|  Autism
is a developmental disorder that is evident in infancy or early childhood
and results in a broad range of behaviors and, often, lifelong impairments.
Now conceptualized as a spectrum of related diagnostic categories, autistic
spectrum disorders (ASD) involve severe difficulties in basic aspects
of social behavior and communication, such as eye contact, facial expressions,
and the development of spontaneous, reciprocal language, which, for reasons
not well understood, are associated with behaviors characterized by repetition
and restriction, ranging from finger-flicking to compulsive touching to
complex preoccupations with electrical appliances or highly circumscribed
interests such as names of radio stations or Scottish clans. Such patterns,
which often change with development, occur in individuals who function
intellectually from profound mental retardation to the superior range
of intelligence. Regardless of the level of functioning, all these cases
share basic disturbances in social interaction and communication.
Manifestations
of the severity of the disorder also range from severe to relatively mild.
The spectrum includes classical cases of autism, as defined by the presence
of severe deficits in the three relevant domains, to cases that include
significant social impairment, but without communication disorder or without
repetitive behaviors. There are also differences in developmental trajectories,
both between ASD and within them. Thus childhood disintegrative disorder
is defined by normal development until at least age 2 years, followed
by loss of skills across multiple areas of functioning. Asperger’s
disorder is defined by severe social deficits and restricted interests
or behaviors that are not accompanied by language delay or mental retardation.
Asperger’s disorder is typically diagnosed later; it is not clear
if behavior patterns actually develop later, or if there is later recognition
of the abnormality of these behaviors as they begin to cause impairment
in social situations.
The
assumption, as discussed in more detail below, is that ASD are of neurobiological
origin beginning before birth or in very early development, related to
complex genetics. It is estimated that in about 10% of cases, ASD is associated
with another identifiable medical condition, the most common being tuberous
sclerosis.
Recently,
the epidemiology of ASD has been of particular interest. Estimates of
prevalence, based on school and state health systems, have skyrocketed,
as have demands for services. Until about 1999, well-designed, large-scale
epidemiological studies indicated that the prevalence of autism had increased
from earlier estimates of 4 to 5 per 10,000 to about 7 per 10,000 for
narrowly defined cases and from 2 per 10,000 to 5 per 10,000 for additional,
more broadly defined cases of ASD. However, in the past year or so, there
have been a number of independent, smaller-scale studies that have yielded
even higher prevalence rates for ASD. For example, a recent study in the
United Kingdom reported prevalence rates of nearly 17 per 10,000 for autism
and 63 per 10,000 of all ASD, including autism. Patterns of intellectual
deficits have also changed, from earlier estimates of mental retardation
in association with autism in 75% to 80% of individuals, to the most recent
U.K. study, which reported associated mental retardation in only 26% of
children with ASD. Together, the broadening of the conceptualization of
ASD to include milder difficulties as well as severe social deficits without
mental handicap, better ascertainment, and the expanded number of referrals
of children have converged to result in a very different characterization
of ASD than might have been made even 10 years ago.
In
addition, our understanding of the nature of the social deficits associated
with ASD, particularly in young and school-age children, has become more
focused than it was 50 years ago when Leo Kanner first used the term autism.
Numerous studies have shown that although ASD affects the acquisition
of social relationships, difficulties do not lie in typical measures of
attachment (e.g., strange situation). Rather, they are found in very basic
social-communicative behaviors, such as use of gaze and facial expression;
in specific contexts in which the coordination of these behaviors results
in a “social-cognitive event,” such as pointing to express
interest in something (e.g., “protodeclarative pointing”);
and in reciprocal relationships, most easily seen in the absence of friendships
with peers. Lorna Wing’s characterization of three forms of social
deficits, ranging from aloof to passive to active, but odd, also provided
a conceptualization of a complex deficit that affects social and communicative
behaviors on multiple levels and in different ways for different children,
though there are some strong similarities (e.g., in deficits in eye contact
and facial expression) that appear across age and intellectual level.
