, 152, 377–382.
tion, 4, 311–331. Burgoine, E., & Wing, L. (1983). Identical triplets
Baron-Cohen, S., Wheelwright, S., Stone, V., & with Asperger’s syndrome. British Journal of
Rutherford, M. (1999). A mathematician, a Psychiatry, 143, 261–265.
physicist and a computer scientist with As- Casanova, M., Buxhoeveden, D. P., Switala, A. E., &
perger syndrome: Performance on folk psy- Roy, E. (2002, February). Minocolumnar pathol-
chology and folk physics tests. Neurocase, ogy in autism. Neurology, 58(3), 428–432.
5(6), 475–483. Charman, T., Swettenham, J., Baron-Cohen, S.,
Bartolucci, G., & Szatmari, P. (1987). Possible Cox, A., Baird, G., & Drew, A. (1997). Infants
similarities between the fragile, X., and As- with autism: An investigation of empathy, pre-
perger’s syndrome [letter]. American Journal tend play, joint attention, and imitation. Devel-
of Disorders of Childhood, 141(6), 601–602. opmental Psychology, 33(5), 781–789.
Bashe, P. R., & Kirby, B. L. (2001). The OASIS Constantino, J., & Todd, R. D. (2003). Autistic
guide to Asperger syndrome. New York: traits in the general population: A twin study.
Crown. Archives of General Psychiatry, 60(5),
Bejerot, S., & Duvner, T. (1995). Asperger’s syn- 524–530.
drome or schizophrenia? Nordic Journal of Cox, A., Klein, K., Charman, T., Baird, G., Baron-
Psychiatry, 49(2), 145. Cohen, S., Swettenham, J., et al. (1999).
Berthier, M. L., Bayes, A., & Tolosa, E. S. (1993). Autism spectrum disorders at 20 and 42
Magnetic resonance imaging in patients with months of age: Stability of clinical and ADI-R
concurrent Tourette’s disorder and Asperger’s diagnosis. Journal of Child Psychology and
syndrome. Journal of the American Academy of Psychiatry, 40(5), 719–732.
Child and Adolescent Psychiatry, 32(3), Dahlgren, S. O., & Trillingsgaard, A. (1996). The-
633–639. ory of mind in non-retarded children with
Berthier, M. L., Starkstein, S. E., & Leiguarda, R. autism and Asperger’s syndrome. A research
(1990). Developmental cortical anomalies in note. Journal of Child Psychology and Psychia-
Asperger’s syndrome: Neuroradiological try, 37, 759–763.
findings in two patients. Journal of Neuropsy- DeLong, G. R., & Dwyer, J. T. (1988). Correlation
chiatry and Clinical Neurosciences, 2(2), of family history with specific autistic sub-
197–201. groups: Asperger’s syndrome and bipolar af-
Bishop, D. V. M. (1989). Autism, Asperger’s syn- fective disease. Journal of Autism and
drome and semantic-pragmatic disorder: Developmental Disorders, 18(4), 593–600.
Where are the boundaries? British Journal of Denckla, M. B. (1983). The neuropsychology of
Disorders of Communications, 24, 107–121. social-emotional learning disabilities. Archives
Bishop, D. V. M. (1998). Development of the Chil- of Neurology, 40, 461–462.
dren’s Communication Checklist (CCC): A Ehlers, S., & Gillberg, C. (1993). The epidemiol-
method for assessing qualitative aspects of ogy of Asperger syndrome: A total population
communicative impairment in children. Jour- study. Journal of Child Psychology and Psychi-
nal of Child Psychology and Psychiatry, 39(6), atry, 34(8), 1327–1350.
879–891. Ehlers, S., Nyden, A., Gillberg, C., & Dahlgren
Bishop, D. V. M. (2000). What’s so special about Sandberg, A. (1997). Asperger syndrome,
Asperger syndrome? The need for further ex- autism and attention disorders: A comparative
ploration of the borderlands of autism. In A. study of the cognitive profiles of 120 children.
Klin, F. R. Volkmar, & S. S. Sparrow (Eds.), Journal of Child Psychology and Psychiatry, 2,
Asperger syndrome (pp. 278–308). New York: 207–217.
Guilford Press. Eisenmajer, R., Prior, M., Leekam, S., Wing, L.,
Bleuler, E. (1951). Lehrbuch der psychiatrie. In Gould, J., Welham, M., et al. (1996). Com-
A. A. Brill (Ed. & Trans.), Textbook of psychi- parison of clinical symptoms in autism and
120 Diagnosis and Classification

Asperger’s disorder. Journal of the American Ghaziuddin, M., Tsai, L. Y., & Ghaziuddin, N.
Academy of Child and Adolescent Psychiatry, (1991). Brief report: Violence in Asperger
35, 1523–1531. Syndrome, a critique. Journal of Autism
Ellis, H. D., Ellis, D. M., Fraser, W., & Deb, S. and Developmental Disorders, 21(3),
(1994). A preliminary study of right hemi- 349–354.
sphere cognitive deficits and impaired social Ghaziuddin, M., Tsai, L. Y., & Ghaziuddin, N.
judgments among young people with Asperger (1992). Brief report: A comparison of the di-
syndrome. European Child and Adolescent Psy- agnostic criteria for Asperger Syndrome. Jour-
chiatry, 3(4), 255–266. nal of Autism and Developmental Disorders,
Everall, I. P., & Le Couteur, A. (1990). Firesetting 22(4), 643–649.
in an adolescent with Asperger’s Syndrome. Ghaziuddin, M., Weidmer-Mikhail, E., & Ghaziud-
British Journal of Psychiatry, 157, 284–287. din, N. (1998). Comorbidity of Asperger
Fombonne, E., & Tidmarsh, L. (2003). Epidemio- syndrome: A preliminary report. Journal of In-
logic data on Asperger disorder. Child and tellectual Disability Research 42(4), 279–283.
Adolescent Psychiatric Clinics of North Amer- Gilchrist, A., Green, J., Cox, A., Burton, D., Rut-
ica, 12, 15–22. ter, M., & Le Couteur, A. (2001). Development
Frith, U. (Ed.). (1991). Autism and Asperger syn- and current functioning in adolescents with
drome. Cambridge, England: Cambridge Uni- Asperger syndrome: A comparative study.
versity Press. Journal of Child Psychology and Psychiatry,
Frith, U. (2004). Emanuel Miller lecture: Confu- 42, 227–240.
sions and controversies about Asperger syn- Gillberg, C. (1985). Asperger’s syndrome and re-
drome. Journal of Child Psychology and current psychosis: A case study. Journal of
Psychiatry, 45(4), 672–686. Autism and Developmental Disorders, 15(4),
Fujikawa, H., Kobayashi, R., Koga, Y., & Murata, 389–397.
T. (1987). A case of Asperger’s syndrome in a Gillberg, C. (1989). Asperger syndrome in 23
nineteen-year-old who showed psychotic Swedish children. Developmental Medicine
breakdown with depressive state and at- and Child Neurology, 31, 520–531.
tempted suicide after entering university. Gillberg, C. (1990). Autism and the pervasive de-
Japanese Journal of Child and Adolescent Psy- velopmental disorders. Journal of Child Psy-
chiatry, 28(4), 217–225. chology and Psychiatry, 31(1), 99–119.
Ghaziuddin, M. (2002). Asperger syndrome: Asso- Gillberg, C. (1991). Outcome in autism and autistic-
ciated psychiatric and medical conditions. like conditions. Journal of the American Acad-
Focus on Autism and Other Developmental emy of Child and Adolescent Psychiatry, 30(3),
Disabilities, 17(3), 138–144. 375–382.
Ghaziuddin, M., Butler, E., Tsai, L., & Ghaziuddin, Gillberg, C. (1998). Asperger syndrome and high-
N. (1994). Is clumsiness a marker for Asperger functioning autism. British Journal of Psychia-
Syndrome? Journal of Intellectual Disabilities try, 172, 200–209.
Research, 38(5), 519–527. Gillberg, C., Gillberg, I. C., Rastam, M., & Wentz,
Ghaziuddin, M., & Gerstein, L. (1996). Pedantic E. (2001). The Asperger Syndrome (and high-
speaking style differentiates Asperger syn- functioning autism) Diagnostic Interview
drome from high-functioning autism. Journal (ASDI ): A preliminary study of a new struc-
of Autism and Developmental Disorders, 26(6), tured clinical interview. Autism: Journal of
585–595. Research and Practice, 5(1), 57–66.
Ghaziuddin, M., Ghaziuddin, N., & Greden, J. Gillberg, C., Gillberg, I. C., & Steffenburg, S.
(2002). Depression in persons with autism: (1992). Siblings and parents of children with
Implications for research and clinical care. autism: A controlled population-based study.
Journal of Autism and Developmental Disor- Developmental Medicine and Child Neurology,
ders, 32(4), 299–306. 34(5), 389–398.
Ghaziuddin, M., Leininger, L., & Tsai, L. (1995). Gillberg, C., & Rastam, M. (1992). Do some cases
Brief report: Thought disorder in Asperger of anorexia nervosa reflect underlying autistic-
syndrome [Comparison with high-functioning like. Behavioural Neurology, 5(1), 27–32.
autism]. Journal of Autism and Developmental Gillberg, I. C., & Gillberg, C. (1989). Asperger
Disorders, 25(3), 311–317. syndrome: Some epidemiological considera-
Ghaziuddin, M., & Mountain-Kimchi, K. (2004). tions. Journal of Child Psychology and Psychi-
Defining the intellectual profile of Asperger atry, 30, 631–638.
syndrome: Comparison with high-functioning Green, J., Gilchrist, A., Burton, D., & Cox, A.
autism. Journal of Autism and Developmental (2000). Social and psychiatric functioning in
Disorders, 34(3), 279–284. adolescents with Asperger syndrome com-
Asperger Syndrome 121

pared with conduct disorder. Journal of Autism Kanner, L. (1954). Discussion of Robinson & Vi-
and Developmental Disorders, 30, 279–293. tale’s paper on “Children with circumscribed
Grossman, J. B., Klin, A., Carter, A. S., & Volk- interests.” American Journal of Orthopsychia-
mar, F. R. (2000). Verbal bias in recognition of try, 24, 764–766.
facial emotions in children with Asperger syn- Kerbeshian, J., & Burd, L. (1986). Asperger’s syn-
drome. Journal of Child Psychology and Psy- drome and Tourette syndrome: The case of the
chiatry, 41(3), 369–379. pinball wizard. British Journal of Psychiatry,
Gunter, H. L., Ghaziuddin, M., & Ellis, H. D. 148, 731–736.
(2002). Asperger syndrome: Tests of right Kim, J., Szatmari, P., Bryson, S., Streiner, D. L., &
hemisphere functioning and interhemispheric Wilson, F. J. (2000). The prevalence of anxiety
communication. Journal of autism and Devel- and mood problems among children with
opmental Disorders, 32(4), 263–281. autism and Asperger syndrome. Autism: Jour-
Happé, F. G. (1995). The role of age and verbal nal of Research and Practice, 4, 117–132.
ability in the theory of mind task performance Klin, A. (1994). Asperger syndrome. Child and
of subjects with autism. Child Development, Adolescent Psychiatry Clinic of North Amer-
66(3), 843–855. ica, 3, 131–148.
Happé, F., Ehlers, S., Fletcher, P., Frith, U., Johans- Klin, A. (2000). Attributing social meaning to am-
son, M., Gillberg, C., et al. (1996). “ Theory of biguous visual stimuli in higher-functioning
mind” in the brain: Evidence from a PET scan autism and Asperger syndrome: The Social At-
study of Asperger syndrome. NeuroReport, tribution Task. Journal of Child Psychology and
8(1), 197–201. Psychiatry, 41(7), 831–846.
Hippler, K., & Klicpera, C. (2003). A retrospective Klin, A., Carter, A., & Sparrow, S. S. (1997). Psycho-
analysis of the clinical case records of “autis- logical assessment of children with autism. In
tic psychopaths” diagnosed by Hans Asperger D. J. Cohen & F. R. Volkmar (Eds.), Handbook
and his team at the University Children’s Hos- of autism and pervasive developmental disorders
pital, Vienna. [Philosophical Transactions of (2nd ed., pp. 418–427). New York: Wiley.
the Royal Society]. Biological Sciences, 358, Klin, A., Jones, W., Schultz, R. T., & Volkmar,
291–301. F. R. (2003). The enactive mind: From actions
Howlin, P. (1997). Prognosis in autism: Do specialist to cognition—lessons from autism. [Philo-
treatments affect long-term outcome. European sophical Transactions of the Royal Society].
Child and Adolescent Psychiatry, 6(2), 55–72. Biological Sciences, 358, 345–360.
Howlin, P. (1999). Children with autism and As- Klin, A., Jones, W., Schultz, R., Volkmar, F., &
perger syndrome: A guide for practitioners and Cohen, D. (2002). Defining and quantifying
carers. New York: Wiley. the social phenotype in autism. American
Howlin, P. (2003). Outcome in high-functioning Journal of Psychiatry, 159(6), 895–908.
adults with autism with and without early lan- Klin, A., Pauls, D., Schultz, R., & Volkmar, F.
guage delays: Implications for the differentia- (in press). Three diagnostic approaches to
tion between autism and Asperger syndrome. Asperger syndrome: Implications for re-
Journal of Autism and Developmental Disor- search. Journal of Autism and Developmental
ders, 33, 3–13. Disorders.
Howlin, P., & Goode, S. (1998). Outcome in adult life Klin, A., Sparrow, S. S., Marans, W. D., Carter,
for people with autism and Asperger’s syndrome. A., & Volkmar, F. R. (2000). Assessment is-
In F. R. Volkmar (Ed.), Autism and pervasive de- sues in children and adolescents with As-
velopmental disorders (pp. 209–241). Cambridge, perger syndrome. In A. Klin, F. R. Volkmar, &
England: Cambridge University Press. S. S. Sparrow (Eds.), Asperger syndrome
Johnson, D. J., & Myklebust, H. R. (1971). Learn- (pp. 309–339). New York: Guilford Press.
ing disabilities. New York: Grune & Stratton. Klin, A., Sparrow, S. S., Volkmar, F. R., Cicchetti,
Jones, P. B., & Kerwin, R. W. (1990). Left tempo- D. V., & Rourke, B. P. (1995). Asperger syn-
ral lobe damage in Asperger’s syndrome. drome. In B. P. Rourke (Ed.), Syndrome of
British Journal of Psychiatry, 156, 570–572. nonverbal learning disabilities: Neurodevelop-
Kaland, N., Moller-Nielsen, A., Callesen, K., mental manifestations (pp. 93–118). New
Mortensen, E. L., Gottlieb, D., & Smith, L. York: Guilford Press.
(2002, May). A new “advanced” test of theory Klin, A., & Volkmar, F. R. (1996). The pervasive
of mind: Evidence from children and adoles- developmental disorders: Nosology and pro-
cents with Asperger syndrome. Journal of Child files of development. In S. Luthar, J. Burack,
Psychology and Psychiatry, 43(4), 517–528. D. Cicchetti, & J. Wiesz (Eds.), Developmental
Kanner, L. (1943). Autistic disturbances of affec- perspectives on risk and psychopathology. New
tive contact. Nervous Child, 2, 217–253. York: Cambridge University Press.
122 Diagnosis and Classification

