Psychology
622, Section V
28 April, 2000
Abstract
Since Leo
Kanner’s discovery of Autism in 1938, many theories of potential causes,
possible treatments, diagnostic criteria, and diagnostic aides have emerged.
While there is no cure or thoroughly effective treatments for autism, there are
many theories regarding the cause of the disorder. Causal theories of Autism as
discussed in this paper are as follows: genetic basis, epileptiform activity,
brain abnormalities, and neurobiological basis. Autistic Disorder also has set
diagnostic criteria as established by the Diagnostic and Statistical Manual
of Mental Disorders IV (1994).
Further aiding in an accurate diagnosis of Autistic Disorder are several
diagnostic tools, which this paper will discuss as follows: Autism Diagnostic
Interview Revised, Autism Diagnostic Observation Schedule, Childhood Autism
Rating Scale, Autism Behavior Checklist, IBSE, and the use of home videos and
Latent Class Models. Studies comparing several of these aids will also be
discussed as well as future diagnostic tools for this disorder with many
unknowns.
On June 29, 1982 a mother gave birth to
her second son. Nothing unusual is noted about the birth or the pregnancy. However, after the first two years of his
life, the mother felt a nagging feeling that the baby boy was not properly
progressing in the areas of speech and motor skills. Rarely did the baby engage in verbalization, nor did the baby
make attempts to crawl, but rather scooted along on his back. The mother assumed that he was just a little
slower than his older brother had been and that he would soon catch up. However, the mother and several others
became concerned when the child scored extremely low on the screening tests for
preschool, especially in areas of language and socialization. Soon the child was shuffled from doctor to
doctor for diagnosis. The mother felt
overwhelmed by the avalanche of questions asked by the doctors. Finally, a diagnosis was made -- Autistic
Disorder.
Introduction
Leo
Kanner, a German born psychiatrist working at John Hopkins Hospital, first
discovered autism in 1938. Observing strange, idiosyncratic behavior in 11
children, Kanner wrote Autistic Disturbances of Affective Contact (1943)
about his observations, which he termed Early Infantile Autism (Lovett
1999). Since Kanner’s original 11
children, the rate of diagnosis for autism has risen dramatically around the
world, “affecting 33-160 per 10,000 using various diagnosis or 5 per 10,000
using the original diagnosis criteria set by Kanner” (Wing, 1997 p. 1765). This paper discusses the following:
1.
Diagnostic
Features of Autism
2.
Theories
of Autism
3.
Diagnostic
Tools
4.
Comparison
of Diagnostic Tools
5.
Future
Diagnostic Tools
This paper focuses on the
diagnosis of Autism in infancy and childhood.
Diagnostic
Features of Autism
Autism, according to Lorna Wing (1997), consists of a “common
triad of impairments: social interaction, communication, imagination, and
behavior” (p. 1761). This triad can
occur on its own or with any other condition and affects males three or four
more times than females (p.1765). The Harvard Mental Health Letter
(1997) states that the majority of autistics are retarded; although one group
termed “autistic savants,” which compromise 10 percent of the autistic
population, has extraordinary talents (p. 2).
While there is no cure, a study conducted by Martin, Scahill, Klin,
& Volkmar (1999) found that 68.8 percent of their participants (taken from
a sample of 26 states and one Canadian province) had used or were using
psychotropic drugs, with the most frequently prescribed medication being
antidepressants (32.1%) (p. 925).
Groups within the Autistic
Population
Wing (1997) categorizes autistics into four groups: aloof group, loner group, active group, and the passive group. The aloof group characterizes the typical autistic person and epitomizes Kanner’s original definition. Those in the aloof group have extremely limited speech and appear indifferent to others, preferring sensory stimuli, such as bright lights. Aloof autistics often engage in odd, repetitive behavior and will often sit for hours arranging things in a specific pattern or order and are very routine oriented (p. 1762). The next group, the active group, interacts with others on a naïve, odd, or inappropriate level and their play is limited to one or two toys. Although they are verbal, their conversations are usually disconnected or exhibit echolaia or may be concerned with one particular object or person (p. 1764). Whereas those in the active group accept human interaction, those in the loner group, according to Wing (1999), function at a high level but prefer to be alone and are concerned only with their own interests (p. 1764). “Often school days are stressful due to nonconformity to the demands of teachers and peers” (Wing, 1999, p. 1764). While those in the loner group have a tendency to avoid human contact, those in the passive group inertly accept approaches from others and are less rigid in their routines (p. 1763). Often “diagnosis in these may be missed until learning problems and interaction with peers begin to emerge in school” (p. 1764).
