Content uploaded by Patricia Howlin
Author content
All content in this area was uploaded by Patricia Howlin on Mar 03, 2017
Content may be subject to copyright.
several studies now suggest that if strict DSM/ICD cri-
teria are applied, a diagnosis of Asperger syndrome be-
comes unlikely, or even impossible (Eisenmajer et al.,
1996; Ghaziuddin, Tsai & Ghaziuddin, 1992; Manjiviona
& Prior, 1995; Mayes & Calhoun, 2001; Miller &
Ozonoff, 1997; Szatmari, Archer, Fisman, Streiner &
Wilson, 1995).
Wing, having been responsible for first bringing
Asperger’s original writings to the attention of clini-
cians in 1981, now expresses her concern that she may
have opened a “Pandora’s box” (Wing, 2000). Her orig-
inal intention, to emphasize the fact that there was no
evidence for a distinction between Asperger syndrome
and autism, has frequently been overlooked. Instead,
her account led to widespread research into the possi-
ble differences between the two conditions. Inadequate
group matchings, small samples, and, above all, lack of
INTRODUCTION
The issue of whether autism and Asperger syn-
drome are different conditions (albeit part of the same
spectrum of disorders) has been a source of continuing
debate over recent years (cf. Klin, Volkmar, & Sparrow,
2000; Schopler, Mesibov, & Kunce, 1998). There is
also considerable disagreement about the validity of the
diagnostic criteria used in DSM-IV (APA, 1994) and
ICD-10 (WHO, 1992) to distinguish between the two
conditions, (Kim, Szatmari, Bryson, Streiner, & Wilson,
2000; Kugler, 1998; Leekam, Libby, Wing, Gould, &
Gillberg, 2000; Manjiviona & Prior, 1999). Indeed,
Outcome in High-Functioning Adults with Autism With
and Without Early Language Delays: Implications for the
Differentiation Between Autism and Asperger Syndrome
Patricia Howlin
1
The question of whether Asperger syndrome and high-functioning autism should be considered
as the same or different conditions has been a source of debate and controversy over recent
years. In the present study, 34 adults with autism who had shown early delays in language
were compared with 42 individuals who were reported to have had no such delays, either in
their use of words or phrases. All participants were at least 18 years of age, had a nonverbal
IQ of 70 or above and met ADI-R criteria for age of onset, communication and social im-
pairments, and stereotyped behaviors. Those in the language delay group were diagnosed as
having high-functioning autism. The remainder were designated as having Asperger syndrome.
The groups were matched for age, nonverbal IQ and gender. No significant differences were
found between the groups either in their total ADI-R algorithm scores, or in their algorithm
scores on individual domains. Social outcome ratings and ADI-R scores based on current func-
tioning also failed to differentiate between the groups. Scores on tests of language compre-
hension and expression were also similar, but in both groups language abilities were well below
chronological age level. The implications of these results with respect to the differences be-
tween Asperger syndrome and high-functioning autism are discussed. The poor performance
on language tests also challenges the assumption that early language development in Asperger
syndrome is essentially normal.
KEY WORDS: High-functioning autism; Asperger syndrome; long-term outcome.
Journal of Autism and Developmental Disorders, Vol. 33, No. 1, February 2003 (© 2003)
30162-3257/03/0200-0003/0 © 2003 Plenum Publishing Corporation
1
Department of Psychology, St. George’s Hospital Medical School,
Cranmer Terrace, London SW17 ORE, U.K.; e-mail: phowlin@
sghms.ac.uk
4Howlin
agreement on diagnostic criteria have meant that few
of these studies have produced any conclusive results
(c.f. Klin et al., 2000). Despite this, the differential di-
agnosis can have implications for services, with indi-
viduals diagnosed as having Asperger syndrome or
Asperger disorder sometimes being deprived of the
level of support that is offered to those diagnosed as
having autism (Klin & Volkmar, 2000; Department of
Health, 2001).
Current Research Evidence
A systematic search of the literature over the past
15 years, using the major databases (MEDLINE, Psych-
INFO, etc.) was conducted together with an additional
trawl of journals with a focus on developmental disor-
ders (Journal of the American Academy of Child and
Adolescent Psychiatry, Journal of Child Psychology
and Psychiatry, Journal of Autism and Developmental
Disorders, Autism, etc.), and recent collections of chap-
ters in edited volumes (Klin et al., 2000; Schopler et al.,
1998). Twenty-six separate studies were identified in
which a direct comparison between individuals with
high-functioning autism (HFA) (IQ 70⫹) and Asperger
syndrome had been conducted and in which diagnostic
criteria and measures of IQ were adequately specified.
Tables I and II summarize these studies, indicating the
general areas investigated, the diagnostic systems used,
the number and ages of participants, and the general
conclusions of each.