Klin, A., & Volkmar, F. R. (1997). Asperger syn- Mawson, D., Grounds, A., & Tantam, D. (1985).
drome. In D. J. Cohen & F. R. Volkmar (Eds.), Violence and Asperger’s syndrome: A case
Handbook of autism and pervasive develop- study. British Journal of Psychiatry, 147,
mental disorders (pp. 94–122). New York: 566–569.
Wiley. Mayes, S. D., Calhoun, S. L., & Crites, D. L.
Klin, A., & Volkmar, F. R. (Eds.). (2003). Asperger (2001). Does DSM-IV Asperger’s disorder
syndrome. Child and adolescent psychiatric clin- exist? Journal of Abnormal Child Psychology,
ics of North America. Philadelphia: Saunders. 29, 263–271.
Klin, A., Volkmar F. R., & Sparrow, S. S. (Eds.). McAlonan, G., Daly, E., Kumari, V., Critchley,
(2000). Asperger syndrome. New York: Guil- H. D., van Amelsvoort, T., Suckling, J., et al.
ford Press. (2002). Brain anatomy and sensorimotor gat-
Klin, A., Volkmar, F. R., Sparrow, S. S., Cicchetti, ing in Asperger’s syndrome. Brain, 125(7),
D. V., & Rourke, B. P. (1995). Validity and 1594–1606.
neuropsychological characterization of As- McKelvey, J. R., Lambert, R., Mottron, L., & Shev-
perger syndrome. Journal of Child Psychology ell, M. I. (1995). Right-hemisphere dysfunc-
and Psychiatry, 36(7), 1127–1140. tion in Asperger’s syndrome. Journal of Child
Le Couteur, A., Bailey, A., Goode, S., Pickles, A., Neurology, 10(4), 310–314.
Robertson, S., Gottesman, I., et al. (1996). A Mesibov, G. B. (1992). Treatment issues with high-
broader phenotype of autism: The Clinical functioning adolescents and adults with
spectrum in twins. Journal of Child Psychol- autism. In E. Schopler & G. B. Mesibov (Eds.),
ogy and Psychiatry, 37(7), 785–801. High-functioning individuals with autism
Leekam, S., Libby, S., Wing, L., Gould, J., & Gill- (pp. 143–156). New York: Plenum Press.
berg, C. (2000). Comparison of ICD-10 and Miles, S. W., & Capelle, P. (1987). Asperger’s syn-
Gillberg’s criteria for Asperger syndrome. drome and aminoaciduria: A case example.
Autism: Journal of Research and Practice, 4, British Journal of Psychiatry, 150, 397–400.
11–28. Miller, J., & Ozonoff, S. (2000). The external va-
Lincoln, A., Courchesne, E., Allen, M., Hanson, E., lidity of Asperger disorder: Lack of evidence
& Ene, M. (1998). Neurobiology of Asperger from the domain of neuropsychology. Journal
syndrome: Seven case studies and quantitative of Abnormal Psychology, 109, 227–238.
magnetic resonance imaging findings. In Murphy, D. G. M., Critchley, H. D., Schmitz, N.,
E. Schopler, G. Mesibov, & L. J. Kunce (Eds.), McAlonan, G., van Amelsvoort, T., Robert-
Asperger syndrome or high-functioning autism? son, D., et al. (2002). Asperger syndrome: A
(pp. 145–166). New York: Plenum Press. proton magnetic resonance spectroscopy study
Lincoln, A. J., Allen, M., & Kilman, A. (1995). The of brain. Archives of General Psychiatry,
assessment and interpretation of intellectual 59(10), 885–892.
abilities in people with autism. In E. Schopler Myklebust, H. R. (1975). Nonverbal learning dis-
& G. Mesibov (Eds.), Learning and cognition abilities: Assessment and intervention. In
in autism (pp. 89–117). New York: Plenum H. R. Myklebust (Ed.), Progress in learning
Press. disabilities (Vol. 3, pp. 85–121). New York:
Littlejohns, C. S., Clarke, D. J., & Corbett, J. A. Grune & Stratton.
(1990). Tourette-like disorder in Asperger’s Myles, B. S., & Simpson, R. (1997). Asperger syn-
syndrome. British Journal of Psychiatry, 156, drome: A guide for educators and parents.
430–433. Austin, TX: ProEd.
Lord, C., Rutter, M. J., DiLavore, P., & Risi, S. Nagy, J., & Szatmari, P. (1986). A chart review of
(1999). Autism diagnostic observation schedule. schizotypal personality disorders in children.
Los Angeles: Western Psychological Services. Journal of Autism and Developmental Disor-
Lotspeich, L., Kwon, H., Schumann, C. M., Fryer, ders, 16(3).
S. L., Goodlin-Jones, B. L., Buonocore, M. H., National Research Council. (2001). Educating chil-
et al. (2004). Investigation of Neuroanatomical dren with autism. Washington, DC: National
Differences Between Autism and Asperger Academies Press.
Syndrome. Archives of General Psychiatry, Newson, E., Dawson, M., & Everaard, T.
61(3), 291–298. (1984–1985). The natural history of able
Marriage, K., Miles, T., Stokes, D., & Davey, M. autistic people: Their management and func-
(1993). Clinical and research implications of tioning in a social context. Communication,
the co-occurrence of Asperger’s and Tourette 19–21.
syndrome. Australian and New Zealander Jour- Ozbayrak, K. R., Kapucu, O., Erdem, E., & Aras,
nal of Psychiatry, 27(4), 666–672. T. (1991). Left occipital hypoperfusion in a
Asperger Syndrome 123

case with Asperger Syndrome. Brain Develop- robehavioral examination of individuals with
ment, 13(6), 454–456. high-functioning and Asperger disorder using
Ozonoff, S. (1998). Assessment and remediation of a fronto-striatal model of dysfunction. Behav-
executive dysfunction in autism and Asperger ioral and Cognitive Neuroscience Reviews,
syndrome. In E. Schopler, G. Mesibov, & L. J. 1(2), 164–177.
Kunce (Eds.), Asperger syndrome or high func- Robinson, J. F., & Vitale, L. J. (1954). Children
tioning autism? (pp. 263–289). New York: with circumscribed interests. American Jour-
Plenum Press. nal of Orthopsychiatry, 24, 755–764.
Ozonoff, S., Dawson, G., & McPartland, J. (2002). A Rourke, B. (1989). Nonverbal learning disabilities:
parent’s guide to Asperger syndrome and high- The syndrome and the model. New York, Guil-
functioning autism. New York: Guilford Press. ford Press.
Ozonoff, S., & Griffith, E. M. (2000). Neuropsycho- Rourke, B. (Ed.). (1995). Syndrome of nonverbal
logical function and the external validity of As- learning disabilities: Manifestations in neuro-
perger syndrome. In A. Klin, F. R. Volkmar, & logical disease disorder and dysfunction. New
S. S. Sparrow (Eds.), Asperger syndrome York: Guilford Press.
(pp. 72–96). New York: Guilford Press. Rourke, B., Young, G. C., & Leenaars, A. A.
Ozonoff, S., Pennington, B. F., & Rogers, S. J. (1989). A childhood learning disability that
(1991). Executive function deficits in high- predisposes those aff licted to adolescent and
functioning autistic individuals: Relationship adult depression and suicide risk. Journal of
to Theory of Mind. Journal of Child Psychol- Learning Disabilities, 22, 169–185.
ogy and Psychiatry, 32(7), 1081–1105. Royers, H., Buysse, A., Ponnet, K., & Pichal, B.
Ozonoff, S., Rogers, S. J., & Pennington, B. F. (2001). Advancing advanced mind-reading
(1991). Asperger’s syndrome: Evidence of an tests: Empathic accuracy in adults with a per-
empirical distinction from high-functioning vasive developmental disorder. Journal of Child
autism. Journal of Child Psychology and Psy- Psychology and Psychiatry, 42(2), 271–278.
chiatry, 32(7), 1107–1122. Rutter, M. (1978). Diagnosis and definition of
Perner, J., & Wimmer, H. (1985). “John thinks that childhood autism. Journal of Autism and
Mary thinks that . . .”: Attribution of second- Childhood Schizophrenia, 8(2), 139–161.
order beliefs by 5–10 year old children. Jour- Rutter, M. (1999). The Emmanual Miller Memorial
nal of Experimental Child Psychology, 39, Lecture 1998: Autism: Two-way interplay be-
437–471. tween research and clinical world. Journal of
Piven, J., Nehme, E., Simon, J., Barta, P., Pearlson, Child Psychology and Psychiatry, 40, 169–188.
G., & Folstein, S. E. (1992). Magnetic reso- Rutter, M., Bailey, A., Bolton, P., & Le Couteur, A.
nance imaging in autism: Measurement of the (1994). Autism and known medical conditions:
cerebellum, pons, and fourth ventricle. Biolog- Myth and substance. Journal of Child Psychol-
ical Psychiatry, 31, 491–504. ogy and Psychiatry, 35(2), 311–322.
Piven, J., Palmer, P., Jacobi, D., Childress, D., & Rutter, M., Bailey, A., Simonoff, E., & Pickles, A.
Arndt, S. (1997). The broader autism pheno- (1997). Genetic inf luences in autism. In D. J.
type: Evidence from a family study of multiple- Cohen & F. R. Volkmar (Eds.), Handbook of
incidence autism families. American Journal autism and pervasive developmental disorders
of Psychiatry, 154, 185–190. (2nd ed., pp. 370–387). New York: Wiley.
Piven, J., Saliba, K., Bailey, J., & Arndt S. (1997). Rutter, M., & Gould, M. (1985). Classification. In
An MRI study of autism: The cerebellum re- M. Rutter & L. Herson (Eds.), Child and
visited. Neurology, 49(2), 546–551. adolescent psychiatry: Modern approaches
Powers, M., & Poland, J. (2002). Asperger syn- (2nd ed., pp. 304–321). Oxford, England:
drome and your child: A parent’s guide. New Blackwell.
York: HarperCollins. Rutter, M., Le Couteur, A., & Lord, C. (2003).
Rapin, I., & Allen, D. (1983). Developmental ADI-R autism diagnostic interview (Revised).
language disorders. In U. Kirk (Ed.), Neu- Los Angeles: Western Psychological Services.
ropsychology of language, reading and spelling. Ryan, R. M. (1992). Treatment-resistant chronic
New York: Academic Press. mental illness: Is it Asperger’s syndrome?
Reitzel, J.-A., & Szatmari, P. (2003). Cognitive and Hospital and Community Psychiatry, 43(8),
academic problems. In M. Prior (Ed.), Learn- 807–811.
ing and behavior problems in Asperger syn- Saliba, J. R., & Griffiths, M. (1990). Brief report:
drome (pp. 35–54). New York: Guilford Press. Autism of the Asperger type associated with
Rinehart, N. J., Bradshaw, J. L., Tonge, B. J., Brere- an autosomal fragile site. Journal of Autism
ton, A. V., & Bellgrove, M. A. (2002). A neu- and Developmental Disorders, 20(4), 569–575.
124 Diagnosis and Classification

Scragg, P., & Shah, A. (1994). Prevalence of As- Taiminen, T. (1994). Asperger’s syndrome or
perger’s syndrome in a secure hospital. British schizophrenia: Is differential diagnosis neces-
Journal of Psychiatry, 165(5), 679–682. sary for adult patients? Nordic Journal of Psy-
Siegel, D. J., Minshew, N. J., & Goldstein, G. chiatry, 48(5), 325–328.
(1996). Wechsler IQ profiles in diagnosis of Tantam, D. (1988a). Annotation: Asperger’s syn-
high-functioning autism. Journal of Autism drome. Journal of Child Psychology and Psy-
and Developmental Disorders, 26(4), 389–406. chiatry, 29(3), 245–255.
Smith, I. M. (2000). Motor functioning in As- Tantam, D. (1988b). Lifelong eccentricity and so-
perger syndrome. In A. Klin, F. R. Volkmar, & cial isolation: II. Asperger’s syndrome or
S. S. Sparrow (Eds.), Asperger syndrome schizoid personality disorder? British Journal
(pp. 97–124). New York: Guilford Press. of Psychiatry, 153, 783–791.
Sparrevohn, R., & Howie, P. M. (1995). Theory of Tantam, D. (1988c). Lifelong eccentricity and so-
mind in children with autistic disorder: Evi- cial isolation: I. Psychiatric, social, and foren-
dence of developmental progression and the sic aspects. British Journal of Psychiatry, 153,
role of verbal ability. Journal of Child Psychol- 777–782.
ogy and Psychiatry, 36(2), 249–263. Tantam, D. (1991). Asperger’s syndrome in adult-
Szatmari, P. (2000). Perspectives on the classifica- hood. In U. Frith (Ed.), Autism and Asperger
tion of Asperger syndrome. In A. Klin, F. R. syndrome (pp. 147–183). Cambridge, Eng-
Volkmar, & S. S. Sparrow (Eds.), Asperger land: Cambridge University Press.
syndrome (pp. 403–417). New York: Guilford Tantam, D., Evered, C., & Hersov, L. (1990). As-
Press. perger’s syndrome and ligamentous laxity.
Szatmari, P., Archer, L., Fisman, S., Streiner, Journal of the American Academy of Child and
D. L., & Wilson, F. (1995). Asperger’s syn- Adolescent Psychiatry, 29(6), 892–896.
drome and autism: Differences in behavior, Tantam, D., Evered, C., & Hersov, L. (1990). As-
cognition, and adaptive functioning. Journal of perger’s syndrome and ligamentous laxity.
the American Academy of Child and Adolescent Journal of the American Academy of Child and
Psychiatry, 34, 1662–1671. Adolescent Psychiatry, 29(6), 892–896.
Szatmari, P., Bartolucci, G., & Bremner, R. Tentler, D., Johannesson, T. M., Johansson, M.,
(1989a). Asperger’s syndrome and autism: Rastamm, M., Gillberg, C., Orsmark, C., et al.
Comparison of early history and outcome. (2003). A candidate region for Asperger syn-
Developmental Medicine and Child Neurology, drome defined by two 17p breakpoints. Euro-
31(6), 709–720. pean Journal of Human Genetics, 11(2),
Szatmari, P., Bremner, R., & Nagy, J. N. (1989b). 189–195.
Asperger’s syndrome: A review of clinical fea- Thomsen, P. H. (1994). Obsessive-compulsive dis-
tures. Canadian Journal of Psychiatry, 34(6), order in children and adolescents: A 6–22-year
554–560. follow-up study: Clinical descriptions of the
Szatmari, P., Bryson, S. E., Boyle, M. H., Streiner, course and continuity of obsessive-compulsive
D. L., & Duku, E. (2003). Predictors of out- symptomatology. European Child and Adoles-
come among high functioning children with cent Psychiatry, 3(2), 82–96.
autism and Asperger syndrome. Journal of Towbin, K. E. (2003). Strategies for pharmacologic
Child Psychology and Psychiatry, 44(4), treatment of high-functioning autism and As-
520–528. perger syndrome. Child and Adolescent Psychi-
Szatmari, P., Bryson, S. E., Streiner, D. L., Wilson, atric Clinics of North America, 12(1), 23–45.
F., Archer, L., & Ryerse, C. (2000). Two-year Tsai, L. Y. (2002). Diagnostic issues in high-
outcome of preschool children with autism or functioning autism. In E. Schopler & G. B. Mesi-
Asperger’s syndrome. American Journal of bov (Ed.), High-functioning individuals with
Psychiatry, 157, 1980–1987. autism (pp. 11–40). New York: Plenum Press.
Szatmari, P., Tuff, L., Finlayson, M. A. J., & Bar- Van Bourgondien, M. E., & Woods, A. V. (1992).
tolucci, G. (1990). Asperger’s syndrome and Vocational possibilities for high-functioning
autism: Neurocognitive aspects. Journal of the adults with autism. In E. Schopler & G. B.
American Academy of Child and Adolescent Mesibov (Eds.), High-functioning individuals
Psychiatry, 29, 130–136. with autism (pp. 227–239). New York: Plenum
Tager-Flusberg, H. (2001). A reexamination of the Press.
theory of mind hypothesis of autism. In J. A. Van Krevelen, D. A. (1963). On the relationship
Burack & T. Charman (Eds.), The development between early infantile autism and autistic
of autism: Perspectives from theory and re- psychopathy. Acta Paedopsychiatrica, 30,
search (pp. 173–193). Mahwah, NJ: Erlbaum. 303–323.
Asperger Syndrome 125

Van Krevelen, D. A. (1971). Early infantile autism cognitive components. Archives of Neurology,
and autistic psychopathy. Journal of Autism 40, 463–468.
and Child Schizophrenia, 1(1), 82–86. Wing, L. (1981). Asperger’s syndrome: A clinical
Volkmar, F. R., & Cohen, D. J. (1989). Disintegra- account. Psychological Medicine, 11, 115–129.
tive disorder or “late onset ” autism. Journal of Wing, L. (1986). Clarification on Asperger’s Syn-
Child Psychology and Psychiatry, 30(5), drome (Letters to the Editor). Journal of
717–724. Autism and Developmental Disorders, 16(4),
Volkmar, F. R., & Cohen, D. J. (1991). Comorbid 513–515.
association of autism and schizophrenia. Wing, L. (1998). The history of Asperger syn-
American Journal of Psychiatry, 148(12), drome. In E. Schopler & G. Mesibov (Eds.),
1705–1707. Asperger syndrome or high-functioning
Volkmar, F. R., & Klin, A. (2000). Diagnostic issues autism? (pp. 11–28). New York: Plenum Press.
in Asperger syndrome. In A. Klin, F. R. Volk- Wing, L. (2000). Past and future of research on As-
mar, & S. S. Sparrow (Eds.), Asperger syn- perger syndrome. In A. Klin, F. R. Volkmar, &
drome (pp. 25–71). New York: Guilford Press. S. S. Sparrow (Eds.), Asperger syndrome
Volkmar, F. R., Klin, A., & Cohen, D. J. (1997). Di- (pp. 418–432). New York: Guilford Press.
agnosis and classification of autism and re- Wing, L., & Gould, J. (1979). Severe impairments
lated conditions: Consensus and issues. In of social interaction and associated abnor-
D. J. Cohen & F. R. Volkmar (Eds.), Handbook malities in children: Epidemiology and classi-
of autism and pervasive developmental disor- fication. Journal of Autism and Childhood
ders (pp. 5–40). New York: Wiley. Schizophrenia, 9, 11–29.
Volkmar, F. R., Klin, A., & Pauls, D. (1998). Wolff, S. (1991). “Schizoid” personality in child-
Nosological and genetic aspects of Asperger hood and adult life: III. The childhood picture.
Syndrome. Journal of Autism and Developmen- British Journal of Psychiatry, 159, 629–635.
tal Disorders, 28(5), 457–463. Wolff, S. (1995). Loner’s: The life path of unusual
Volkmar, F. R., Klin, A., Schultz, R. B., Bronen, R., children. London: Routledge.
Marans, W. D., Sparrow, S. S., et al. (1996). Wolff, S., & Barlow, A. (1979). Schizoid personal-
Grand rounds in child psychiatry: Asperger ity in childhood: A comparative study of
syndrome. Journal of the American Academy of schizoid, autistic and normal children. Jour-
Child and Adolescent Psychiatry, 35, 118–123. nal of Child Psychology and Psychiatry, 20,
Volkmar, F. R., Klin, A., Siegel, B., Szatmari, P., 19–46.
Lord, C., Campbell, M., et al. (1994). DSM-IV Wolff, S., & Chick, J. (1980). Schizoid personality
Autism / Pervasive Developmental Disorder in childhood: A controlled follow-up study.
Field Trial. American Journal of Psychiatry, Psychological Medicine, 10, 85–100.
151, 1361–1367. World Health Organization. (1993). International
Volkmar, F. R., Lord, C., Bailey, A., Schultz, R. T., classification of diseases [Chapter V. Mental
& Klin, A. (2004). Autism and pervasive de- and behavioral disorders (including disorders
velopmental disorders. Journal of Child Psy- of psychological development): Diagnostic cri-
chology and Psychiatry, 45(1), 1–36. teria for research] (10th ed.). Geneva,
Weimer, A. K., Schatz, A. M., Lincoln, A., Ballan- Switzerland: Author.
tyne, A. O., & Trauner, D. A. (2001). “Motor ” Ziatas, K., Durkin, K., & Pratt, C. (2003). Differ-
impairment in Asperger syndrome: Evidence ences in assertive speech acts produced by
for a deficit in proprioception. Journal of De- children with autism, Asperger syndrome, spe-
velopmental and Behavioral Pediatrics, 22(2), cific language impairment, and normal devel-
92–101. opment. Development and Psychopathology,
Weintraub, S., & Mesulam, M. M. (1983). Devel- 15(1), 73–94.
opmental learning disabilities of the right
hemisphere: Emotional, interpersonal, and