Detection
Although
sometimes detected in a school setting, more frequently it is parents who
detect autistic symptoms in their child, usually before the first two or three
years (Harvard Mental Health Letter, 1997, p. 1). In fact, a study by Adrein et al (1993)
confirms that in a majority of cases, autistic signs are apparent from
birth. The study consisted of 10 boys
and two girls diagnosed with infantile autism and eight boys and 4 girls who
were normal. All participants had filmed during the first two years of their
lives. Two diagnosis blind raters
analyzed each film using an IBSE scale.
“A Mann-Whitney U test was used for comparisons between behaviors of
normal and autistic children and a Wilcoxon matched pairs tests was used for
longitudinal analysis of behaviors in normal and autistic children” (Adrein et
al, 1993, p.218). The results show that
before the age of one, the autistic group differed significantly from the
normal group in the following areas: socialization, communication, lack of
appropriate facial expressions, hypotonia, and attention. After one year of age, eight new symptoms
emerged and are noted as follows: ignores people, prefers aloneness, no eye
contact, lack of appropriate expressive postures/gestures, too calm, adaptation
to environment, and no expression of emotions (Adrien et al, 1993, p. 219).
Thus, often detection of autism is possible in the very first months of the
child’s life.
The
genetic bases for autism stems from twin and family studies. According to the Harvard Mental Health
Letter (1997), the rate of autism for identical twins is 90% and 95% for
fraternal twins (p. 3). Also, the Harvard
Mental Health Letter (1997) states: “Among brothers and sisters of autistic
persons, the rate of both autism itself and milder related symptoms is 50 to 100
times higher than the average” (p.3).
Further evidence for a genetic basis comes from a study conducted in the
1980’s in Utah. The study found that
among 11 families with an autistic father, “more than half of the 44 children
were autistic” (Harvard Mental Health Letter, 1997, p. 3). These findings suggest a genetic component
for autism.
Given
that a major element of autism is a deficiency in communication, theories
regarding epileptiform activity as a causal factor for autism are being
explored. One reason for considering
epileptiform activity as theory lies in the fact that such activity has been
linked to causing ahasia in Landau-Kleffner syndrome and because of the fact
that one-third of autistic children experience seizures by adolescence (Lewine, 2000, p. 457 & Wing, 1997, p.
1764). Research conducted by Lewine
(2000) suggests “that there is a subset of children with ASD’s who demonstrate
clinically relevant epileptiform activity during slow-wave sleep, and that this
activity may be present even in the absence of a clinical seizure disorder”
(Lewine, 2000, p. 457). Lewine’s method
included using Magnetoencephalography to identify epileptiform activity during
stage III sleep in six children with Landau-Kleffner syndrome and in 50
children with autism. The results of the study revealed that 82% of those with
autism showed epileptiform activity in the same intra/perisylvian regions as
those with Landau-Kleffner syndrome (Lewine, 2000, p. 457). Lewine (2000) suggests that in the future,
strategies aimed at controlling the epileptiform activity may “lead to
significant improvement in language and autistic features” (p. 457).
A third theory regarding origin of
autistic features is a neurobioligical one.
In fact, studies using Positron Emission Tomography suggest
abnormalities in the brain at the subcellular level. Positron Emission Tomography (PET) is a “techinique that can be
used to quantitatively probe various neurochemical systems” (Trifiletti, 1999,
p.68). Trifiletti (1999) reports that a
study conducted by Chugani et al shows evidence for “significant abnormalities
in development of serotonin synthesis in autistic patients” (p. 68). The study, using PET, investigated serotonin
synthesis capacity in 30 autistic patients, eight normal siblings, and 16
children with epilepsy. Chugani et al
discovered a decline in serotonin synthesis capacity among non-autistic
children and a smaller increase in serotonin synthesis capacity within the
autistic population (Trifileti, 1999, p. 68).
“Differences in serotonin synthetic capacity between autistic and
nonautistic are most marked during infancy” (Trifileti, 1999, p. 68). This discovery agrees with other findings
that autistic symptoms form in infancy.
Abnormal
Brain Structure
The last
theory regarding the cause of autism discussed in this paper is that of
abnormal brain structure. The Harvard
Mental Health Letter (1997) reports that:
many
[autistic patients] have abnormally small and densely packed neuron in the
amygadala and in the hippocampus …[and] abnormally low blood circulation
in
parts of
the cerebral cortex during intellectual operations and a reduction in the
number
of cells relaying inhibitory messages from the body movement centers in
the
cerebellum to the cerebral cortex (p. 3).