Table I. Summary of Findings in Studies Comparing Individuals with Asperger Syndrome
and High-Functioning Autism Matched for Full-Scale IQ
Principal area of study, Author Age n
(and criteria for AS) Mean (range) HFA:AS Conclusions
General clinical characteristics
1. Szatmari et al., 1990 (Wing’s criteria)
2. Mayes & Calhoun, 2000 (modified DSM-IV)
3. Ozonoff et al., 2000 (DSM-IV)
Obstetric/early history and motor abnormalities
4. Szatmari et al., 1989 (Wing’s criteria)
5. Gillberg & Gillberg, 1989 (Gillberg criteria)
Neuropsychological and language profiles
6. Ozonoff et al., 1991 (modified ICD-10)
7. Klin et al., 1995 (modified ICD-10)
8. Ozonoff et al., 2000 (DSM-IV)
9. Rinehart et al., 2000 (DSM-IV)
10. Szatmari et al., 1990 (DSM-III)
11. Iwanaga et al., 2000 (DSM-IV)
12. Klin, 2000 (DSM-IV)
Behavioural and psychiatric disturbance
13. Szatmari, 1989 (Wing’s criteria)
HFA 23, AS14 yr
6.0 yr
13.5 yr
HFA 23 AS 14
10 yr
11–12 yr
15–16 yr
13.5 yr
10 yr
HFA 23, AS14
5–6 yr
19–20 yr
HFA 25 AS 14
17:26
23:24
23:12
25:28
23:23
13:10
19:21
23:12
12:12
17:26
15:10
20:20
25:28
HFA ⬎social impairments, language prob-
lems, stereotypies, and preoccupations
No difference on any of 71 variables (IQ,
language, symptomatology, motor
coordination, emotion or behavior)
HFA ⬎ADI scores at 4; few differences in
current scores. No differences in social
functioning. AS ⬎special interests;
HFA ⬎insistence on sameness
HFA ⬎early abnormalities in social,
language, and behavior
AS ⬎clumsy
AS ⬎VIQ, and ⬎VIQ-PIQ difference.
AS ⬎verbal memory, executive function
& TOM scores. HFA ⬎CARS scores
HFA ⬎PIQ and better motor skills; AS ⬎
VIQ and higher verbal ability
Few differences but AS ⬎VIQ-PIQ differ-
ences; HFA ⬎expressive problems. No
differences on executive function/TOM
tasks
HFA ⬎problems in shifting attention
Few differences in IQ profiles, motor, or
other test scores
No significant differences on gross or fine
motor tasks, and most nonverbal tests.
AS ⬎HFA on some verbal tasks
No differences on social attribution task
HFA ⬎bizarre preoccupations; AS? ⬎
psychiatric problems
The majority of participants were children, with only
three research groups including any adults (Gilchrist,
Green, Cox, Rutter & Le Couteur, 2001; Klin, 2000;
Szatmari et al., 1989, 1990, 1995). In many cases strict
DSM-IV criteria for Asperger disorder were not used,
principally because, as noted above, if the hierarchical
guidelines proposed by ICD-10 or DSM-IV/DSM-
IV(TR) (APA, 1994, 2000) are followed, only a small
proportion of cases then meet the necessary criteria. In-
stead, this label tended to be used for individuals who,
Implications for the Differentiation Between Autism and Asperger Syndrome 5
although often meeting criteria for autism, had not
shown early language delays. Moreover, many of the
studies reviewed failed to match participants with As-
perger syndrome and those with HFA on the basis of IQ.