Rett Syndrome: A Pervasive Developmental Disorder


At least in its classical form, Rett syndrome is with this disorder. This false lead, coupled with
a phenotypically distinct progressive X-linked very limited exposure (Rett, 1969, 1977) of
dominant neurodevelopmental disorder that this information in the English language med-
almost exclusively affects females. The char- ical literature, resulted in a general failure to
acteristic pattern of cognitive and functional recognize Rett syndrome, previously termed
stagnation with subsequent deterioration pro- cerebroatrophic hyperammonemia as a nosolog-
foundly impairs postnatal brain growth and ical entity.
development. Rett syndrome represents one of Unaware of Rett’s work, Bengt Hagberg
the most common causes of mental retardation was working with patients displaying similar
in females, second only to Down syndrome symptoms in Sweden. In 1980, he presented
(Ellaway & Christodoulou, 1999). Stereotypic a paper at the European Federation of Child
hand movements, typically at midline, are one Neurology Societies describing 16 girls he
of the most prominent symptoms. The disorder had observed (Hagberg, 1980). Later, he and
is the first human disease that has been found several colleagues (Hagberg, Aicardi, Dias, &
to result from defects in a protein involved in Ramos, 1983) published in the Annals of Neu-
the regulation of gene expression through its rology their report of 35 girls from France,
interaction with methylated DNA. As such, Portugal, and Sweden with Rett syndrome.
Rett syndrome could hold the key to our under- This landmark account awakened the recogni-
standing of human disorders ranging from tion and interest of clinicians and researchers
some forms of learning disability to autism. and provided credit to Dr. Rett for his pioneer-
Andreas Rett, an Austrian physician, first ing efforts on the disorder that bears his name.
described the disorder following his serendipi- Researchers have learned much about this dis-
tous discovery of two girls seated in his waiting order over the past 2 decades; however, much
room who displayed strikingly similar hand- remains to be discovered.
wringing mannerisms. When he discussed this
with his receptionist, they were able to review CLINICAL PRESENTATION AND
patient records and identify six additional NATURAL HISTORY
patients with similar behavioral characteristics
and developmental histories. Unable to find Individuals with Rett syndrome exhibit a
a known classification for the disorder, Rett unique and characteristic course of develop-
(1966) published a report (in German) de- ment (Naidu, Murphy, Moser, & Rett, 1986).
scribing 22 girls with a syndrome consisting of Prenatal and perinatal histories of these
stereotypic hand movements, dementia, autistic persons are generally unremarkable. Although
behavior, ataxia, cortical atrophy, and hyper- minor pre- and perinatal problems (e.g., mild
ammonemia ( blood ammonia). The reported hypotonia, tremulous neck movements, a low
increased levels of hyperammonemia were sub- intensity of interpersonal contact, and abnor-
sequently found to be only rarely associated mal hand use and language development) can be

Rett Syndrome: A Pervasive Developmental Disorder 127

identified retrospectively in as many as 80% loss of acquired speech, voluntary grasping,

of the girls (Burford, Keer, & Macleod, 2003; and the purposeful use of the hands (Charman
Charman et al., 2002; Hanefeld, 1985; Leonard et al., 2002). The girls begin to lack sustained
& Bower, 1998; Nomura & Segawa, 1990; interest in persons or objects and demonstrate
Opitz & Lewin, 1987; Sekul & Percy, 1992), it limited interpersonal contact; however, they
is unlikely that these mild symptoms would be maintain eye contact (Holm, 1985; Trevathan &
detected as relevant even with detailed neuro- Naidu, 1988; Witt-Engerstrom, 1987). This
logical or developmental assessment. Excess deterioration occurs very quickly, typically
levels of hand patting, waving, and involuntary within 1 year or less, resulting in severe-to-
movements including alternate opening and profound disabilities and stereotyped behaviors.
closing of the fingers, twisting of the wrists and Deceleration of head growth (acquired micro-
arms, or nonspecific circulating hand-mouth cephaly), coarse, jerky movements of the trunk
movements appear to be the most characteristic and limbs, a stiff-legged, broad-based gait with
early warning signals for the syndrome (Holm, somewhat short steps and swaying movements
1985; Kerr, Montague, & Stephenson, 1987; of the shoulders when walking accompany the
Witt-Engerstrom, 1987). Given the wide range developmental deterioration (Coleman & Gill-
of functioning at this age, such soft signs would berg, 1985; Hanefeld, 1985; Kerr & Stephen-
generally be dismissed. Thus, parents generally son, 1986; Naidu et al., 1986; Percy, Zoghbi, &
report normal physical and mental development Riccardi, 1985). With the loss of purposeful
for the first 6 to 8 months of life as evidenced hand movements, the most prominent symptom
by physical growth and psychomotor and verbal of the syndrome appears, in the form of stereo-
behavior (Gillberg, 1987; Sekul & Percy, 1992). typic hand-to-mouth movements, hand clasp-
This apparently normal period of development ing, and “ hand washing” (see Figure 5.1;
is followed by a slowing or cessation of the ac- Ishikawa et al., 1978; Leiber, 1985).
quisition of developmental milestones with sig- The developmental regression seems to
nificant deviations in the acquisition of skills plateau during the early school years. In some
requiring balance (e.g., walking) in many cases. cases, parents report their daughters attempt
By 15 months, approximately half of the girls to increase their functional use of retained
demonstrate serious developmental delays and skills. The girls become more responsive to
abnormal neurological signs or symptoms. By 3 their environment. As children with Rett syn-
years of age, the children have experienced a drome approach adolescence, they are fre-
rapid deterioration of behavior as evidenced by quently subject to increased spasticity and

Figure 5.1 Stereotypic hand movements in Rett syndrome.

128 Diagnosis and Classification

vasomotor disturbances of the lower limbs, Stage 1 are nonspecific and are not predic-
possible loss of existing ambulation, scoliosis, tive of subsequent deterioration.
and a diminished rate of growth. Facial gri- Stage 2 (onset 1 to 4 years): During this
macing, bruxism (teeth grinding), hyperventi- stage, the syndrome becomes significantly
lation, apnea ( breath holding), aerophagia (air more pronounced as the children lose previ-
swallowing), constipation, and seizure activity ously acquired abilities. A relatively well-
also may accompany the syndrome (Trevathan demarcated period of rapidly declining
& Naidu, 1988). social interaction, stagnation or loss of ac-
Hagberg and Witt-Engerstrom (1986) pro- quired cognitive abilities, and loss of pur-
posed a staging system to facilitate the charac- poseful hand use and speech is evident in
terization of the disorder patterns and profiles most cases. Stereotypic movements, often
from infancy through adolescence. Their sys- virtually continuous during waking hours,
tem suggests four clinical stages and was de- become a prominent symptom. The intellec-
rived from a synthesis of clinical observations tual functioning of the girl with Rett syn-
over the years in 50 Swedish cases of Rett syn- drome during this stage is generally reported
drome. The stage patterns provide average to fall within the severe-to-profound range
guidelines for use when confronted with the di- of mental retardation. Ataxic/apraxic gait
agnostic problems resulting from the complex abnormalities are observed in ambulatory
symptomatology and longitudinal profile of girls. Also when the child is awake, she may
the condition. The four stages are Early Onset display aberrant breathing patterns. Hyper-
Stagnation Stage, Rapid Destructive Stage, ventilation and respiratory pauses (gener-
Pseudostationary Stage, and Late Motor De- ally lasting 30 to 40 seconds) are most
generation Stage. This staging system has common.
gained general acceptance; however, the names This deterioration frequently has
assigned to each stage have been criticized been sufficiently dramatic to simulate a
(Opitz, 1986; Trevathan & Naidu, 1988), and toxic or encephalitic state (Hagberg &
the stages are commonly referred to simply by Witt-Engerstrom, 1986). Parents frequently
number. The symptoms and features of each of report that their children seem irritable. Un-
the four stages are as follows: provoked episodes of screaming and sponta-
neous tantrums are common (Coleman et al.,
Stage 1 (onset 6 to 18 months): The clinical 1988). Seizure activity is present in approxi-
profile at this stage suggests a deterioration, mately one-fourth of the girls during this
or at least a general slowing down (stagna- stage. Sleep abnormalities including delayed
tion) of motor development. Hypotonia is sleep onset and increased night awakenings
typically noted. Deviation from normal de- are manifested by more than three-fourths
velopment is often compensated for or hid- of the girls (Piazza, Fisher, Kiesewetter,
den, in part, by the rapid developmental Bowman, & Moser, 1990).
speed of infancy. For example, most of the Stage 3 (onset 2 to 10 years): Persons with
children can sit independently prior to the Rett syndrome generally demonstrate di-
initiation of this stage but many fail to de- minishing autistic symptomatology and im-
velop the subsequent postural skills needed proved social interaction during this
for balance when crawling, standing, and period. They appear to be more aware of
walking. Yet, Sekul and Percy (1992) report their surroundings and seem to make at-
that as many as 60% of the girls appear to tempts at using residual functional skills.
compensate for this delay with alternative Communication skills are reported to im-
means of locomotion (e.g., rolling, creeping, prove with better interaction. Some girls
or shuffling). Approximately 80% of per- employ eye pointing, babbling, or even
sons with Rett syndrome will attain inde- word pieces to signal communicative intent
pendent ambulation. Thus, additional gross (Sekul & Percy, 1992). Seizures occur in up
motor abilities often are learned during this to 80% of the girls with Rett syndrome
stage, but are delayed in their appearance, (Coleman et al., 1988). Spasticity or rigid-
even as others are lost. The symptoms of ity and scoliosis tend to progress, and jerky,
Rett Syndrome: A Pervasive Developmental Disorder 129

truncal ataxia and apraxia become promi- Hagberg, Berg, & Steffenburg, 2001; Iyama,
nent. 1993). These deaths are reported throughout
Stage 4 (onset 10+ years): Progressive mus- the first 3 decades of the life span with their
cle wasting, scoliosis, spasticity, and rigid- peak occurrence falling in the second decade
ity frequently are displayed during this of life. Cardiac disturbances (Percy, 1992;
late stage. Decreasing mobility and late- Sekul et al., 1991), breathing dysfunctions, and
stage second neuron abnormalities (e.g., seizures (Hagberg, 1989) have been suggested
drop-foot abnormalities, remarkably plantar- as possible causes for these premature deaths.
flexed feet) may require the use of a wheel- The four-stage clinical pattern and profile
chair. Spinal cord dysfunction appears to act for Rett syndrome has been reported to be “a
in conjunction with extrapyramidal features sometimes crude and a somewhat simplistic
to lessen mobility (Witt-Engerstrom & Hag- frame” for specifically characterizing and
berg, 1990). Interestingly, this stage is covering the whole profile in all cases (Hag-
marked by increased motor deterioration berg & Witt-Engerstrom, 1986, p. 58). Transi-
only. The person’s cognitive functioning re- tions between stages are often indistinct and
mains stable, while social interaction (eye may be difficult to discern precisely for re-
contact) and attentiveness improve. Seizure search purposes (Philippart, 1986). Even so,
activity often becomes less problematic, al- the staging system has been found to be a use-
lowing for a decrease in the anticonvulsant ful instrument for a more systematic registra-
regimen for some girls. tion, thought, and approach to the complex
clinical manifestations of individuals with the
Although persons over the age of 25 have Rett syndrome as they progress through the
been identified, the oldest being 76 years of disorder.
age (International Rett Syndrome Association,
personal communication, 2002), little system- DIAGNOSTIC CRITERIA
atic research has been conducted with this
group to provide information on the course of The diagnostic criteria for Rett syndrome, ini-
the disorder past adolescence. There is some tially developed in the mid-1980s (Hagberg,
evidence to suggest that the life expectancy of Goutieres, Hanefeld, Rett, & Wilson, 1985;
persons with Rett syndrome may be dimin- The Rett Syndrome Diagnostic Criteria Work
ished; however, precise information is not Group, 1988), have recently been revised (Hag-
available at this time. A recent case study of a berg, Hanefeld, Percy, & Skjeldal, 2002). The
60-year-old woman with Rett syndrome was re- fourth edition (revised) of the Diagnostic and
ported by Jacobsen, Viken, and von Tetzchner Statistical Manual of Mental Disorders (DSM-
(2001). With proper attention and care, this IV-TR) has included Rett syndrome (termed
person was able to regain numerous lost skills Rett’s Disorder) as a subcategory of Pervasive
including ambulation. The report stresses the Developmental Disorder (American Psychiatric
importance of environmental adaptations nec- Association, 1994). The DSM-IV-TR diagnostic
essary for rehabilitation. Failure to keep accu- criteria for 299.80 Rett’s Disorder generally
rate developmental histories, infrequent coincide with those initially proposed by the
diagnostic evaluations of adults, and the proba- Rett Syndrome Diagnostic Criteria Work Group
bility of secondary contractures masking the (1988). The DSM-IV-TR criteria, however, dif-
more classic signs of the disorder impede the fer in some significant ways from the most
identification of older persons with Rett syn- recent revision of the diagnostic criteria devel-
drome. Malnutrition, uncontrolled seizures, oped by members of an international panel of
swallowing difficulties, and health problems experts convened by the International Rett Syn-
secondary to immobility increase the risk for a drome Association (IRSA: Hagberg et al.,
shortened life span in persons with Rett syn- 2002). For example, the DSM-IV-TR indicates
drome. Premature deaths in ambulatory girls “apparently normal psychomotor development
thought to be healthy, except for demonstrating through the first 5 months after birth” whereas
Rett syndrome, and whose seizures were under Hagberg and his associates (2002) indicate that
control have been reported (Hagberg, 1989; psychomotor development is “largely normal
130 Diagnosis and Classification

through the first six months or may be delayed similar to those reported for Rett syndrome. In
from birth” (p. 295). Deceleration of head fact, six males have been registered with the
growth between 5 and 48 months of age is a International Rett Syndrome Association (per-
necessary criterion within the DSM-IV-TR. The sonal communication, 2002). All but one of
revised IRSA diagnostic criteria, on the other these male cases, however, fail to meet the
hand, indicate that postnatal deceleration of strict criteria necessary to be included as con-
head growth will be evident in the majority of firmed classical cases of Rett syndrome. One
cases, but not all. A slight variation between report (Topcu et al., 2002) describes a male
the DSM-IV-TR and the IRSA diagnostic crite- who does, in fact, display classical Rett syn-
ria involves the established age limits for vari- drome and is mosaic (somatic mosaicism) for
ous symptoms. The DSM-IV-TR has adopted a the truncating MECP2 mutation. Nevertheless,
5-month upper age level for normal psychomo- Rett syndrome is almost exclusively seen in
tor development and the lower age limit for loss girls due to the predominant occurrence of mu-
of purposeful hand skills. Thus, these criteria tations on the paternal X chromosome, and also
coincide with the 5-month lower limit for the the presumed early postnatal lethal effect of the
appearance of decelerated head growth. The disease-causing mutations in hemizygous boys
adoption of the 5-month upper age limit for nor- (Topcu et al., 2002).
mal development may prove too restrictive. The
Rett Syndrome Diagnostic Criteria Work Group VARIANTS OF RETT SYNDROME
(1988) provided a footnote to their criteria to
indicate that “apparently normal development The mutation of MECP2, a regulating gene on
may appear for up to 18 months” (p. 426). No the X chromosome, has been linked to the Rett
such notation is provided within the current syndrome (Amir et al., 1999; Hoffbuhr, Moses,
DSM-IV criteria. Jerdonek, & Naidu, 2002). A number of re-
The Hagberg et al. (2002) criteria specify searchers have completed surveys of the muta-
supportive criteria such as disturbances in tions found in various populations of individuals
breathing when awake, bruxism, impaired with Rett syndrome (e.g., Bieber Nielsen et al.,
sleep patterns, impaired muscle tone, muscle 2001; Bienvenu et al., 2000; Cheadle, Gill, &
wasting, and dystonia, to mention a few. The Fleming, 2000; Wan, Lee, & Zhang, 1999;
DSM-IV-TR includes EEG abnormalities, Xiang et al., 2000) and have reported the identi-
seizures, nonspecific brain-imaging abnormal- fication of mutations in MECP2 in 35% to 87%
ities, and severe or profound mental retarda- of those tested.
tion (Axis II ) as associated features and Although mutations of the MECP2 gene ap-
disorders. Thus, the two sources differ in the pear to play a critical role in the development
associated and supportive features of the dis- of Rett syndrome, they cannot yet serve as dis-
order. Moreover, the current DSM-IV-TR crite- tinctive diagnostic markers because other dis-
ria fail to indicate exclusionary criteria that orders have been linked to mutations of the
are critical for differential diagnosis. Table 5.1 MECP2 gene. These mutations have been doc-
provides the revised diagnostic criteria for umented in persons affected with severe en-
Rett syndrome specified by Hagberg and his cephalopathy (Hoffbuhr, Devaney, et al., 2001;
associates (2002) and Rett’s Disorder as pro- Wan et al., 1999), X-linked mental retardation
vided in the DSM-IV-TR. (Meloni, Bruttini, & Longo, 2000), infantile
Classical Rett syndrome, to date, is almost autism (Beyer et al., 2002), mild learning dis-
exclusively described in female patients. Some ability (Hoffbuhr, Devaney, et al., 2001), and
researchers (Clayton-Smith, Watson, Ramsden, an Angelman syndrome-like phenotype (P. Wat-
& Black, 2000; Coleman, 1990; Eeg-Olofsson, son, Black, & Ramsden, 2001). Therefore, as
Al-Zuhair, Teebi, Zaki, & Daoud, 1990; no specific diagnostic marker variable has yet
Jan, Dooley, & Gordon, 1999; Philippart, 1990; been identified for Rett syndrome, clinical ho-
Topcu et al., 2002; Topcu, Topaloglu, Renda, mogeneity is essential for epidemiological and
Berker, & Turanli, 1991), however, have pre- research purposes. The strict diagnostic crite-
sented case studies of males displaying behav- ria previously discussed, however, may result
ioral symptoms and developmental histories in a failure to recognize the spectrum of pheno-
Rett Syndrome: A Pervasive Developmental Disorder 131