These findings suggest an explanation for the
flattening of emotional responses exhibited in many with autistic disorder due
to the fact that there are abnormalities in the amygdala and cerebral cortex,
which control emotional responses and sensory information respectively.
Further
support for the theory of brain abnormalities is found in a study conducted by
Muller et al (1999) in which they used PET to map language and auditory
perception areas within the brain.
Muller et al, using 5 autistic and 5 matched non-autistic adults,
conducted a series of PET scans under the following conditions: “(a) rest; (b) listening to a psendorandom
sequence of tones; (c) listening to 10 structurally simple sentences; (d)
sentence repetition; and (e) sentence generation” (Muller et al, 1999, p.
22). The results of listening to tones
showed “peak activation in the bilateral primary and secondary auditory cortex”
(Muller et al, 1999, p. 24) for the control group and “weaker, but significant
activation in the left anterior cingulategyrus” (Muller et al, 1999, p. 24) for
the autistic group. Moreover,
differences between the autistic group and the control group were found upon
analyzing brain activation patterns for listening to sentences. The control group exhibited “peak activation
in the lateral temporal cortex” where as the autistic group showed “2 peaks
found in the right middle frontal and left superior temporal syri, …indicating a
reduced left dominance” (Muller et al, 1999, p. 25). The two groups differed again in activation patterns for sentence
repetition. In contrast to the control
group, Muller et al (1999) found that “the autistic group showed no activation
in the left caudate nucleus, right thalamus, or midbrain” (p. 26). Lastly, upon comparing activation patterns
for sentence generation, the control group showed a "peak in the inferior
and middle frontal gyri, frontal operculum, and inferior temporal region of the
left hemisphere” (Muller et al, 1999, p. 26).
In contrast, the autistic group exhibited only one significant
activation in the left middle frontal gyrus. (Muller
et al, 1999, p.26). Overall, Muller et al (1999)
reported that thalamic activations were consistently right dominated in
autistic group and left dominant in the control group (p.26). These findings emphasize that abnormal brain
functioning, especially in the areas of hemispheric dominance, may be the
causal factor for the signs and symptoms of Autistic Disorder.
Diagnostic
Tools
One of the
most frequently used diagnostic tools for the identification of Autistic
Disorder is the Diagnostic and Statistical Manual of Mental Disorders IV. In fact, most insurance companies require
the inclusion of DSM identification numbers on all reports. Therefore, the DSM has become the “gold
standard” in America for the diagnosis of Autistic Disorder. While Kanner first
identified autism in 1943, it was not until the printing of the DSM-III (1980),
that the American Psychological Association classified the disorder, thus
making the diagnosis of autism readily available (Volkmar et al, 1994, p. 1361).
Autistic Disorder was “included in a new class, the pervasive
developmental disorders” (Volkmar et al, 1994, p. 1361). With the revision of the DSM-III-R, the
criteria for autistic disorder were expanded to account for developmental
changes and to lower the age of onset (Folkmar et al, 1994, p. 1362). However, the DSM-III-R did not incorporate
differential diagnoses, such as those seen in the ICD-10 (1990), which
included: Rhett syndrome, Heller’s Syndrome, and Asperger syndrome (Volkmar et
al, 1993, p. 1362). Noticing a high
rate of false positive cases, Volkmar et al conducted a study “to identify
issues that needed resolution for DSM-IV” in which “ a series of literature
reviews and data analysis were undertaken” (Volkmar et al, 1993, p. 1362). The results of the study show that the
DSM-III-R had a sensitivity of (.93) and a specificity of (.79), whereas the ICD-10
results are (.87) and (.98), the highest specificity and consequently, the
highest rate of agreement among the clinicians participating in the study
(Volkmar et al, 1993, p. 1362). As a
result of these findings, the DSM-IV compromises criterion of both the
DSM-III-R and the ICD-10, making for an accurate diagnostic tool for autistic
disorder.
Terms
commonly used in diagnosis of autism according to a survey conducted by Bennett
(1993) include (in rank order): severe communication difficulties, autistic
features, and autistic (p. 345). The
survey consisted of a list of six common descriptive terms—autistic, autistic
features, autistic syndrome, pervasive developmental disorder, severe
communication difficulties, and asperger, drawn from interviews with a group of
educational psychologists. Speech
therapists, specialist teachers, pediatricians, and educational psychologists,
who were all involved in the early identification of children showing autistic
features, participated in the survey (Bennett, 1993, pp. 343-344). Although the DSM-IV classifies autistic
features as Autistic Disorder under the category of pervasive developmental
disorders, there is evidence that using such terminology may actually inhibit
the diagnostic process. In fact, in the
survey conducted by Bennett (1996), 78 % of practitioners reported
complications in diagnosis/description using the terms “due to negative
reactions of parents or disagreements between professionals” (Bennett, 1993, p.