Thus the reported differences between the groups could
be due to cognitive disparities rather than a true diag-
nostic differentiation. Table I summarizes the findings
of 10 separate investigations in which participants were
matched for full-scale IQ (three studies cover several
different areas and are listed more than once). The find-
Table II. Summary of Findings in Studies Comparing Individuals with Asperger Syndrome
and High-Functioning Autism (AS Groups Higher IQ)
Principal area of study, Author Age n
(and criteria for AS) Mean (range) HFA:AS Conclusions
General clinical characteristics
1. Szatmari et al., 1995 (modified ICD-10 criteria)
2. Eisenmajer et al., 1996 (clinical diagnosis)
3. Kurita, 1997 (ICD-10)
4. Gilchrist et al., 2001 (ICD-10)
Obstetric/early history and motor skills
5. Ghaziuddin et al., 1994 (ICD-10)
6. Ghaziuddin et al., 1995 (ICD-10)
7. Ghaziuddin & Butler, 1998 (DSM-IV/ICD-10)
8. Manjiviona & Prior, 1999 (modified ICD-10)
Neuropsychological and language profiles
9. Ghaziuddin & Gerstein, 1996 (ICD-10)
10. Ehlers et al., 1997 (Gillberg criteria)
11. Pomeroy, 1998 (non-language impaired PDD)
12. Manjiviona & Prior, 1999
(Modified/strict DSM-IV)
13. Miller & Ozonoff, 2000 (DSM-IV)
Behavioural and psychiatric problems
14. Ghaziuddin et al., 1995 (ICD-10)
15. Tonge et al., 1999 (DSM-IV)
16. Kim, 2000 (modified DSM-IV))
5 yr
10 yr
5–6 yr
HFA 21 yr;
AS 14 yr
12–13 yr
13–14 yr
10–11 yr
11 yr
15–16 yr
10 yr
7–8 yr
10–11 yr
10 yr
12 yr
AS 9.9; HFA
7.4 yr
12 yr
47:21
48:69
16:26
13:20
9:11
9:11
12:12
9:12
13:17
40:40
13:15
21:35
26:14
8:12
75:52
40:19
HFA ⬎social and adaptive problems,
rituals and early language delays. No
differences in current nonverbal/
communication/motor skills
Few differences on any variables;
delayed language only significant
differences in HFA group
Few significant differences but HFA ⬎
on some CARS items
HFA ⬎ADI problems age 4–5 yrs. No
difference in current functioning on
ADI or ADOS
No significant differences in motor
skills
No significant differences in neonatal
optimality scores
No significant differences in motor
skills when IQ controlled for
AS ⬎PIQ but no differences in motor
skills
AS ⬎pedantic speech
Diffs in cognitive profiles; AS higher
scores on most subtests
AS ⬎VIQ; also ⬎language compre-
hension and expression
AS ⬎VIQ no difference on neuropsy-
chological profiles
No significant differences in motor
skills, executive function or TOM
when IQ co-varied
AS ⬎disorganized thought; few other
significant differences
AS ⬎psychopathology
No differences in
depression/anxiety/mood
6Howlin
ings are variable. Of five reports on early history and
general clinical characteristics, three (studies 1, 3, and
4 in Table I) suggest higher rates of problems in HFA
groups. However, these differences were not necessar-
ily maintained as children grew older (study 3). There
was greater evidence of special interests in the Asperger
group, whereas insistence on sameness was more com-
mon in the autism group (study 3). Only one study
(no. 5) found evidence of increased clumsiness in in-
dividuals with Asperger syndrome. Of seven studies
examining neuropsychological or linguistic function-
ing, four (nos. 6, 7, 8, and 11) reported higher verbal
skills in the Asperger groups. There were few differ-
ences in scores on tasks of social understanding, The-
ory of Mind or executive function (nos. 8 & 12) and
those that were found may have been related to higher
verbal skills in the Asperger group (no. 6). Rates of
psychiatric disturbance did not differ markedly al-
though individuals with HFA tended to show more
bizarre pre-occupations (no. 13).
Table 2 summarises the findings of sixteen sepa-
rate studies in which the Asperger syndrome groups
were of higher overall IQ than the Autism groups. Of
eight with a focus on early history, general clinical
severity or motor problems, five reported few differ-
ences. Ghaziuddin and Butler (1998; no. 7) identified
more motor difficulties in their HFA group, but the dif-
ference disappeared when IQ was controlled for and
both groups were relatively impaired compared to pop-
ulation norms. The results of Szatmari et al. (1995,
no. 1) suggested that early group differences in lan-
guage abilities diminished over time. Similarly,
Gilchrist et al. (2001, no. 4) found that initial differ-
ences in symptom severity tended to decline with age
and although there were differences on ADI scores in
early childhood (with the HFA group showing more
difficulties) there were no significant differences in
adolescence and early adulthood. Of five reports on
neuropsychological and linguistic functioning, the
Asperger groups generally showed superior verbal
skills. However, such differences disappeared when
IQ was controlled for (no. 13). In the area of
behavioural/psychiatric disturbance two studies (14
and 15) found evidence of increased pathology in the
Asperger groups but this was not found in study 16.
Overall, there appears to be no consistent evidence
that there are any major differences in rates of social,
emotional and psychiatric problems, current sympto-
matology, motor clumsiness or neuropsychological pro-
files between the two groups. Moreover, there are
suggestions that group differences found in early child-
hood may decrease with age (Gilchrist et al., 2001;
Ozonoff, South, & Miller, 2000; Szatmari et al., 1995).
Several studies do, however, suggest that even if chil-
dren are matched on full-scale IQ, the Asperger groups
have better developed verbal skills in certain areas.
The fact that individuals with autism who, by defini-
tion, are significantly delayed in their language de-
velopment continue to show poorer linguistic skills
than those with Asperger syndrome (who, by definition,
are not delayed) is perhaps not surprising (Szatmari,
2000). However, because few studies have examined
differences in these groups beyond mid adolescence,
it is possible that the apparent linguistic superiority of
the Asperger groups might reduce with age. The pre-
sent paper, therefore, focuses on differences in older
individuals, all aged 18⫹, and all with a nonverbal IQ
of at least 70.