TABLE 5.1 Diagnostic Criteria for Rett Syndrome and Rett’s Disorder
Diagnostic Criteria Rett’s Disorder

Necessary Criteria Necessary criteria

1. Apparently normal prenatal and perinatal history A. All of the following:
2. Psychomotor development largely normal through 1. Apparently normal prenatal and perinatal
the first 6 months or may be delayed from birtha development
3. Normal head circumference at birth 2. Apparently normal psychomotor development
4. Postnatal deceleration of head growth in the through the first 5 months after birth
majority B. Onset of all of the following after the period of
5. Loss of achieved purposeful hand skills between normal development:
ages 6 months and 2 1⁄ 2 years 1. Deceleration of head growth between ages 5
6. Stereotypic hand movements such as hand and 48 months
wringing/squeezing, clapping/tapping, mouthing 2. Loss of previously acquired purposeful hand
and washing/rubbing automatisms skills between ages 5 and 30 months with
7. Emerging social withdrawal, communication subsequent development of stereotyped hand
dysfunction, loss of learned words, and cognitive movements (e.g., hand-wringing and hand-
impairment washing)
8. Impaired (dyspraxic) or failing locomotion 3. Loss of social engagement early in the course
(although often social interaction develops later)
Supportive Criteria 4. Appearance of poorly coordinated gait or trunk
1. Disturbances of breathing while awake ( hyper- movements
ventilation, breath holding, forced expulsion of air 5. Severely impaired expressive and receptive
or saliva, air swallowing) language development with severe psychomotor
2. Bruxism (teeth grinding) retardation
3. Impaired sleep pattern from early infancy
4. Abnormal muscle tone successively associated with Associated Features and Disorders
muscle wasting and dystonia 1. Severe or profound mental retardation (Axis II )
5. Peripheral vascomotor disturbances 2. Increased frequency of EEG abnormalities and
6. Scoliosis/ kyphosis progressing through childhood seizure disorders
7. Growth retardation 3. Nonspecific brain imaging abnormalities
8. Hypotrophic small and cold feet: small, thin hands

Exclusionary Criteria
1. Organomegaly or other signs of storage disease
2. Retinopathy, optic atrophy, or cataract
3. Evidence of perinatal or postnatal brain damage
4. Existence of identifiable metabolic or other
progressive neurological disorder
5. Acquired neurological disorder resulting from severe
infections or head trauma
Development may appear to be normal for up to 18 months.
Source: From “An Update on Clinically Applicable Diagnostic Criteria in Rett Syndrome,” by B. Hagberg,
F. Hanefeld, A. Percy, and O. Skjeldal, 2002, European Journal of Paediatric Neurology, 6, pp. 292–297; the diag-
nostic criteria and information on Rett’s Disorder is from Diagnostic and Statistical Manual of Mental Disorders,
fourth edition, text revision, American Psychiatric Association, 1994, Washington, DC: Author. Reprinted with

typic manifestations that might be included his associates (2002) have developed revised
under the Rett syndrome classification. criteria for the delineation of variant pheno-
Clinical variants of children with similar types. These criteria are provided in Table 5.2.
clinical courses who do not fulfill all the cur- Atypical forms include individuals with devel-
rent diagnostic criteria have been recognized opmental delays prior to regression or who lack
in the literature (Hagberg, 1995; Hagberg & the initial period of normal development (con-
Skjeldal, 1994; Huppke, Held, Laccone, & genital variant) or who display an early psy-
Hanefeld, 2003; Zappella, 1992). Hagberg and chomotor delay but without regression until
132 Diagnosis and Classification

TABLE 5.2 Diagnostic Criteria for the the variant forms of Rett syndrome may indi-
Delineation of Variant Phenotypes cate the presence of genocopying mutations or
A. Must meet at least three of the following Main (yet unidentified) mutations in regulatory ele-
Criteria: ments of MECP2 (Shahbazian & Zoghbi, 2002).
1. Absence or reduction of hand skills
2. Reduction or loss of babble speech
3. Monotonous pattern of hand stereotypies
4. Reduction or loss of communication skills
5. Deceleration of head growth from first years The presentation of Rett syndrome differs con-
of life siderably depending on the stage and age of
6. Rett syndrome disease profile: a regression observation. For example, a child 4 or 5 years
stage followed by a recovery interaction of age with classical Rett syndrome can be cor-
contrasting with slow neuromotor regression rectly diagnosed with relative ease. Due to
B. Must meet at least 6 of the following Supportive vague symptomology, diagnosis during infancy
Criteria is frequently misinterpreted. Likewise, the late
1. Breathing irregularities stage in adolescence displays a common, com-
2. Bloating/air swallowing
3. Bruxism ( harsh sounding teeth grinding)
plex picture of severe multiple disabilities
4. Abnormal locomotion with secondary contractures resembling any
5. Scoliosis/ kyphosis number of disorders and, therefore, often is
6. Lower limb amyotrophy (weakening and misdiagnosed. To fully understand this condi-
wasting of the muscle) tion, the diagnostician must recognize and
7. Cold, purplish feet, usually growth impaired consider the entire disease process (Trevathan
8. Sleep disturbances including night screaming
& Naidu, 1988).
9. Laughing/screaming spells
Only 18% to 23% of the estimated 8,000 to
10. Diminished response to pain 10,000 individuals in the United States af-
11. Intense eye contact /eye pointing flicted with Rett syndrome have been identi-
Note: From “An Update on Clinically Applicable Diag-
fied thus far (International Rett Syndrome
nostic Criteria in Rett Syndrome,” by B. Hagberg, F. Association, personal communication, 2002;
Hanefeld, A. Percy, and O. Skjeldal, 2002, European Moser, 1986). Lack of awareness of this disor-
Journal of Paediatric Neurology, 6, pp. 292–297. der on the part of physicians and clinicians is
undoubtedly a major contributing factor to this
state of affairs. Even when physicians are
school age (delayed onset variant); a group aware of the Rett syndrome, an accurate diag-
whose supportive characteristics do not appear nosis is not always forthcoming. Table 5.3 pre-
until late childhood (atypical or “ form fruste”); sents some of the clinical characteristics by
those with a family recurrence (familial vari- stage and differential diagnoses often assigned
ant); and those who have retained some speech to persons with Rett syndrome.
(“preserved speech variant ”). In the latter The most common nonspecific diagnosis for
group, words or even short sentences may be children with Rett syndrome above age 1 year is
retained, although they are typically not em- reported to be early infantile autism (DSM-IV-
ployed in a functional manner. The head cir- TR—299.00 Autistic Disorder; Olsson, 1987).
cumference of individuals with the form fruste In fact, many children with Rett syndrome
variant often fall within normal limits, yet sig- seem to fulfill the necessary criteria to estab-
nificantly below the norm (Hagberg, Stenbom, lish the diagnosis of infantile autism (Gillberg,
& Witt-Engerstrom, 2000). Individuals with 1987; Olsson, 1987; Olsson & Rett, 1985).
the congenital variant often display a more se- Thus, some researchers (Allen, 1988; Gillberg,
vere phenotype than that found in classical Rett 1989) argued that perhaps Rett syndrome might
syndrome (Hagberg et al., 2000). Mutations in best be thought of as a subtype of autism or
the MECP2 gene have been identified in as overlapping diagnostic entities. The DSM-IV-
many as 50% of the individuals tested with TR includes Autistic Disorder and Rett’s
variant forms of Rett syndrome (Bieber Nielsen Disorder as subcategories of Pervasive De-
et al., 2001; Cheadle et al., 2000). The lower velopmental Disorder. The behavioral pat-
percentage of identified MECP2 mutations in terns, progression, and prognosis of these two
TABLE 5.3 Rett Syndrome: Clinical Characteristics and Dif ferential Diagnosis by Stage

Stage Clinical Characteristics Differential Diagnosis

Early onset stagnation stage Developmental stagnation Benign congenital hypotonia

Onset: 6 to 18 months Deceleration of head/ brain growth Prader-Willi syndrome
Duration: months Disinterest in play activity Cerebral palsy
Rapid destructive stage Rapid developmental regression Autism
Onset: 1 to 4 years with irritability Psychosis
Duration: weeks to months Loss of hand use Hearing or visual disturbance
Seizures Encephalitis
Hand stereotypies: wringing, clap- Infantile spasms (West syndrome)
ping, tapping, mouthing Tuberous sclerosis
Autistic manifestations Ornithine carbamoyl transferase
Loss of expressive language deficiency
Insomnia Phenylketonuria
Self-abusive behavior (e.g., chew- Infantile neuronal ceroid lipofusci-
ing fingers, slapping face) nosis (INCL)
Plateau stage Severe mental retardation /apparent Spastic ataxic cerebral palsy
Onset: 2 to 10 years dementia Spinocerebral degeneration
Duration: months to years Amelioration of autistic features Leukodystrphies or other storage
Seizures disorders
Typical hand stereotypies: wring- Neuroaxonal dystrophy
ing, tapping, mouthing Lennox-Gastaut syndrome
Prominent ataxia and apraxia Angelman syndrome
Hyperventilation, breath-holding,
Apnea during wakefulness
Weight loss with excellent appetite
Early scoliosis
Late motor deterioration Combined upper and lower motor Unknown degenerative disorder
Onset: 10+ years neuron signs
Duration: years Progressive scoliosis, muscle
wasting, and rigidity
Decreasing mobility: wheelchair-
Growth retardation
Improved eye contact
Staring, unfathomable gaze
Virtual absence of expressive and
receptive language
Trophic disturbance of feet
Reduced seizure frequency

Source: “Diagnostic Criteria for Rett Syndrome,” by The Rett Syndrome Diagnostic Criteria Work Group, 1988,
Annals of Neurology, 23, pp. 425–428. Reprinted with permission.

134 Diagnosis and Classification

conditions, nevertheless, differ significantly. TABLE 5.4 Comparison of Rett Syndrome and
Clinically important differences have been Infantile Autism
identified between the Rett syndrome (espe- Rett Syndrome
cially during the latter two stages) and other 1. Normal development to 6 to 18 months
conditions with autism or autistic traits (Naidu 2. Progressive loss of speech and hand function
et al., 1990; Olsson, 1987; Olsson & Rett, 1985, 3. Profound mental retardation in all functional
1990; Percy, Zoghbi, Lewis, & Jankovic, 1988).
4. Acquired microcephaly, growth retardation,
A basic distinction between the two disorders decreased weight gain
can be made on the basis of motor behavioral 5. Stereotypic hand movements always present
analysis (Olsson & Rett, 1985, 1987; Percy, 6. Progressive gait difficulties, with gait and trun-
Zoghbi, et al., 1988). “Whereas autism repre- cal apraxia and ataxia; some may become nonam-
sents a regression of verbal but not motor skills; bulatory
Rett syndrome involves the simultaneous re- 7. Language always absent
8. Eye contact present, and sometimes very intense
gression of both skills” (Percy, Zoghbi, et al., 9. Little interest in manipulating objects
1988, p. S67). Stereotypic behavior associated 10. Seizures in at least 70% in early childhood (vari-
with infantile autism is generally more complex ous seizure types)
and, unlike that in Rett syndrome, often in- 11. Bruxism, hyperventilation with air-swallowing
volves the manipulation of an object with and breath-holding common
preservation of the pincer grasp. Children with 12. Choreoathetoid movements and dystonia may be
Rett syndrome reportedly differ from children
with infantile autism with respect to their res- Infantile Autism
piratory pattern (displaying breath-holding, 1. Onset from early infancy
hyperventilation, and air-saliva expulsion); the 2. Loss of previously acquired skills does not occur
3. More scatter of intellectual function. Visual-
presence of ataxia and apraxia, bruxism, hy-
spatial and manipulative skills often better than
poactivity, and a general slowness of move- apparent verbal skills
ments; and an absence of purposeful hand 4. Physical development normal in the majority
movement (Gillberg, 1986; Olsson, 1987; 5. Stereotypic behavior is more varied in manifes-
Percy, Zoghbi, et al., 1988; Sekul & Percy, tation and is always more complex; midline man-
1992). Persons with Rett syndrome demon- ifestations rare
strate a very restricted repertoire of movements 6. Gait and other gross motor functions normal in
first decade of life
that appear monotonous in both form and speed
7. Language sometimes absent; if present, peculiar
(Olsson & Rett, 1985). Van Acker (1987) re- speech patterns always present; markedly
ported that the stereotypic behaviors of persons impaired nonverbal communication
with Rett syndrome were displayed in patterned 8. Eye contact with others typically avoided or
sequences with significant conditional proba- inappropriate
bilities, whereas those of persons with infantile 9. Stereotypic ritualistic behavior usually involves
autism were displayed in a random fashion. skillful but odd manipulation of objects or sen-
sory self-stimulation
Budden (1986) has presented another critical 10. Seizures (usually temporal-limbic complex par-
feature that may help in the differential diagno- tial) in 25% in late adolescence and adulthood
sis of Rett syndrome and infantile autism: Per- 11. Bruxism, hyperventilation, and breath-holding
sons with Rett syndrome frequently develop not typical
appropriate speech before the onset of symp- 12. Dystonia and chorea not present*
toms. On the other hand, children with autism *Extrapyramidal signs may appear in some patients
differ from those with Rett syndrome in that with autism after puberty.
they display overactivity and inappropriate vo- Source: “ The Clinical Recognition and Differential
calizations, and tend to replicate simple motor Diagnosis of Rett Syndrome,” by E. Trevathan and S.
activities or complex movements within a rich Naidu, 1988, Journal of Child Neurology, 3(Suppl.),
repertoire of motor behavior (Percy, Zoghbi, pp. S6–S16. Reprinted with permission.
et al., 1988). Table 5.4 presents a comparison of
the clinical manifestations differentiating Rett
syndrome from infantile autism.
Rett Syndrome: A Pervasive Developmental Disorder 135

Considering the present concept of infantile Both disorders cause a rapid regression of psy-
autism as a behavioral syndrome, the initial chomotor development and the manifestation of
differential diagnosis of Rett syndrome for hand and finger stereotypies at approximately
some children may prove somewhat problem- the same age. As INCL progresses, however,
atic. Infantile autism, however, is very rare in myoclonus and retinal degeneration becomes
a female, which means that the mere presence apparent and differentiates the two disorders
of severe autistic symptomology in a girl under (Sekul & Percy, 1992). Failure to designate
2 years should prompt the consideration of retinopathy or optic atrophy as exclusionary
Rett syndrome in the differential diagnosis. criteria for Rett’s Disorder within the DSM-IV-
One must also be aware, however, that “a large TR may complicate the differential diagnosis of
percentage of children with Rett syndrome age this disorder.
0 to 6 months or older than 3 to 5 years are not A report by Philippart (1993) links Rett syn-
autistic” (Olsson & Rett, 1987). Thus, physi- drome with tuberous sclerosis, a nerocutaneous
cians and clinicians alike must realize that the disorder. Although this disorder may show ini-
presence of an autistic behavioral syndrome is tial similarities to Rett syndrome, close exami-
not an obligatory condition for a diagnosis of nation of the skin with a Wood’s lamp and
Rett syndrome (Olsson & Rett, 1987). the presence of serial computed tomography
Millichap (1987) has suggested that Rett abnormalities will distinguish this disorder
syndrome might represent a variant of Child- (Sekul & Percy, 1992). Chromosomal disorders,
hood Disintegrative Disorder (Heller’s syn- such as Angelman syndrome (“ happy puppet
drome). Children with Childhood Disintegrative syndrome”) can display similar features to Rett
Disorder develop their symptoms later (e.g., at syndrome. Children with Angelman syndrome,
least age 2 and typically age 3 or 4 years) than however, fail to display a period of normal
individuals with Rett syndrome and normal development and subsequent rapid regression.
neurological findings are reported in persons Acute and chronic encephalitis may be dis-
with the former disease. In a comparison of tinguished by examination of the cerebrospinal
two boys with Childhood Disintegrative Disor- fluid (CSF) and by a characteristic elec-
der and six girls with Rett syndrome, Burd, troencphalography. The loss of language and
Fisher, and Kerbeshian (1989) reported per- the development of seizures in preschool-age
sons afflicted with these disorders differed children are similar in both Rett syndrome and
from children with classic autism. Children with the Landau-Kleffner syndrome. Head circum-
Childhood Disintegrative Disorder (Heller’s ference growth and motor skills are preserved
syndrome) and Rett syndrome displayed nor- in the Landau-Kleffner syndrome.
mal prenatal and perinatal periods, followed In summary, the clinical identification of
by marked developmental regression after Rett syndrome rests on the careful exploration
which they acquired few or no new skills. The of clinical manifestations and the specific
authors have suggested that these children pattern of symptom progression. The differen-
should be distinguished from those with classic tial diagnosis based on clinical observation, al-
autism, and should be classified as “pervasive though frequently difficult at presentation
disintegrative disorder, Heller type” and “per- (especially during the earliest stages), becomes
vasive disintegrative disorder, Rett type.” The much easier after follow-up over several months
DSM-IV-TR has included both Rett’s Disorder to a few years. A diagnosis of Rett syndrome
(Rett syndrome) and Childhood Disintegrative should involve a molecular genetic analysis to
Disorder (Heller’s syndrome) as subcategories identify mutations in the MECP2 gene.
of Pervasive Developmental Disorders.
Stage 2 developmental regressions often sug- EPIDEMIOLOGY
gest neurodegenerative diseases. The earliest
stages of Rett syndrome are difficult to distin- Rett syndrome has been reported to exist on all
guish from infantile neuronal ceroid lipofuscin- the populated continents and in most countries
sis (INCL), an autosomal recessive disease of the world (e.g., Budden, 1986; Goutieres
especially frequent in the Finnish population. & Aicardi, 1985; Hanaoka, Ishikawa, &
136 Diagnosis and Classification