Characterizations
of a unique and specific social-communicative deficit associated with
ASD in adolescence and adulthood have been more difficult. When ASD is
associated with cognitive and language delays, individuals with autism,
compared to those with equivalent levels of retardation but without autism,
have more difficulty with both basic and more contextual aspects of social
interaction and relationships as a group. In particular, when ASD is not
associated with language deficit or mental retardation, there are a number
of compensatory “strategies,” as well as comorbidity with
anxiety disorders, depression, and other psychiatric disorders, which
affect the manifestation of the disorder. While standardized diagnostic
instruments such as the Autism Diagnostic Interview-Revised (ADI-R) and
Autism Diagnostic Observation Schedule (ADOS) have been able to operationalize
the definitions of ASD described in DSM-IV and ICD-10,
a current goal is to modify these tools to be able to quantify the severity
of the disorder as a whole, or at least for the primary domains of the
disorder (social reciprocity, communication, restrictive behaviors), independent
of chronological age, developmental level, or degree of mental handicap/language
impairment. An example of this process is the work of Tanguay and Robertson,
identifying factors within the social deficit as measured on the ADI-R
and the ADOS. In our own research, as shown in Figure
1, we have been able to discriminate clearly between children with
nonspectrum disorders and children with various ASD on measures such as
the ADOS social and communication scales. However, there is very significant
overlap among groups of children with clinical diagnoses of pervasive
developmental disorder-not otherwise specified and those with clinical
diagnoses of autism, both with and without mental retardation. A challenge
for the future is to generate scores that are meaningful beyond the broad
differentiations.
Recent
research in neuroimaging and neuropsychology has contributed to a better
understanding of the synergy between social and cognitive models of deficits
in ASD. From recent empirical findings by Schultz indicating different
areas of brain activation for adults with ASD during an observation of
a social situation, to the finding of Dawson and her colleagues of poor
response to their names in infants in whom autism is later diagnosed,
research has highlighted the intersection of deficits in specific behaviors
(e.g., looking at faces) and more holistic processing of social “presses,”
such as in orienting to someone’s voice. The forerunners of such
hypotheses, including the concept of autism as rooted in delayed or failed
acquisition of a “theory of mind” and even earlier deficits
in joint attention, have been shown to be part of developmental processes
that are crucial aspects of the overall pattern of ASD throughout childhood.
There
have also been attempts to generate separate measures of various aspects
of the phenotype. Standard diagnostic instruments yield scores in each
of three defining domains, but, on the whole, factor analyses and other
statistical techniques, all with relatively small samples, have tended
to yield one clear “autism spectrum” factor and often a strong
effect of the degree of cognitive impairment and/or verbal skills. Some
studies have found less robust factors that include various repetitive
behaviors, but these are less internally consistent and less consistent
across populations. Consideration of these issues, as well as the clinical
observation of the enormous diversity within ASD, has led to various proposals
of subtypes that differ from those proposed in DSM-IV and ICD-10
and that might be tied more closely to differences in neurobiological
factors. Such subtypes include distinctions related to IQ, language level,
presence of seizures, early developmental trajectory, and comorbidity
with other childhood psychiatric disorders. To date, none of these potential
subtypes has been consistently replicated beyond features that are already
part of the defining variable. However, some findings of different family
history data for nonverbal versus verbal individuals and for individuals
with normal versus delayed language milestones are quite intriguing. This
is an area of intense interest for genetics that will require large samples,
the development of sophisticated measures, and complex statistics to yield
replicable findings.
One
question that remains to be addressed is whether familial transmission
is different for social deficits, language abnormalities, and repetitive
behaviors. Twin studies have found that there is almost as much variation
in IQ and autistic symptoms within monozygotic twin pairs as within dizygotic
pairs. Families with a child with Asperger’s disorder show an increased
rate of autism as well as Asperger’s disorder and milder phenotypes.
Although a significant proportion of children with autism have mental
retardation, the rates of mental retardation without autism are not increased
in families of children with ASD, suggesting that the mental retardation
that occurs in some cases of autism is a part of the disorder, rather
than a separately transmitted phenomenon.
In
summary, substantial progress has been made in the ability to describe
accurately the social-communicative and behavioral deficits that comprise
ASD. Measurement tools that allow the categorization of autism and the
autistic spectrum in reliable and valid ways are available. There is a
strong empirical basis for the treatment of many of the symptoms of autism
as continuous dimensions. Categorical approaches do not fully represent
the range or the significance of these behaviors. Replicable thresholds
can be set, using specific methods, such as the ADI-R and the ADOS, for
the presence of ASD, based on observation or report of a constellation
of behaviors. Yet boundaries are not yet well defined among the ASD or
between individual behaviors. Identifying in relatives those behaviors
that are clearly associated with impairment and measurable personality
characteristics, which may or may not be pathological, in order to define
a unique broader common variant has been difficult. Much work remains
for those seeking to understand better the psychological and developmental
nature of these disorders and to provide better methods of quantification
of the disorder or, ideally, aspects of the disorder, for neurobiological
and genetic studies.
top of page
|