345). From his survey, Bennett (1993) suggests, “diagnostic description should
be made only after a period of multi-disciplinary assessment, observation, and
interviews” (p. 346).
Other Diagnostic Tools
As of
yet autistic disorder has no biological cause.
Therefore, diagnosis must be based on behavioral criteria determined by
the gathering of a careful developmental history. This paper will discuss the following diagnostic tools that can
be used in conjunction with the DSM-IV: Autism Diagnostic Interview Revised,
Autism Diagnostic Observation Schedule, Childhood Autism Rating Scale, Autism
Behavior Checklist, IBSE, and the use of home videos and Latent Class
Models.
The
Autism Diagnostic Interview Revised (ADI-R) is a semi-structured, investigator
based interview administered to caregivers of persons for whom autism is
possible. The interview consists of 111 items that explore the following areas:
communication ability, use of vocabulary and grammar, ability to engage in
social chat, “prosodic” conventions, and the ability to engage in social
exchanges. The interview also measures
emotional and non- emotional gestures and records abnormalities in
communication such as echolalia and bizarre communication efforts (Tanguay,
1998, p. 272). The ADI-R also comes
with a diagnostic algorithm in which ADI-R items are matched to three DSM-IV
symptom domains for autism to determine whether the person would receive a
diagnosis according to the DSM-IV criteria (Tanguay, 1998, p. 272). The questions are addressed using specific
probes for each item, such as emotional empathy, play behaviors, and
friendships. For each response given,
the examiner asks additional questions to “pin down” the symptoms (Tanguay,
1988, p. 272). A numerical score is
assigned to each item ranging from 0 to 3, with three indicating very
abnormal. Also, each time the guardian
describes a behavior that indicates abnormality; the interviewer asks for
specific examples and records the examples to be included in a written report
(Tanguay, 1988, p. 272). Overall, the
ADI-R adequately detects those behaviors conducive to autistic disorder.
ADOS
The next
diagnostic tool, the Autism Diagnostic Observation Schedule (ADOS), evaluates
the patient’s socialization skills. The
ADOS encompasses nine social “presses” in which the interview sets up a social
situation, such as playground scenario, and the child or adolescent’s responses
are taped for scoring. In the PLADOS,
used for nonverbal children, a child is filmed and is invited to play with
various toys or engage in social activities, such as a birthday party (Tanguay,
1998, p. 272). This instrument best
gauges a child’s socialization skills.
A
further diagnostic tool, the Childhood Autism Rating Scale (CARS), consists of
15 items and employs a 7-point Likert scale ranging from 1 to 4. The higher the score is, the greater the
deviance from the norm. While
conducting the test, the interviewer refers to unique narrative descriptions
for selecting scale values of each item.
Following the completion of the 15 questions, the item scores are summed
to obtain the total CARS score.
Estimates of reliability for the CARS are high. In fact, in an independent investigation
conducted by Garfin, McCallon, and Cox (1988) found that after a 12-month interval, test-retest
reliability for 91 cases shows a rating of .79. Thus, the CARS is a reliable tool for diagnosing autism in
children (Eaves, 1993, p. 483).
Another
reliable instrument for diagnosing autism is the Autism Behavior Checklist
(ABC). With the ABC, caretakers
respond to 57 items as being present or absent. The contribution of any given ABC item depends upon its
pre-determined weight. The weight of
each item is based on the item’s contribution to the prediction of autism. Following the completion of the checklist,
items are collapsed into five scales: sensory, relating, body and object use,
language, and social and self-help. The
sum of the scores in each of the scales constitutes the total ABC score. The reliability of the checklist is found to
be .82 (Eaves, 1993, p. 483). This
checklist adequately detects signs of autism.
The IBSE
assesses the autistic behaviors exhibited by young children. This diagnostic
tool consists of a clinical and quantitative assessment of behavior, which is
used to distinguish those behaviors exhibiting autistic attributes. The scale consists of 33 items that are
rated on a five-point scale ranging from 0 for never to 4 for continuously. The
33 items are classified under six main categories: socialization,
communication, adaptation to environment, tonus motility, emotional and
instinctual reactions, and attention and perception. A global score is obtained by adding all the scores of the items
on the rating scale. This instrument
is useful in assessing autistic behaviors in young children (Adrien et al,
1993, p. 618).