Kamoshita, 1985; Kerr & Stephenson, 1986; studies suggesting the prevalence of Rett syn-
Moodley, 1992). Thus, the literature supports drome has perhaps been overestimated.
the view that Rett syndrome does not seem to No matter which estimate is employed, Rett
be a rare disorder and that it is more or less syndrome seems to be significantly more preva-
universal; additionally less than 2 per 100 lent among girls than phenylketonuria (PKU), a
cases of Rett syndrome display familial rela- condition for which all neonates are screened in
tionships (Zoghbi, 1988). Thus, the pattern of the majority of developed countries (Hagberg,
occurrence of Rett syndrome is dissimilar 1985). As progressive brain disorders and meta-
to that of traditional inborn errors of metabo- bolic diseases together constitute only 5% to
lism (e.g., glactosemia, Hartnup’s disease, ke- 6% (1.5 to 2.0 per 10,000 children) of the eti-
toaciduria, phenylketonuria), which often ologies in persons with severe or profound men-
display strong geographic, ethnic, and familial tal retardation, the Rett syndrome should be
accumulation. considered as an important etiologic factor in
Approximately 2,329 cases of the Rett syn- females, second only to Down syndrome. In
drome worldwide are registered with the Inter- fact, this syndrome might well be responsible
national Rett Syndrome Association, with the for one-fourth to one-third of progressive devel-
following distribution: United States, 1,887; opmental disabilities among females (Hagberg,
Canada, 101; Mexico, 9; and other foreign, 312 1985).
(International Rett Syndrome Association, per-
sonal communication, 2002). The prevalence of ETIOLOGY
the Rett syndrome has been studied, based on
the Swedish registry for mental retardation Since Rett syndrome, at least in its classical
and surveys of neuropediatricians, in a part of form, had only been reported in females cou-
southwestern Sweden comprising five counties pled with a number of familial cases (repre-
and the city of Gothenburg (Hagberg, 1985). In sented in Table 5.5), researchers suspected the
a population of 315,469 children and adoles- existence of an X-linked dominant genetic in-
cents, 6 to 17 years of age, 13 cases were de- heritance as the basis for this disorder. It was
tected, all girls. The corresponding prevalence hypothesized that the genetic mutation would
was about 1 per 15,000 live female births. In be lethal in males, resulting in a spontaneous
their study of 5,400 consecutive referrals to the abortion of the fetus (Comings, 1986; Ric-
pediatric neurology center in western Scotland, cardi, 1986). Genetic mapping in the familial
Kerr and Stephenson (1985) identified 19 cases cases identified a mutation at the Xq28 locus
of Rett syndrome. The resulting prevalence rate (Sirianni, Naidu, Pereira, Pillotto, & Hoffman,
was very similar to that reported by Hagberg 1998). In 1999, researchers (Amir et al., 1999)
(1985); 1 per 12,000 to 13,000 females (Kerr & identified mutations in the MECP2 gene as the
Stephenson, 1986). cause of Rett syndrome.
A prevalence study was conducted within During normal human development, many
several geographic areas of the state of Texas in genes are expressed in a tissue-specific manner.
the United States by researchers at the Baylor That is, the gene must only function in the cre-
College of Medicine (Kozinetz et al., 1993). ation of specific cells. In fact, more than one-
The study employed the Texas Rett Syndrome third of our genes are expressed only within the
Registry and explored females ages 2 through brain. Many of these genes are needed during
18 years. This study was the first with a large critical periods of central nervous system
ethnic mix that would allow an exploration of (CNS) development, and then their expression
racial /ethnic group-specific prevalence of Rett must be turned off. Other genes are required
syndrome. A prevalence estimate of approxi- only after birth and turning these genes on and
mately 1 per 22,800 live female births was off at appropriate times is critical for normal
reported with no significant differences in and proper development (Lombroso, 2000). The
prevalence estimates by race/ethnicity (African MECP2 gene encodes the development of the
American, Caucasian, and Hispanic). This es- transcriptional silencing methyl-CpG binding
timate is lower than that reported in earlier protein-2. This protein plays an important role
Rett Syndrome: A Pervasive Developmental Disorder 137

TABLE 5.5 Familial Cases in Rett Syndrome

Number of Pairs
Monozygotic twins
Both females aff licted 8
Only one female aff licted 1
Dizygotic twins (female/female)
Both females aff licted 2
Only one female aff licted 5
Dizygotic twins (female/male)
Female aff licted 8
Full sisters 6
Half-sisters 2
Full cousins 1
Second cousins 2
Second half-cousins 1
Aunt—niece 1
Great-grand aunt—niece 1
Sister and half-brother, both have children with Rett syndrome 1
Mother—daughter 1
Source: “Genetic Aspects of Rett Syndrome,” by H. Y. Zoghbi, 1988, Journal of Child
Neurology, 3(Suppl.), pp. S76–S78. Adapted with permission.

in the “silencing” of various genes during CNS Belichenko, Woodcock, and Woodcock (2001)
development. This protein binds to prescribed suggest the need to explore further those cases
methylated cytosine nucleotides (CpG dinu- in which no mutations of the MECP2 gene have
cleotides) on the DNA. The bound DNA- been identified for individuals who have been
MECP2 complex then interacts with a histone clinically diagnosed with Rett syndrome. They
deacetylase complex and the transcriptional suggest that additional novel MECP2 muta-
co-repressor Sin3A. Together, these repressors tions, other genetic mutations, or external fac-
alter the chromatin making the genes inaccessi- tors may yet be identified to play a role in the
ble to transcriptional activators—in essence, occurrence of Rett syndrome.
silencing the further transcription of that gene. Numerous different mutations of the
The mutations in the MECP2 gene in Rett syn- MECP2 gene have been identified in individu-
drome result in a failure to produce the MECP2 als with Rett syndrome. Van den Veyver and
protein. Thus, the genes that MECP2 would Zoghbi (2001) found 30 missense mutations
normally silence continue to engage in ongoing (single-base changes that result in the substitu-
transcription. The number and nature of the tion of one amino acid for another in the protein
genes for which MECP2 is meant to suppress product), 22 of which are in the methyl-CpG
throughout the genome is as yet unknown binding domain (MBD) of MECP2. Addition-
(Singer & Naidu, 2001); although there is rea- ally they found 35 nonsense mutations (single-
son to believe these genes regulate the develop- base changes that create a termination condon
ment and mature function of the brain and resulting in a shortened, dysfunctional protein
central nervous system (Shahbazian & Zoghbi, product), 12 frameshift mutations (changes in
2002; Webb & Latif, 2001). Genetic analyses the DNA chain that occur when the number of
have reported as many as 87% of the females nucleotides inserted or deleted is not a multiple
displaying the classical Rett syndrome and 50% of three, so that every condon beyond that point
of those displaying a variant form test positive is read incorrectly during translation), and 1
for mutations of the MECP2 gene (Bieber splice-site mutation (failure to remove intron,
Nielsen et al., 2001). These numbers suggest or noncoding DNA, sequences prior to protein
strong support for the current criteria used translation altering the sequence of the protein
for clinical diagnosis of Rett syndrome. Kerr, product).
138 Diagnosis and Classification

Most mutations occur de novo (as new inactivation is skewed. That is, a greater number
spontaneous mutations in the generation of the of the genes on either the paternal or maternal X
sperm or ovum; Dragich, Houwink-Manville, chromosome remain active. Approximately 10%
& Schanen, 2000). Kondo et al. (2000) propose of females in the general population displayed
that Rett syndrome is more frequent in females skewed inactivation of their X chromosomes
since 88% of the sporadic or de novo cases in (Lyon, 1972). As the cell reproduces, the same
their study appeared to have paternally de- gene on each X chromosome remains
rived mutations in MECP2. In contrast, the fa- active (or inactive) in each copy. The severity
milial cases and mutations in a number of of the symptoms in Rett syndrome may be re-
males suspected of displaying a variant of Rett lated to both the number and the location of
syndrome were determined to be maternally the genes with the mutated MECP2 gene. If the
derived. Girard et al. (2001) report that de active gene in most cells holds the mutation
novo MECP2 mutations with paternal origin (skewed XCI ); symptoms will likely be more
were identified in 71% of the families studied. severe. Likewise, if the cells with the active
Two cases of maternal origin MECP2 muta- gene displaying the mutation are more often in-
tion were reported. The high frequency of volved in the encoding for the development of
male germ-line transmission of the mutation is critical tissues (e.g., brain cells) symptoms
consistent with a predominant occurrence of would be more dramatic. Bieber Nielsen et al.
the disease in females (as males do not inherit (2001), however, failed to identify any correla-
their X chromosome from their father). tion between the X chromosome inactivation
If separate mutations affect a single gene, pattern in peripheral blood samples with the
the condition is termed allelic heterogeneity, clinical presentation or severity of symptoms in
and this is found in the majority of disease- a group of Danish women with Rett syndrome.
causing mutations studied to date. Interest- To date, mutations in the MECP2 gene have
ingly, even when different regions of the same been identified in many, but not all individuals
gene are mutated, they often result in the diagnosed with the Rett syndrome (e.g., Bien-
same clinical phenotype among affected venu et al., 2000; Hoffbuhr et al., 2001; Xiang
individuals. Occasionally, however, different et al., 2000). MECP2 mutations have been iden-
mutations within the same gene will result tified in 70% to 90% of sporadic or de novo
in different clinical presentations. This may cases and approximately 50% of familial cases
explain some of the Rett syndrome variants. of Rett syndrome (Shahbazian & Zoghbi, 2001).
One study (Huppke, Laccone, Kramer, Engel, Failure to identify the mutation may result
& Hanefeld, 2000), however, failed to corre- from many factors including the accuracy of the
late the type of mutation and the phenotype. original diagnosis and the methodology used
Thus, factors other than type or position of the to screen for the mutations (e.g., secondary
mutation may influence the severity of the structural content prediction [SSCP], gene se-
symptoms. quencing, denaturing high performance liquid
Another factor hypothesized to play an im- chromatography [DHPLC]). Failure to identify
portant role in the symptoms associated with mutations in the MECP2 gene in individuals
Rett syndrome involves the pattern of X chro- displaying Rett syndrome also might indicate
mosome inactivation (XCI; Amir et al., 2000; that mutations will be found in other genes
Hoffbuhr et al., 2001; Takagi, 2001; Zoghbi, within the same enzymatic pathway as the
Percy, Schultz, & Fill, 1990). At conception, fe- MECP2 gene. Thus, genes encoding for other
males have two X chromosomes, with similar proteins required in the proper methylation of
genes on each (a gene pair). Having the genes the specific DNA sequences, or the deacetyla-
on both X chromosomes remain active results tion of histones might be involved (Lombroso,
in severe effects; thus, one of the genes in each 2000). Only the coding region of the MECP2
pair must be inactivated. Typically, the pattern gene has been carefully analyzed, so mutations
of inactivation is random between the genes on in regulatory elements (similar to other known
the X chromosome provided by the mother and C ➝ T transition mutation disorders) could ac-
those on the X chromosome provided by the count for those cases in which no mutation has
father. In some cases, however, the pattern of been identified (Shahbazian & Zoghbi, 2002).
Rett Syndrome: A Pervasive Developmental Disorder 139

As mentioned, mutations of the MECP2 the result of arrested brain development (Arm-
gene have also resulted in disorders that are strong, 2001).
phenotypically distinct from Rett syndrome Brucke et al. (1988) report the most con-
(e.g., nonspecific X-linked mental retardation, spicuous finding from their autopsy studies
congenital encephalopathy with respiratory ar- was the underpigmentation of the substantia
rest; e.g., Clayton-Smith et al., 2000; Hoffbuhr nigra (especially the zona compacta). It con-
et al., 2001). Thus, the mutation in the MECP2 tained many fewer well-pigmented neurons
gene cannot serve as a diagnostic marker vari- for the age of the person (53% to 73%), and
able for Rett syndrome. Genetic research will fewer pigmented granules per neuron,
need to further explore the correlation of geno- whereas the total number of nigral neurons
type and phenotype. More importantly, work to and the triphasic substructure of neurome-
identify the “downstream” genes that are af- lanin were within the normal range. The basal
fected by the mutation of the MECP2 gene and ganglia of some patients showed mild gliosis.
any other gene along the enzymatic pathway These findings were supported in a study of
will help improve our understanding of the 38 patients with Rett syndrome conducted in
pathomechanisms of the disorder. Sweden by Lekman et al. (1989). Brucke
et al. (1988) also reported low melanin con-
NEUROPATHOLOGY tent in the locus coeruleus. As the melanin
pigmentation in the substantia nigra normally
Neuropathological studies have been conducted increases with age, this lack of pigmentation
in Rett syndrome that may assist in a better un- serves as “evidence of a retardation in matu-
derstanding of how ineffective MECP2 activity ration of these neurons (in the substantia
produces Rett syndrome. Although individuals nigra and locus coeruleus) which possibly
with Rett syndrome are typically small for their leads to a decreased synthesis rate of dopa-
age, only the brain weighs less than expected mine and a compensatory enhancement in its
for the height and weight of the individual turnover rate” (Brucke et al., 1988, p. 323). In-
(Armstrong, Dunn, Schultz, et al., 1999). This creased levels of dopamine and serotonin
supports the view that the MECP2 gene appears metabolites in their subject support their hy-
to have its greatest impact on downstream pothesis. The abnormalities of the substantia
genes that code for the development and mature nigra and the related changes in dopamine syn-
function of brain tissue. Accumulated autopsy thesis could account for the prominent move-
studies (e.g., Brucke, Sofic, Killian, Rett, & ment disorder associated with Rett syndrome.
Riederer, 1988; Harding, Tudmay, & Wilson, Receptors for serotonin are hugely in-
1985; Jellinger & Seiteberger, 1986; Missliwetz creased in the brain stem at all ages in persons
& Depastas, 1985) suggest that the brain of the with Rett syndrome (Kerr & Witt-Engerstrom,
individual with Rett syndrome weighs signifi- 2001). Serotonin plays an early role in the de-
cantly less than that of controls matched for termination of later cortical function (Lager-
both age and height. The weight of the average crantz & Srinivasan, 1991) and is also an
brain from an individual with classical Rett important neuromodulator and neurotransmit-
syndrome weighed 950g, which is the brain ter in the mature brain.
weight of a normal 1-year-old child (Arm- Jellinger, Armstrong, Zoghbi, and Percy
strong, 2001). This researcher goes on to note (1988) presented evidence from electron mi-
that many individuals with form fruste and pre- croscopic studies of abnormal neurites in the
served speech variants of Rett syndrome dis- frontal cortex and caudate nucleus with greatly
play head circumferences (and presumably reduced axonal or dendritic connections. In a
brain weights) within normal limits. The de- study of two females with Rett syndrome
crease in brain size and weight appears to begin (Armstrong, 1992), the cortical neurons ap-
after birth (in most cases), starting at 3 to 4 peared to be less mature and demonstrated sig-
months of age. Brain size and weight appear to nificantly decreased dendritic arborization
stabilize in later childhood, and the absence of that did not appear to be age related. This could
markers for significant degenerative disorder be the result of the general growth arrest mani-
support the idea that the decreased brain size is fested in Rett syndrome and may, in part,
140 Diagnosis and Classification

explain the acquired microcephaly witnessed procedure has been developed for the post-
in Stage 1 of the disorder. mortem examination (Percy, Hass, Kolodnyu,
Golgi studies of cerebral cortical dendrites Moser, & Naidu, 1988), and is available from
(Armstrong, Dunn, Antalffy, & Trivedi, 1995) IRSA. Advance arrangements must be made
have identified a reduced dendritic arboriza- with a pathologist so that tissues can be frozen,
tion in the pyramidal neurons of Layers III, IV, optimally within 4 to 6 hours following death.
and V in frontal motor and inferior temporal
regions in the brains of individuals with Rett NEUROANATOMY
syndrome. In the affected regions, apical and
basal dendrites are selectively reduced; basal The development of sophisticated methods for
dendrites of Layers III and V in the frontal visualization of the human central nervous
and motor cortex; basal dendrites of Layer IV system in vivo has provided a means to quan-
of the subiculum, and apical dendrites of Layer tify brain structure and function in persons
V of the motor cortex. The dendrites of the with brain dysfunction. Routine neuroimaging
hippocampal and occipital regions are not sig- studies in persons with Rett syndrome, how-
nificantly reduced. The decreased dendritic ever, have only revealed occasional nonspe-
branching in Rett syndrome suggests that the cific changes. Serial computed tomography
synaptic input is reduced. (CT) and magnetic resonance imaging (MRI )
A study by Cornford, Philippart, Jacobs, studies have shown evidence of progressive
Scheibel, and Vinters (1994) reports on the brain atrophy, particularly in the frontal and
neuronal changes in the brain of a girl with Rett temporal regions, in some girls after age 2
syndrome observed in a frontal lobe biopsy per- years (Krageloh-Mann, Schroth, Niemann, &
formed at age 3 years and in the postmortem Michaelis, 1989; Nihei & Naitoh, 1990; No-
brain at age 15 years. Widespread neuronal mi- mura, Segawa, & Hasegawa, 1984; Yano et al.,
tochondrial inclusions and the appearance of 1991). These findings are consistent with the
dendritic retraction in Golgi-stained cortical pathological findings of Jellinger and Seite-
pyramidal and Purkinji neurons were the most berger (1986) reported earlier.
significant neuropathological features. The Decreased cerebral blood flow (to 88% of
Golgi preparations of the frontal cortex and that noted in age-matched controls) was re-
cerebellar folia (autopsy brain) manifested ported in seven persons with Rett syndrome
truncation and thickening of the dendrites and a (Nielsen, Friberg, Lou, Lassen, & Sam, 1990).
degenerate appearance of cortical pyramidal Single photon emission computed tomography
neurons similar to that of an aged brain. Thus, demonstrated significantly decreased cerebral
neuronal and mitochondrial deterioration ap- blood flow to the prefrontal and temporal re-
peared to continue after stabilization of the gions, whereas that to the primary sensorimotor
neurological deterioration (at 3 years) in Rett cortex remained unaffected. A similar pattern
syndrome. This has led these authors to specu- of cerebral blood flow is observed in infants
late, “Rett syndrome could result from inade- (Chugani, Phelps, & Mazziotta, 1987) suggest-
quate maintenance of a full array of neuronal ing that this finding may reflect the growth ar-
contacts, similar to the aging process, in which rest noted in Rett syndrome. Further evidence
such dendritic regression apparently occurs of abnormal cerebral blood flow in Rett syn-
over the span of many years” (p. 430). drome was reported by Yoshikawa et al. (1991).
The pathogenic mechanisms of the morpho- The developmental increase of the frontal-
logical brain lesions and their relations to to-temporal cerebral blood flow ratio demon-
clinical and neurochemical findings in Rett strated in age-matched controls was not ob-
syndrome remain unknown. Neuropathological served in six females with Rett syndrome.
studies at autopsy serve a critical role in at- Employing quantitative methods of analysis
tempts to better understand Rett syndrome. in neuroimaging, Cassanova and associates
The International Rett Syndrome Association (1991) reported smaller cerebral hemispheres
urges parents of persons with this disorder to in 8 persons with Rett syndrome when com-
consider the gift of autopsy should their chil- pared with controls. A decreased area of cau-
dren die prematurely. To this end, a uniform date nucleus, even when the overall smaller
Rett Syndrome: A Pervasive Developmental Disorder 141

brain area was taken into account, also was neuroimaging studies could make a significant
noted in the girls with Rett syndrome. Smaller contribution to our understanding of the etiol-
cerebral hemispheres, basal ganglia, corpus ogy, homogeneity, and pathogenesis of this
callosum, cerebellar hemispheres, inferior disorder. Segawa (2001) has presented a rather
olive, and anterior vermis were reported in 13 interesting discussion of the pathophysiology
females with Rett syndrome compared with 10 of Rett syndrome by attempting to relate many
female control subjects (Murakami et al., of the neuropathological findings to clinical
1992). Reiss et al. (1993) completed a quantita- characteristics of the disorder.
tive neuroimaging study that provided in vivo
neuroanatomical correlates of the neurological NEUROCHEMICAL ALTERATIONS
and developmental features of Rett syndrome.
This group reports reduced cerebral volume, a The progressive nature of Rett syndrome after
disproportionate reduction in brain tissue vol- an apparently normal pre- and neonatal period
umes, with a greater decrease of gray matter to is highly suggestive of a metabolic disorder
white, regional variation in the percentage of similar to PKU. The pathogenesis of the dis-
cortical gray matter (with frontal regions show- order, however, remains a mystery. Extensive
ing the greatest decrease), reduced volume of research exploring biogenic amines and endor-
subcortical gray matter (with the caudate nu- phins through the analyses of serum amino
cleus showing significant volume reduction), acids, urine amino and organic acids, lysoso-
and increased cerebrospinal fluid volume when mal enzymes, and routine chemistries has been
controlling for brain volume differences in the undertaken. One hypothesis is that symptoms
females with Rett syndrome. These findings, result from an abnormality in the dopamine
especially those related to the caudate nucleus, system, a neurotransmitter system that regu-
are of interest from a clinical standpoint as lates the control of voluntary movements in the
they may help explain the significant motor extrapyramidal system. This hypothesis, based
and cognitive-developmental symptoms pres- on the decreased pigmentation of the substan-
ent in Rett syndrome. tia nigra and the prominent movement dis-
Magnetic resonance imaging (MRI ) studies orders suggestive of extrapyramidal dysfunc-
have compared volumetric brain analyses of in- tion (Zoghbi, Percy, Glaze, Butler, & Riccardi,
dividuals with Rett syndrome to age-matched 1985), led Nomura and her associates (No-
controls (Gotoh et al., 2001; Subramaniam, mura et al., 1984; Nomura, Segawa, & Hig-
Naidu, & Reiss, 1997). Global reductions in urashi, 1985) to speculate that as the disease
grey matter volumes are reported; with exag- progresses the dopamine system becomes hy-
gerated loss within the prefrontal, posterior peractive due to postsynaptic supersensitivity
frontal, and anterior temporal regions. Subra- caused by hypoactive dopamine neurons. Ab-
maniam et al. (1997) reported a preferential re- normality in the dopamine system is supported
duction in the volume of the caudate nucleus. by the finding of a decrease of biogenic amine
Gotoh et al. (2001) identified thinning of the metabolites in cerebrospinal fluid in six chil-
corpus callosum, widening of the prepontin cis- dren with Rett syndrome, the most significant
tern, a narrowing of the brain stem, and cere- reductions being in homovanillic acid (HVA),
bellar atrophy in some individuals with Rett the major dopamine metabolite (Zoghbi et al.,
syndrome. No evidence of active degenerative 1985). These findings were extended to 32
disease, however, had been seen, and the whole girls found to have significant reductions in
brain seemed to be affected by the atrophy. HVA and 4-hydroxy-3-methoxyphenylethylene
The consistency of data obtained through glycol (HMPG), the metabolites of dopamine
the neuroimaging research with results from and norepinephrine (Zoghbi et al., 1989; Zoghbi
neuropathological investigations supports the et al., 1985). An abnormality in this system is
need for continued neuroimaging studies in further supported by demonstration in an au-
Rett syndrome. Longitudinal studies of sub- topsy study of decreasing binding of 3H spiper-
jects from the time the children manifest the one, a ligand with high affinity for dopamine
earliest signs of the syndrome would be espe- D2 receptors, in the putamen (Riederer et al.,
cially enlightening. Information gained from 1985). Hand-mouth stereotypies, hypotonia,
142 Diagnosis and Classification

and ataxia similar to that seen in Rett transferase (ChAT) activity in a series of
syndrome are demonstrated in boys with postmortem brain studies. The decreased
the Lesch Nyhan syndrome. In that disorder, ChAT activity may be related to loss of cholin-
all biochemical aspects of the function of ergic cells, qualitatively similar to the loss in
dopamine neuron terminals in the corpus Alzheimer’s disease. These studies (Wenk
striatum have been found to be decreased up to et al., 1991, 1993) also report decreased ChAT
10% to 30% of control values in autopsy studies activity in the hippocampus and thalamus con-
of three affected cases (Lloyd et al., 1981). sistent with a loss of cholinergic cells in the me-
Studies, however, have failed to replicate the dial septum and vertical limb of the Broca and
findings that abnormal neurotransmitter levels the pedunculopontine tegmental nucleus. This
characterize the Rett syndrome. Unlike the six loss of cholinergic cells throughout the basal
cases reported by Zoghbi et al. (1985) where forebrain might well be responsible for the cog-
norepinephrine and dopamine metabolites were nitive stagnation and memory loss characteris-
reduced in comparison to control individuals, tic of Rett syndrome, as has been suggested for
subjects in a more recent investigation (Harris the dementia associated with Alzheimer’s and
et al., 1986) did not demonstrate a reduction in Parkinson’s diseases (Collerton, 1986; White-
metabolites of either of these neurotransmitter house, Price, Clark, Coyle, & DeLong, 1981).
substances. Reduction of dopamine D2 receptor Persons with Rett syndrome have been re-
binding in the putamen, as found in the autopsy ported to display a remarkable tolerance for
study by Riederer et al. (1985) with 3H spiper- pain and a high rate of stereotyped behavior,
one, was not demonstrated in living subjects by seizure activity, and respiratory disturbances.
in vivo positron-emission tomography (PET) These symptoms have been induced in labora-
scanning. Additionally, low normal receptor tory animals exposed to elevated endorphin
binding instead of dopamine receptor supersen- levels. Several research groups have therefore
sitivity was reported (in direct opposition to studied the ß-endorphin system in females
the results reported by Nomura et al., 1985). with Rett syndrome. Budden, Myer, and Butler
Similar findings are reported in a second study (1990) found elevated ß-endorphin immunoreac-
(Riederer et al., 1986), where preliminary bio- tivity in the cerebrospinal fluid of 11 out of the
chemical analyses on plasma, urine, cere- 12 girls studied. These findings were extended
brospinal fluid, and postmortem brain areas as elevated cerebrospinal fluid ß-endorphins
indicated no disturbance of neurotransmitter were reported in 90% of more than 150 persons
function. These researchers suggest various with Rett syndrome (Myer, Tripathi, Brase, &
drug therapies administered to the girls prior to Dewey, 1992). The degree of elevation of the ß-
sample testing and undernutrition (a problem endorphins, however, did not correlate with the
common to girls with the Rett syndrome) might severity of the symptoms (e.g., stereotypy,
influence the synthesis and turnover of these breathing disturbance) or stage of the disorder.
biogenic amines. An alternative hypothesis Contradictory results, however, have been re-
suggests that such a deficit might also be trig- ported (Genazzani, Zappella, Nalin, Hayek, &
gered as a primary consequence of the disease Facchinetti, 1989; Gillberg, Terenius, Hagberg,
process. Efforts to understand the often contra- Witt-Engerstrom, & Eriksson, 1990). These re-
dictory results are complicated further by the searchers report significantly lower levels of ß-
differing laboratory procedures employed by endorphins in girls with Rett syndrome
investigators throughout the world. compared with age-matched controls. As in the
Rett syndrome appears to share neurochemi- biogenic amine studies reported earlier, the
cal features (without the associated neuro- basis for these contradictory findings remains
pathological features) with some age-related unknown.
neurodegenerative diseases such as Alzheimer’s Perhaps the most promising lead for a neuro-
and Parkinson’s diseases. Wenk and his associ- chemical marker for Rett syndrome has devel-
ates (Wenk, Naidu, Casanova, Kitt, & Moser, oped from a study of five autopsied cases
1991; Wenk, O’Leary, Nemeroff, Bissette, exploring brain and cerebrospinal fluid glycol-
Moser, & Naidu, 1993) have reported decreased ipids (Lekman, Hagberg, & Svennerholm,
cortical and subcortical levels of choline acetyl- 1991). The concentrations of two major brain
Rett Syndrome: A Pervasive Developmental Disorder 143

gangliosides, GD1a in frontal gray matter and syndrome (Zappella, 1990; Zappella & Genaz-
GD1b in temporal gray matter, appear to be zani, 1986; Zappella, Genazzani, Facchinetti,
lowered selectively in the cerebral cortex and & Hayek, 1990). Improvements in communica-
cerebellum in girls with Rett syndrome. The tion and a decreased frequency of agitation
ganglioside GD1a is thought to play an impor- episodes were initially reported; however, these
tant role in synaptogenesis since it is prominent improvements failed to recur following the
in synaptic membranes and high concentrations “ washout ” phase of the study. Uncontrolled
are reported during the time when nerve ending studies exploring the effect of other anti-
growth and synapse formation are most intense Parkinsonian drugs aimed at the monoamine
(25th fetal week until age 2 years). Ganglioside system (e.g., L-dopa, pergolide, deprenyl) have
GD1b is rich in axons and accrues more slowly, been undertaken but have failed to provide evi-
reaching maximum concentration at age 20 dence of improvement on a consistent basis.
years. The reduction in GD1a would help ex- L-dopa and Sinemet (DuPont) have been re-
plain the pathogenic findings of decreased den- ported to show benefit for a limited number
dritic arborization reported by Armstrong of patients in the later stages of the disorder
(1992). These findings appear to be specific for when increasing rigidity appeared (Percy &
Rett syndrome; however, replication of a larger Hagberg, 1992). Tetrabenazine, a monoamine
series is required to validate the results. depleter and blocker, resulted in an exacerba-
Synapses within the cerebral cortex, basal tion of symptoms in one patient (Sekul & Percy,
ganglia, and brain stem responsible for move- 1992). Egger, Hofacker, Schiel, and Holthausen
ment and breathing employ the excitatory amino (1992), reported that magnesium orotate or cit-
acid neurotransmitter glutamate. Glutamate- rate (4 to 10 mg/ kg/day) initially given as an
mediated neurotransmission appears to be dis- anticonvulsant (after more traditional anticon-
rupted in persons with Rett syndrome. Two vulsants had failed) resulted in a decrease in
studies (Hamberger, Gillberg, Palm, & Hag- hyperventilation in a girl with Rett syndrome.
berg, 1992; Lappalainien & Riikonen, 1996) They extended their findings to six additional
of the cerebrospinal fluid in girls with Rett patients with Rett syndrome. A decrease in hy-
syndrome report elevated levels of glutamate perventilation was reported in all girls and par-
in young children and significantly reduced ents reported a decrease in their daughters’
levels in older individuals. A study using MR hand stereotypies and episodes of agitation.
spectroscopy (Pan, Lane, Hetherington, & Convulsions were reduced in four of the girls.
Percy, 1999) also replicated these findings. Serum magnesium levels were normal for all
This would suggest that the synaptic levels of patients prior to the start of the treatment, sug-
glutamate may be elevated and could account gesting that the magnesium was acting pharma-
for the cortical hyperexcitability, seizures, cologically rather than correcting a deficit. The
and possibly dendritic pathology and abnor- researchers suggest that the magnesium is coun-
malities in Rett syndrome. teracting intracellular lactic acidosis and serv-
ing as a N-methyl-D-asparate channel blocker,
DRUG THERAPY thus reducing excitotoxic neuronal damage.
A randomized double-blind controlled
To date, there is no cure for the Rett syndrome, crossover trial of L-carnitine (a natural amino
and therapeutic interventions directed at the acid that breaks down long-chain fatty acids in
fundamental mechanisms underlying Rett syn- the blood transporting the particles to the mi-
drome, while limited in number and scope, tochondrial membrane for increased energy
have failed to demonstrate any lasting or sub- production) was performed with 35 girls with
stantive improvements. Rett syndrome (Ellaway et al., 1999). The
study reported improvements in eye contact,
Treatment of Underlying Causes concentration and attention span, reduced day-
time somnolence, increased vocalization, and
Based on the biochemical findings in the re- increased mobility; assessments by both med-
search, the effect of bromocriptine, a dopamine ical personnel and caregivers indicated that
agonist, has been explored in 12 girls with Rett the girls appeared happier.
144 Diagnosis and Classification

Neurotransmitter precursor therapy has (p. 238) during sleep. Robb, Harden, and Boyd
been attempted with the amino acids tyrosine (1989) report, that in their study of 52 girls
(the precursor for dopamine and norepineph- with Rett syndrome, discharges, consisting of
rine) and tryptophan (precursor for serotonin) short waves or spikes, were a common feature.
in nine girls (Nielsen, Lou, & Andresen, 1990). These discharges could be infrequent or almost
No clinical performance or EEG pattern continuous and characteristically were most
changes were observed as a result of treatment. prominent around the middle third of the head.
A double-blind controlled crossover trial of Pronounced EEG abnormalities were most
melatonin (a natural hormone secreted by the often found from 3 to 10 years of age and
pineal gland to help promote drowsiness) was tended to become less severe during the sec-
conducted by McArthur and Budden (1998) to ond decade of life (Niedermeyer et al., 1986;
explore its efficacy on sleep disturbances as- Rett, 1986). Glaze and associates (1987), de-
sociated with Rett syndrome. They report the scribed the progressive changes in the EGG
melatonin decreased sleep onset latency, im- and correlated it with the clinical staging sys-
proved total sleep, and improved sleep effi- tem. Their work is summarized in Table 5.6.
ciency—especially in those subjects with the The EEG changes generally appear at the be-
most disturbed sleep patterns during baseline. ginning of Stage 2 and then follow a stepwise
Naltrexone, an opiate antagonist, was em- progression, with slowing, loss of normal sleep
ployed in a double-blind crossover trial (Percy characteristics, multifocal abnormalities, and,
et al., 1991). The use of an opiate antagonist finally, generalized slow spike and wave activ-
was attempted due to the reports of elevated ity (Verma, Chheda, & Nigro, 1986).
levels of ß-endorphins in girls with Rett syn- Although all girls with Rett syndrome
drome. The motor behavior and other symp- demonstrate abnormal EEG tracings, seizure
toms of the disorder displayed no improvement activity is not universal. Naidu et al. (1986) re-
during the naltrexone treatment phase. In fact, ported that approximately 84% of the girls in
the girls’ performance on the Bayley Scales of their study demonstrated seizures. The most
Infant Development worsened during the phase common types of clinical seizures include
of naltroxone treatment (Percy & Hagberg, generalized tonic-clonic and partial complex
1992). Given the lack of a biochemical marker seizures. Infantile spasms with hypsaarryth-
and a limited understanding of the biological mia, however, may be an early symptom
basis of Rett syndrome, it may not be surpris- (Iyama, 1993). Selection of a specific medica-
ing that drug therapy has not proven particu- tion should be based on clinical seizure type
larly effective. and EEG pattern. Several clinicians (Adkins,
1986; Budden, 1986; Naidu et al., 1986; Philip-
EEG Profile and Seizure Control part, 1986) agree that standard dosages of
Tegretol (carbamazepine) constitute the best
At this time, we must be satisfied to provide seizure management program. Adrenocorticotr-
suitable medication aimed at symptomatic phic hormone or prednisone has been helpful
relief (e.g., seizure activity). The electroen- in treating infantile spasms (Sekul & Percy,
cephalogram (EEG) has been demonstrated to 1992). Hagberg (1985) warns, however, that
be significantly abnormal for persons with Rett many girls with Rett syndrome overreact and
syndrome throughout all but the earliest stage must therefore be taken off the medication.
of the disorder. A study based on the EEG Haas and his associates (1986) have employed
records of 44 persons with Rett syndrome found the ketogenic diet to reduce seizures in girls
abnormal EEG tracings to be almost universal with seizures failing to respond to medications.
(Niedermeyer, Rett, Renner, Murphy, & Naidu, As the girls enter late adolescence, seizure ac-
1986). Abnormal sleep patterns have also been tivity may decrease allowing modification of
noted. Haas, Rice, Trauner, and Merritt (1986) their medication regimen. Staring spells, eye
report the “presence of intermittent episodes of rolling, and other episodic behavior may be ob-
high amplitude bursts of spike wave or slow served in individuals with Rett syndrome and
wave discharges followed by a brief period of do not always indicate seizure activity (Garo-
relative suppression of background activity” falo, Drury, & Goldstein, 1988). Therefore,
Rett Syndrome: A Pervasive Developmental Disorder 145