As noted
earlier, home movies provide an excellent diagnostic tool for the screening of
early symptoms of autism. The practice
of reviewing home movies permits a longitudinal observation with an accurate
depiction of the child’s behavior before or at the moment of the onset, and
subsequently, aides in the informant’s account of the child. Also, by pin
pointing the first signs and symptoms of autism in the child, a more
individualized treatment program can be undertaken (Adrien et al, 1993, p.
621).
LCM
Lastly,
Latent Class Models (LCM) can be useful in the evaluation of diagnostic
criteria lacking a gold standard, such as with the diagnosis of autism. First used by Young to evaluate diagnostic
criteria for schizophrenia, LCMs provide an assessment of diagnostic accuracy
when no base rate can be obtained (Szatmari, Volkmar, & Walter, 1995, p.
216). With the LCM approach, “expressions
are obtained for the probability of an individual’s having any particular
combination of observed results on the diagnostic criteria” (Szatmari, Volkmar,
& Walter, 1995, p. 216). To calculate the estimate of
false-positive—false-negative errors, the unknown
variable is assigned a value of either present or absent. The false- positive rate = 1- specificity
and the false-negative
rate = 1- sensitivity. This formula independently
predicts how close the judgements of the behaviors/criteria for any patient is
true (Szatmari, Volkmar, & Walter, 1995, p. 216). This model aids in establishing which diagnostic tool to
use.
As noted
above, there are many diagnostic instruments for evaluating suspected autistic
children and adolescents. Consequently,
comparisons have been made between various diagnostic tools. This paper will discuss the comparison of
the ABC and CARS and the comparison of the combined ADI, ADOS, and PADOS with
the DSM-IV criteria.
To
establish the relationship between the Autism Behavior Checklist and the
Childhood Autism Rating Scale, Eaves & Milner (1993) sampled 77 subjects
(48 with a firm diagnosis or autism) and compared sensitivity ratings and
correlations obtained by computing the Pearson r. Participants were administered both scales. The results are as follows: the ABC
sensitivity rating equaled 82% and the CARS sensitivity rating equaled 98%. A moderate correlation (.67) between the two
scales was found. This relationship indicates that the CARS and ABS measure
similar, but not identical constructs.
However, this experiment failed to examine the accuracy of the two tests
in identifying those who are autistic within a group because no estimate of
specificity of non-autistic subjects correctly identified could be
generated. However, based on
sensitivity scores, the CARS appears to be a better diagnostic tool than the
ABC (Eaves & Milner, 1993, p.281-286).
Tanguay, Robertson, & Derrick (1998) also compared two diagnostic instruments. Tanguay, Robertson, & Derrick (1998) compared the ADI, ADOS, and PADOS to the DSM-IV to “assess whether ‘social communication’ could be used to assess severity of symptoms in autism” (p. 211). The 90 participants, referred by the Autism Society of Kentuckiana and by clinicians in Kentucky, were tested using the three instruments, and were evaluated by independent clinicians using the DSM-IV. The results show that based on “current’ symptoms, of the 63 autistic subjects, the DSM-IV diagnosed 32 correctly while the ADI-R Autism Algorithm diagnosed 42 autistic subjects correctly. Further analysis using the Pearson r revealed that the first and second domains within the DSM-IV correlates well with the social communication domains while the third domain of the DSM-IV did not correlate well. Thus, Tanguay, Robertson, & Derreck, suggest that clinicians need to “judiciously apply DSM-IV criteria to diagnose sever forms of autism” (p. 275). However, the DSM-IV still remains the standard for the diagnosis of Autistic Disorder.
In the
last 50 years since Kanner’s first discovery, many theories of potential
causes, possible treatments, diagnostic criteria, and diagnostic instruments
have emerged. While there are many
unknowns in the areas of cause and treatment, there is a set criteria as established
by the DSM-IV and there are several well tested and effective diagnostic
techniques for detecting autism, such as the following: Autism Diagnostic
Interview Revised, Autism Diagnostic Observation Schedule, Childhood Autism
Rating Scale, Autism Behavior Checklist, IBSE, and the use of home videos and
Latent Class Models. Much research has
been done in the development and implementation of these diagnostic aides, as
well as research comparing various diagnostic tools. Also, examining serotonin synthesis production and the use of PET
for finding both a cause and a diagnosis appears promising. However, after 50 years of research,
Autistic Disorder remains enigmatic.