TABLE 5.6 Correlation of EEG Characteristics with Clinical Stages

EEG Characteristics

State Stage 1 Stage 2 Stage 3 Stage 4

Awake Normal or minimal Marked slowing of Further, gradual Absence of occipital-

slowing of occipital- occipital-dominant slowing of occipital- dominant rhythm;
dominant rhythm and rhythm and back- dominant rhythm, with marked slowing of
background activity ground activity; rare its subsequent disap- background activity
focal spike or sharp- pearance; moderate- (delta frequencies);
wave discharges to-marked slowing of multifocal spike and/or
background activity; sharp-wave discharges
appearance of or generalized slow
multifocal spike and spike-wave pattern
sharp-wave discharges;
during latter part of
this stage, appearance
of generalized slow
spike-wave pattern
Asleep Normal, with well- Less well-defined Absent vertex Almost continuous
defined vertex trans- vertex transients and transients and spindles generalized slow
ients and sleep spindles and subsequent during NREM sleep; spike-wave activity
spindles loss of these sleep multifocal spike
characteristics; ap- and/or sharp-wave
pearance of focal or discharges, with later
multifocal spike and/ development of
or sharp-wave dis- generalized slow
charges spike-wave pattern
during NREM sleep

Source: “Rett Syndrome in the Electroencephalographic Characteristics with Clinical Staging,” by D. G. Glaze,
J. D. Frost, H. Y. Zoghbi, & A. K. Percy, 1987, Archives of Neurology, 44, pp. 1053–1056. Copyright 1987, Amer-
ican Medical Association.

these behaviors should not be treated as benefit from adequate caloric intake must be
seizures without EEG documentation. considered when examining the malnutrition
displayed in some persons with Rett syndrome
GROWTH PATTERNS AND NUTRITION (Missliwetz & Depastas, 1985). Motil, Schultz,
Brown, Glaze, and Percy (1994) report that en-
A pattern of deceleration across all growth ergy expenditure associated with involuntary
measurements, following the first 6 months of motor movement places these girls in a situa-
life, is witnessed in most persons with Rett syn- tion of lower energy balance (energy intake ver-
drome (Schultz et al., 1993). The exact cause of sus expenditure) than controls. The lowered
this remains unknown. A systemic deficiency energy balance in girls with Rett syndrome par-
in mitochondrial energy production as re- alleled their degree of height and weight
ported from muscle biopsy tissue has been deficits, despite similar dietary energy intakes
suggested (Coker & Melnyk, 1991; Schultz between groups. Haas and his associates (Haas
et al., 1993). The best evidence to date, how- & Rice, 1985; Haas et al., 1986) have reported
ever, suggests nutritional, rather than chromo- improved weight gain in conjunction with di-
somal, neurological, or hormonal factors minished stereotyped behavior and better
underlie this failure to grow. Some evidence seizure control with the implementation of a
suggests defects in carbohydrate (Clark et al., high-calorie, high-fat ketogenic diet.
1990; Haas & Rice, 1985; Haas et al., 1986), Reilly and Cass (2001) explored the issue of
ascorbic acid, and glutathione (Sofic, Ried- growth failure in Rett syndrome. They paid spe-
erer, Killian, & Rett, 1987) metabolism. Mal- cial attention to the impact of the feeding prob-
absorption of critical nutrients and failure to lems experienced by many of these individuals.
146 Diagnosis and Classification

Through their own clinical experience and the measure of cognitive ability or of the motor and
results of a questionnaire, these researchers verbal disability. Thus, the assessment of cogni-
identified several common feeding problems tive functioning in persons with Rett syndrome
(see Table 5.7). They have developed a series is particularly problematic.
of management protocols involving dietary sup- Given the effect of Rett syndrome on verbal
plementation, food texture modifications, and and motor behavior, assessment employing tra-
posture modifications. A nutritionist should be ditional standardized measures of cognitive
available to consult with parents and program ability is counterindicated. These measures
staff relative to diet if weight gain is a problem. would be invalid because they would generate
Constipation is a common problem experi- scores that would be indicative of the verbal
enced by persons with Rett syndrome. Many of and motor disability instead of cognitive abil-
the girls fail to consume adequate fluids and ity. For this reason, the cognitive abilities of
fiber which may result in an impacted bowel girls with Rett syndrome traditionally have
(Hunter, 1987). Dietary measures (ingestion been assessed with instruments designed for
of fiber, mineral oil, fruit with high liquid con- evaluating infants. These measures can be ad-
tent, etc.) may prove adequate, although artifi- ministered to persons with very little verbal
cial laxatives, enemas, or suppositories are and motor ability. Generally, however, they re-
often required (Naidu et al., 1990). quire caregivers and others familiar with the
child to make judgments about the child’s abil-
COGNITIVE AND ity and developmental functioning level to
ADAPTIVE FUNCTIONING support data collected through direct observa-
tion. Many of these measures allow flexibility
Rett syndrome is characterized by an especially in the materials employed in testing to maxi-
debilitating combination of an extrapyramidal mize child interest. Thus, these assessments
movement disorder, delayed response latencies, are thought to provide a more accurate picture
and the loss of acquired speech. Such a combi- of developmental functioning than traditional
nation of disabilities significantly limits our IQ tests. Although using infant assessments for
ability to estimate the cognitive functioning of older children with severe impairments is a
persons with this disorder. Traditional methods common practice, these tests were normed and
of cognitive assessment require either unim- standardized on infants. Thus, the score at-
paired motor or verbal responses from the indi- tained might best be interpreted as an estimate.
vidual to generate valid estimates. If the person Along those lines, Demeter (2000) cautions
is required to employ a means of responding that when assessing individuals with severe or
that in itself is impaired, we can never be profound disabilities such as those found in
certain whether the assessed functioning is a Rett syndrome, a standardized instrument is
unlikely to measure the full range of a child’s
TABLE 5.7 Feeding Problems Common in ability. Noting that discrepancies exist be-
Persons with Rett Syndrome tween parental reports of ability and the abil-
Inability to consume adequate calories orally to meet
ity levels indicated by test results, Demeter
energy requirements. suggests that persons with Rett syndrome
Evidence of ongoing aspiration during oral feeding.
often show situation-specific skills that tradi-
tional measures miss. Some children can and
Oral feeding is stressful for the caregiver, the indi-
vidual with Rett syndrome, or both.
do learn to interact within their environment to
get their needs met, yet this capacity may be
Mealtimes are protracted, leaving limited time for
other daily activities.
missed by assessment or assumed to be non-
existent. Demeter suggests that when attempt-
Oral intake is erratic.
ing to assess a child with Rett syndrome, re-
Oral supplementation has failed.
searchers should use criterion-referenced tests
Chronic food/ liquid refusal or aversive behavior dur- to measure a child’s ability to learn new skills
ing mealtime has developed.
as well as multiple observations in natural en-
A safe route for providing regular medication is re- vironments. Further, because children with
quired (e.g., with food or liquid intake).
Rett syndrome are highly social, assessment
Rett Syndrome: A Pervasive Developmental Disorder 147

should use social interests or people versus ob- the student is aware that she should activate a
jects to elicit responses. switch but does so randomly)? Further, the
Physical and neurological impairments in- small amount of stimuli provided as distracters
herent in Rett syndrome (e.g., apraxia) may im- in measures such as the PPVT-R allow for a
pede a person’s ability to act on or respond to 25% chance that the student will select the cor-
stimuli in the consistent manner often necessi- rect answer. Administration of assessment in-
tated in standardized testing situations. As a struments to persons with severe disabilities
result, scoring should be completed in a way typically requires that the individual be famil-
that does not penalize these individuals for in- iar with the response capabilities of the person
stances when there is no response. Test admin- being tested. Problems can arise when the ad-
istrators should take into account the child’s ministrator scores an item as correct based on
successful attempts divided by the total number a vague implied response or assumed capabili-
of actual attempts, omitting the times the child ties based on previous knowledge of the child.
was nonresponsive. For example, if a child was Another consideration when assessing the
asked to point to a particular picture and was cognitive abilities with Rett syndrome is visual
only able to do so 6 of the 10 times requested, acuity. In a 1996 study, von Tetzchner et al.,
and got 4 of the 6 attempts correct, then the used the Fagen Test of Infant Intelligence to as-
percentage would be calculated using 4⁄6 as the sess visual function and visual information
raw score instead of 4⁄10. The end result is an ex- processing in a group of girls with Rett syn-
ploration of the ratio of correct to incorrect re- drome and then compared them with a control
sponses free of trials in which the movement group of age-matched typically developing
disorders associated with Rett syndrome pre- children. Results indicated that all the girls in
cluded item selection. the study (n = 41) displayed a visual acuity
One assessment tool suggested by the Inter- below what would be expected in fully devel-
national Rett Syndrome Association (IRSA, oped vision. The authors suggest this may be
2002) for assessing cognitive ability in persons due to syndrome-related arrested cortical and
with Rett syndrome is the Peabody Picture Vo- retinal development or attentional deficits that
cabulary Test-Revised (PPVT-R). The PPVT-R cause misleading results. Either way, the au-
requires that students point to the one picture thors suggest that persons with Rett syndrome
that the test administrator verbally labels, out need additional time to process information be
of four displayed. The IRSA suggests a modi- it novel or familiar. Additional response time
fied test protocol (enlarging pictures and must be incorporated into any assessment in-
spreading them farther apart to ease the inter- volving children with Rett syndrome.
pretation of eye-gaze responses, testing over Much of the research on Rett syndrome sug-
multiple days to guard against “down days” and gests that the individuals function within the
possible misinterpretation) to assist in provid- severe-to-profound range of mental retardation.
ing an accurate picture of the child’s skill level. Perry, Sarlo-McGarvey, and Haddad (1991)
One must note, however, that the PPVT-R “. . . employed the Cattell Infant Intelligence Scale
is not a comprehensive test of general intelli- (Cattell, 1940) to assess the cognitive func-
gence, instead it measures only one important tioning of 15 girls with Rett syndrome. Each of
facet of intelligence: vocabulary” (Dunn & the girls tested below the 8-month cognitive
Dunn, 1981, p. 2). Like other measures em- functioning level, with an average of 3 months,
ployed to assess the cognitive ability of per- suggesting profound cognitive deficits. There
sons with Rett syndrome, the validity of the is, however, considerable debate about whether
scores attained must be explored further. children with Rett syndrome experience a true
Assessment measures that employ eye gaze dementia involving cognitive degeneration that
or switch activation may be useful for persons results in severe or profound mental retar-
with Rett syndrome. The question remains dation (e.g., Budden, Meek, & Henighan, 1990;
whether the score attained could be replicated Charnov, Stach, & Didonato, 1989; Rolando,
if the same measure were administered again: 1985) or a cognitive arrest or stagnation at the
Does the test measure cognitive ability or sim- point of the initial motor and language regres-
ply provide an opportunity for responding (e.g., sion (e.g., Fontanesi & Haas, 1988; Kerr et al.,
148 Diagnosis and Classification

1987; Naidu et al., 1986). Charnov and her as- Development (1975), they found the girls re-
sociates (Charnov et al., 1989) evaluated the sembled one another and remained relatively
developmental histories of 16 girls with Rett stable over the 3 years in the sensorimotor do-
syndrome through parent interview (of prere- mains of Object Permanence (Piagetian Stages
gression development) and current functioning I and II ), Vocal Imitation (Piagetian Stages 0
assessment employing the Birth to Three De- and I ), Gestural Imitation (Piagetian Stages 0
velopmental Scale (Bangs & Dodson, 1979). In and I ), and Scheme Actions (Piagetian Stages
all developmental areas, they report that the II and III ). There was considerable individual
current functioning of each subject was sub- variation among the girls in relation to the
stantially below that which the children were sensorimotor domain of Means Ends Abilities,
attributed to have achieved prior to the onset of ranging from Piagetian Stages I to IV. The
the disorder. Interestingly, the developmental girls also demonstrated the greatest level of
skills profile at the time of testing mirrored individual improvement over the 3-year study
that of the preregression skills, although at a (from Piagetian Stage I to Stages II and IV for
lower level of functioning. three of the girls).
There is some support for a hypothesis that Current best estimates suggest that persons
Rett syndrome results in cognitive arrest or with Rett syndrome function within the severe-
stagnation at the developmental level achieved to-profound range of mental retardation follow-
at the age of onset of the condition in combina- ing the regression in Stage 2 of the disorder.
tion with a severe extrapyramidal movement and The areas of gross motor activity and daily liv-
expressive language disorder as opposed to cog- ing skills appear to be more advanced than
nitive dementia (Hagberg & Witt-Engerstrom, other adaptive functions and the girls are ca-
1986; Kerr & Stephenson, 1986; Stephenson & pable of some cognitive improvement over
Kerr, 1987). The extent and timing of the cogni- time. Further research is needed to gain a
tive stagnation may be the result of the differen- fuller understanding of the cognitive abilities
tial timing of the genetic mutation and the of girls with Rett syndrome. Multidisciplinary
number and location of the cells impacted with assessments of children with Rett syndrome
the defective MECP2 gene (Hoffbuhr et al., are required to gather information on factors
2001). Fontanesi and Haas (1988) evaluated the that may influence cognitive development and
cognitive functioning of 18 girls with Rett syn- performance. Inconsistency of response in
drome. They administered the Vineland Adap- the test situation is a common report across
tive Behavior Scales and the Bayley Scales of the studies. This makes standardized testing
Adaptive Abilities in addition to examining difficult, and care should be taken not only
medical and developmental histories and inter- to provide standardized administrations of
viewing parents regarding the age at which their cognitive tests, but also to look across the
daughters attained developmental milestones. daily functioning of the individuals to identify
Their results indicated that “skills not depen- actions that indicate target abilities. For ex-
dent on either language or fine motor function ample, in one of our studies (Van Acker &
are retained at a developmental level equivalent Grant, 1995), one of the girls failed to demon-
to the age of onset ” (p. S23). Gross motor func- strate awareness of a covered object during
tioning, daily living skills, and object perma- testing, suggesting a lack of object perma-
nence were found to be relatively preserved, nence. Later in the testing situation, however,
whereas fine motor control and language func- she dropped a toy and it rolled under the table.
tioning displayed substantial degeneration from She repeatedly looked under the table for the
preregression developmental levels. missing item.
In a longitudinal study of the cognitive
skills in six girls with Rett syndrome, Woody- COMMUNICATION
att and Ozanne (1993) reported patterns of
similarity across the group as well as individ- Communication abilities and subsequent pro-
ual variations across subjects and within gramming for persons with Rett syndrome have
subjects over a period of 3 years. Using the Uz- been the topic of very limited empirical re-
giris and Hunt Scales of Infant Psychological search. Prior to regression, most of the girls
Rett Syndrome: A Pervasive Developmental Disorder 149

with this disorder are reported to have devel- facial and gestural communicative behaviors
oped single spoken words and/or word combi- and developed a “dictionary” for those who in-
nations. Comprehension skills appropriate to teract frequently with their child (Interna-
the child’s age also are typically noted (Bud- tional Rett Syndrome Association, 1990).
den, Meek, et al., 1990; Woodyatt & Ozanne, Until recently, persons with multiple dis-
1992). The loss of language skills during re- abilities, such as those displayed in Rett syn-
gression is sufficiently rapid to be mistaken for drome, would have been deemed unlikely
hearing loss. Following the regression phase of candidates to benefit from formal communica-
the disorder, speech and language skills are ob- tion intervention. In the past, the literature has
served to be severely impaired. Verbalizations asserted a powerful relationship between cog-
are typically nonexistent or limited to “non- nition and language development. Intentional
functional” consonant-vowel combinations, ex- communication was thought to require perfor-
cept in the estimated 2% to 4% of girls who mance at Piaget’s (1929) sensorimotor devel-
fall within the preserved speech variant of Rett opment Stage V (Bates, Benigni, Bretherton,
syndrome (Budden, Meek, et al., 1990; Skjel- Camaioni, & Volterra, 1979; Bates, Camaioni,
dal, von Tetzchner, Jacobsen, Smith, & Heil- & Volterra, 1975; Kahn, 1977, 1981, 1984).
berg, 1995; Zappella, 1992). Girls with the Formal testing suggests that girls with Rett
preserved speech variant may retain full sen- syndrome function at a presymbolic language
tence use and may even slowly show an in- level (Woodyatt & Ozanne, 1992, 1993).
crease in vocabulary (von Tetzchner, 1997). Again, one cannot determine whether these re-
However, for the most part, these girls employ sults relate primarily to cognitive deficits or to
their language in a nonfunctional manner. They an expressive language and/or motor disorder.
continue to display significant communication Numerous anecdotal reports however, suggest
impairment with pronounced pragmatic diffi- that these girls may well understand more than
culties (Gillberg, 1997). As the girls approach they can express (Weisz, 1987). Moreover, the
adolescence, improved eye contact, social in- assumption that intentional communication re-
teraction, and communicative intentionality quires Stage V functioning has been called
are reported (Woodyatt & Ozanne, 1993). into question (Reichle & Keogh, 1986; Rice,
Individuals with Rett syndrome become a 1983).
part of the estimated one million children and Persons with Rett syndrome, in fact, have
adults in the United States who are unable to been taught to employ augmentative com-
communicate orally (Diggs, 1981). Most suc- munication systems that involve eye pointing,
cessful communication programs for these communication boards (pictures), facial
girls take advantage of their limited commu- expressions, gestures, and the activation
nicative behaviors (e.g., Donnellan, Mirenda, of switches or computers to indicate some
Mesaros, & Fassbender, 1984). Vocalizations, limited volitional gestures. Van Acker and
facial expressions, gestures, walking toward a Grant (1995) employed a dynamic computer-
desired item or activity, and eye gaze are com- graphic presentation contingent on touch-
mon communicative behaviors displayed by screen activation with three girls to select
persons with Rett syndrome. Parents and edu- desired food items. Results indicated that
cators must attune themselves to their child’s two of the three girls demonstrated a reduc-
communicative behavior and respond contin- tion in stereotypic hand use and significantly
gently to these signals. As with any child, the increased item requesting when provided
critical element in the development of a mean- computer-based requesting instruction. Inter-
ingful communication system depends on the estingly, cognitive assessment of all three
contingent interaction of the person and her en- girls indicated a functioning level below the
vironment (Lewis, Alessandri, & Sullivan, Piagetian Sensorimotor Stage V, yet benefit
1990). When others learn to detect the commu- from instruction was obvious.
nicative behaviors and to respond to them in a Similarly, Hetzroni, Rubin, and Konkol
systematic fashion, a formal and effective sys- (2002) were able to teach three girls with Rett
tem of communication can be developed. Par- syndrome to match picture symbols (PCS or
ents have taken photographs of their daughter’s orthography) or picture symbols plus text to
150 Diagnosis and Classification

the spoken request to “ touch .” Their addressing symptoms by maintaining or im-

results suggested the girls were able to demon- proving functional movement, mobility, pre-
strate a steady learning curve over the symbols venting deformities, and keeping the girls in
employed and to display a partial retention of contingent contact with their environments.
these symbols over time. During the periods of regression associated
Speech and language services appear to be with the disorder, therapeutic intervention is
warranted for persons with Rett syndrome. Pro- especially important and should be more fre-
gramming should emphasize functional recep- quent, because transition skills are at risk
tive and expressive language skills, as well as (Hanks, 1990). Although persons with Rett
cognitive, social pragmatic, and affective com- syndrome display numerous similarities, their
munication skills. Early intervention could specific therapeutic problems and responses to
target eye contact and attending to the environ- treatment vary dramatically (Hanks, 1986,
ment. Cause-and-effect relationships that allow 1990; Lieb-Lundell, 1988; Sponseller, 1989).
the children to gain an understanding of their The therapeutic intervention program, there-
ability to affect their environment are essential fore, must be highly individualized.
for communication training. Though Rett Apraxia and ataxia are frequently the earli-
syndrome affects a person’s ability to attend to est manifestations of motor problems in Rett
objects and stimuli, seemingly reducing the syndrome. Hypotonia interferes with postural
capability to discriminate between items and stability. Fitzgerald, Jankovic, Glaze, Schultz,
therefore making communication intervention and Percy (1990) suggest that the typical
difficult, there is evidence that children with “jerky truncal tremors” may reflect a derange-
Rett syndrome can sustain attention in pre- ment of postural reflexes rather than cerebel-
ferred activities for longer than average periods lar pathology. The girls develop compensatory
of time (J. Watson, Umansky, Marcy, & Repa- increased tone to achieve stability, resulting in
choli, 1996). Implications of this research sup- abnormal movement patterns. A marked fixing
port the idea that interventions should strive or locking of their joints into positions of sta-
to identify and use activities, people, and items bility to counter disruption in balance is typi-
preferred by the child to enhance motivation cal, inhibiting their ability to shift positions.
to learn. Simple switches to activate toys or Thus, the legs are often kept in wide abduction
computer monitors may enhance an awareness while sitting and standing (see Figure 5.2) and
of cause-and-effect relationships. Musical in- weight shift is absent (Hanks, 1986).
struments also have been shown to promote the The girls often demonstrate expressions of
child’s desire to interact with her environ- agitation and fear in response to any movements
ment (Wesecky, 1986; Zappella, 1986). Given
the importance of communication skills, re-
searchers need to identify the range of commu-
nication functioning in, and effective methods
to teach improved communication skills to, per-
sons with Rett syndrome.


Persons with Rett syndrome exhibit multiple

orthopedic and motor movement disorders.
These disorders can vary significantly across
the different phases of the syndrome. Physical
and occupational therapists, therefore, must
play an important role in the care of individu-
als with Rett syndrome (Braddock, Braddock,
& Graham, 1993; Hanks, 1986, 1990). Inten-
sive therapy, while failing to alter the actual Figure 5.2 Stereotypic hand movements in Rett
course of the disease, has been successful in syndrome.
Rett Syndrome: A Pervasive Developmental Disorder 151

that are not self-initiated. Similar voluntary respiratory dysrhythmia ( hyperventilation and
movements, however, are not related to these apnea; Kerr et al., 1990) exacerbated stereo-
stress reactions. Lieb-Lundell (1988) reports, typic movements. Increased stereotypic hand
“No amount of practice or exposure alters this movements in Rett syndrome have been re-
response of fear to extrinsically initiated ported to coincide with improved EEG trac-
movement ” (p. 533). ings. Niedermeyer and Naidu (1990) have
Several therapeutic interventions have suggested that this passive finger movement
been found to be successful in the treatment of may block focal and multifocal spike activity.
apraxia-ataxia. Tone reduction techniques simi- This phenomenon has been reported previously
lar to those used with patients afflicted with in Rolandic epilepsy (Sekul & Percy, 1992).
cerebral palsy or impaired through a stroke are Despite their neurological origin, the
appropriate. Interventions might include (1) use stereotypic behaviors of persons with Rett syn-
of the therapy ball, (2) balance-stimulating drome appear to be influenced, at least mini-
floor activities, (3) segmental rolling, and mally, by environmental stimuli. An analogue
(4) rotation and weight shift activities (Hanks, study of the stereotypic behavior of two young
1986, 1990). Vestibular movement activities women with Rett syndrome demonstrated that
(e.g., merry-go-rounds, swings) have also been their moment-to-moment expression appeared
reported as helpful if the child will tolerate this to be influenced by environmental contingen-
intervention (Havlak & Covington, 1989). cies (Wehmeyer, Bourland, & Ingram, 1993).
Efforts to facilitate normal movement They found that both participants increased
while maintaining reduced tone are usually not or decreased their level of stereotypic re-
successful, however (Hanks, 1990). As the sponding as the result of various consequences
girls often resist being moved, close physical within their social setting. For example, one of
contact and a slow, firm approach aimed to the girls would demonstrate a significant in-
reduce the child’s anxiety during physical as- crease in stereotypy when placed in a task de-
sistance is suggested. Individuals with Rett mand situation. The authors speculated that
syndrome often have very long response la- escape and/or avoidance of task engagement
tency. Therapists must provide verbal direc- was negatively reinforcing her stereotypy. Van
tions and encouragement and provide the time Acker (1987) reported a 40% reduction in the
needed for a response. stereotypic hand movements displayed by one
Stereotyped hand movements represent one girl when involved in computer-assisted in-
of the most distinguishing characteristics of struction. These findings were extended in a
the Rett syndrome. Hand wringing, hand wash- later study where the stereotypic behavior of
ing, hand tapping, and hand-to-mouth move- two additional girls displayed a significant de-
ments are common stereotypies resulting in crease during computer-assisted requesting.
the loss of purposeful hand function. These Though stereotypic hand movements are an
movements appear to evolve with age, proceed- intrinsic part of Rett syndrome, there is some
ing from simple, rapid movements to a slower, evidence that girls with the preserved speech
more complex form and ultimately to slow, variant of Rett syndrome retain more functional
less complicated repetitive movements (Clare, hand use than girls with classical Rett syn-
1986). Most researchers consider the stereo- drome. The preserved speech variant is usually
typic hand movements to represent primary associated with a higher ability level in the cog-
circular reactions resulting from an underlying nitive, speech, and motor domains. Umansky
extrapyramidal disorder. Thus, they represent et al. (2001) suggest that differing levels of
nonvolitional movements. Persons with Rett MECP2 protein function may account for dif-
syndrome appear to require great concentra- ferences in object-oriented hand use in girls
tion and effort to break out of the stereotypic with the preserved speech variant. Further-
movements for even brief attempts at purpose- more, as a result of an intervention study con-
ful hand use. ducted with one 6-year-old girl with the
Stereotypic behaviors are more remarkable preserved speech variant of Rett syndrome,
under stressful situations and diminish or dis- Umansky et al. believe that social facilitation
appear momentarily when changing posture of object-oriented hand use is an important fac-
or eating. In some individuals, periods of tor in preservation of this function. Outside
152 Diagnosis and Classification

encouragement of hand use contributes to maintenance of treatment effects has not been
a child’s ability to retain hand movements documented. Treatments with various medica-
and skill. tions such as L-dopa, haloperidol, 5-HTP, and
The suspected neurophysiological etiology various anticonvulsants have proved unreward-
of stereotypy in Rett syndrome has led re- ing thus far (Percy et al., 1985).
searchers to caution practitioners attempting Splinting has been found to be successful in
to modify these behaviors. As these behaviors interrupting hand-to-mouth (Hanks, 1986) and
appear to represent basically involuntary hand wringing (Aron, 1990; Naganuma, 1988)
movements, efforts to change this behavior, es- movements, thus allowing the girls to direct
pecially through the use of aversive conse- their attention to tasks and persons in their en-
quences, seem ill advised (Hanks, 1990). vironment and to reduce the risk of skin break-
Moreover, previous attempts to use operant down related to these high-rate behaviors. In
conditioning procedures to alter symptomatic fact, some persons with Rett syndrome have
behaviors of persons with Rett syndrome have demonstrated improved functional hand use
not proven particularly promising. For exam- while splints were in use (Naganuma, 1988).
ple, differential reinforcement, response inter- Tuten and Miedaner (1989), however, were un-
ruption, and contingent hand restraint have able to replicate the effectiveness of hand
been used to decrease self-injurious behavior splints. The two subjects involved in this study
(Iwata, Pace, Willis, Gamache, & Hyman, displayed no decrease in stereotyped hand
1986). They reported that the hand biting of wringing, nor any subsequent increase in func-
these girls, though moderately reduced tional hand use as a result of hand splint appli-
through intervention, appeared to “ be related cation. Sharpe (1992) employed an alternating
to organic predisposition rather than being treatments design to compare the effectiveness
shaped inadvertently by the environment ” of bilateral hand splints and an elbow orthosis
(p. 164). Facial screening has been attempted, for two girls with Rett syndrome. Both girls
without success, to decrease breathing irregu- demonstrated a decrease in stereotypic hand
larities (Lugaresi, Cirignotta, & Montagna, movements and a corresponding increase in
1985). Recently, however, operant approaches toy play with the use of the elbow orthosis.
employing prompting, backward chaining, The bilateral hand splints had no obvious treat-
shaping, and positive reinforcement have ment effect. No studies have demonstrated any
shown success (Bat-Haee, 1994; Piazza, An- maintenance effects of hand splints over time
derson, & Fisher, 1993; Van Acker & Grant, once the splints have been removed. Addition-
1995). Piazza and her associates (1993) were ally, many persons with Rett syndrome are un-
able to reestablish functional self-feeding in able to tolerate the application of hand or arm
three girls with Rett syndrome through the use splints for even short periods.
of graduated prompting and positive reinforce- Instructional strategies that reinforce be-
ment. These researchers suggest that feeding haviors incompatible with stereotypic move-
may represent a good first step in developing ments are recommended. Many of the girls
improved hand use as the girls generally enjoy accept having the stereotypic behavior inter-
eating and the food serves as a natural positive rupted. In fact, they often seem to relax when
reinforcer. They caution parents and practi- their stereotypic behavior is stopped for short
tioners to be aware that the period needed for periods. This can be accomplished by having
skill acquisition may be longer than that dis- the girls placed into the prone position or by
played in other children with severe or pro- simply holding one hand while assisting func-
found disabilities. Moreover, because these tional hand use (see Figure 5.3). Hanks (1986)
girls display variable progress from day to day, reports, “ Toys that combine bright colors and
data collection to plot change over the course sound and require input from the child are
of the training can help parents avoid becom- helpful in keeping the child involved with the
ing discouraged. Thus, operant procedures, es- environment and making attempts to use her
pecially nonaversive approaches, may prove hands” (p. 250). Encouraging the child to use
effective in the treatment of some behavioral her hands for reaching or switch activation can
characteristics of the Rett syndrome; however, potentially reduce stereotypic hand movements
Rett Syndrome: A Pervasive Developmental Disorder 153

skills has been helpful in the improvement of

range of motion and the reduction of discom-
fort (Hanks, 1986; Lieb-Lundell, 1988; Schle-
ichkorn, 1987). Tone reduction activities such
as rotation, weight shift, and vibration have
been reported to result in a temporary reduc-
tion of spasticity (Hanks, 1986, 1990).
Ambulation remains one of the critical
skills to develop and maintain in persons with
Rett syndrome. Many of the girls fail to de-
velop this skill prior to Stage II, while others
Figure 5.3 Interruption of stereotypic behavior to lose this ability as part of the rapid motor de-
promote functional hand use.
generation. As spasticity and apraxia increase,
as reaching for an object and hand-wringing the girls often lose many functional gross
behaviors are incompatible (J. Watson, Uman- motor skills that they had previously achieved.
sky, Marcy, Johnston, & Repacholi, 1996). Additionally, these girls often manifest spatial
Music therapy has also proved useful in the pro- disorientation. Their perception of an upright
motion of functional hand use. The music ap- posture results in a forward, backward, or lat-
pears to increase the level of awareness and the eral leaning. Ambulation may be lost, espe-
instruments motivate efforts to reach out and cially for girls with an orientation toward
interact (Wesecky, 1986). Battery-operated backward leaning, as they are unable to initi-
toys and computers modified to respond to an ate their forward weight shift. They fear
easily activated switch provide an almost limit- falling when attempting weight shift in a for-
less array of possibilities not only to decrease ward direction. The abnormal gait pattern typ-
stereotyped behavior (Van Acker, 1987; Van ically displayed in this syndrome results from
Acker & Grant, 1995) but also to increase a combination of spasticity, ataxia, apraxia,
functional hand use, communication, and cog- compensatory spinal rigidity, and spatial dis-
nitive development (Hanks, 1986; Sponseller, orientation. Asymmetries may develop as one
1989; Zappella, 1986). leg becomes stiffer or weaker. Weight shift is
Bumin, Uyanik, Yilmaz, Kayihan, and accomplished through lateral rocking and
Topcu (2003) report that following 8 weeks of trunk rotation is lost. More simply, because
exposure to the Halliwick method of hy- the center of balance is off, girls often think
drotherapy, an 11-year-old girl with Rett they are falling forward when they are stand-
syndrome displayed increased balance while ing normally. They try to compensate by lean-
walking, decreased stereotypic movements, ing back, fall over, and over time may lose the
and improved purposeful hand use (e.g., im- ability to stand (Kerr et al., 1990).
proved self-feeding). Independent standing and ambulation rep-
Spasticity has been reported as a typical resent a realistic therapy goal, especially in
problem during Stage II of the disorder. This the early years. Many girls while unable to
spasticity may “ vary from a mild increase in walk independently can do so with assistance.
tone in the gastroc-soleus complex, resulting in Such aided ambulation should be encouraged.
toe walking, to severe involvement throughout Weight-bearing exercises, walking, and gait
the body affecting even respiration and swal- training have been successful (Hanks, 1986,
lowing” (Hanks, 1986, p. 248). The resulting 1989, 1990). If appropriate, activities designed
muscle imbalance may lead to severe contrac- to elicit righting and equilibrium responses
tures, especially distally (e.g., a downward might include use of the large therapy ball
pointing of the foot). This spasticity may also (prone positioning and then standing, leaning
be responsible, at least in part, for the high in- forward), weight shifting (seating the child on
cidence of scoliosis in girls with Rett syn- a bench and then tipping the bench slightly
drome (Hanks, 1986; Sponseller, 1989). backward), and ultimately active shift forward
Hydrotherapy emphasizing movement in to come from sitting to standing (Hanks,
the water, range of motion, and basic water 1990).
154 Diagnosis and Classification

Frequently, foot deformities (e.g., ankle Placing the child on her side with the apex of
pronation, plantar flexion inversion and toe the curve down may prove helpful. Exercises
curling) must be corrected with polypropylene designed to maximize use of the muscles the
ankle-foot orthosis, hinged to allow dorsiflex- girl avoids using are in order (e.g., feed and
ion. Attempts to correct pronation foot defor- lead the child by the hand on her hypotonic
mities with below-ankle orthosis has not proven side). Good positi