The three main questions posed were:
1. Do high-functioning adults with autism (IQ
70⫹) who showed early language delays dif-
fer in their early symptomatology (as assessed
by parental report) from those with no early
language delays?
2. Do ADI-R scores (current and retrospective) for
social, communication, and ritualistic/stereo-
typed behaviors indicate differences between the
groups?
3. Are there other differences with respect to so-
cial and linguistic functioning between the two
groups in adulthood?
METHOD
Participants
The participants in the study were selected from
patients attending a specialist diagnostic and assess-
ment clinic for adults with developmental disorders.
All had been referred by their local services (family
doctors, social or psychiatric services) because of in-
dications that they had an autistic spectrum disorder.
Some had been diagnosed with an autistic spectrum
disorder when younger, but many had not and they or
their parents were generally seeking a formal diagno-
sis or clarification of an earlier diagnosis. All indi-
viduals who met the following criteria were included
in the study:
●Diagnosis of autism or Asperger confirmed on
the basis of the ADI-R (conducted by the au-
thor), plus previous clinical records, contempo-
raneous reports, and assessment of current
functioning
●18 years or older
●Nonverbal IQ within the normal range (i.e., 70⫹)
Measures
Cognitive and Linguistic Tests
Nonverbal IQ scores were obtained either from the
Raven’s Matrices (Raven, 1956), the Wechsler Adult
Intelligence Scale (WAIS-R or WAIS-III, Wechsler,
1981, 1997), or the Wechsler Abbreviated Scale of In-
telligence (WASI; Wechsler, 1999). Comprehension for
single words was assessed using the British Picture Vo-
cabulary Scale (Dunn, Dunn, Wetton, & Burley, 1997;
Dunn, Dunn, Whetton, & Pintillie, 1982) and vocabu-
lary by the Expressive One Word Picture Vocabulary
Test (Gardner, 1982). Although these are among the
very few tests that are available for assessing language
beyond middle childhood, most individuals were, nev-
ertheless, above the chronological age ceiling and hence
a standard score could not be calculated. Age equiva-
lent scores were therefore used as an alternative, and
for those scoring above the maximum a randomized age
equivalent was assigned (see Mawhood, Howlin, &
Rutter, 2000). In addition, because of the problems in-
herent in deriving age-equivalent scores in this way,
ratings of language level were used (0 ⫽above test
ceiling; 1 ⫽15 yr to ceiling; 2 ⫽12–14.9 yr; 3 ⫽
10–11.9 yr; 4 ⫽8–9.9 yr; 5 ⫽⬍8yr).
ADI-R Scores and Group Classification
A detailed developmental history was obtained for
all participants by means of the ADI-R (Lord, Rutter,
& Le Couteur, 1994). In the majority of cases the in-
formants were parents (usually the mother), although
in three cases siblings provided the information and one
informant was a stepmother (she had also been the
person’s childminder when he was tiny). ADI-R algo-
rithm scores for impairments in communication, reci-
procal social interaction, and repetitive and stereotyped
behaviors/interests, together with the total score for all
three domains were calculated for each individual. Rat-
ings of fine and gross motor coordination were also
obtained from the ADI-R, because problems in these
areas are often noted as being more frequent in people
with Asperger syndrome (Volkmar & Klin, 2000).
Out of a total of 76 individuals meeting inclusion
criteria, 34 (25 men, 9 women) were reported by the
informant to have shown definite delays in language as
defined by the ADI-R (i.e., no single word speech by
24 months age and/or no phrase speech by 3 years). All
individuals in this group were diagnosed as having autism
Implications for the Differentiation Between Autism and Asperger Syndrome 7
in that they scored at or above the cut off for abnormal-
ities on each of the domains (communication, social un-
derstanding, stereotyped behaviors, and age) assessed by
the ADI-R. Forty-two individuals (35 men, 7 women)
who also met full ADI-R criteria were reported to have
had no delays, either in their use of words or phrases.
Although individuals in this group would not meet cur-
rent DSM-IV criteria for Asperger disorder, for the pur-
pose of the present comparison, and in keeping with
other recent studies (Kim et al., 2000; Manjiviona &
Prior, 1999; Szatmari et al., 1995) they were designated
as having Asperger syndrome. There was no signifi-
cant difference in the age of individuals in the two
groups (average age in the autism group, 27.6 yr, SD
10.0 yr; in the Asperger group, 26.1 yr, SD 5.9 yr; t ⫽
80, p⫽.43). The ratio of males to females in the autism
group was 1.0:2.8; in the Asperger group it was higher
at 1.0:5.0 but the difference was not significant (X
2
⫽
1.09; p⫽.29).
Findings
Note that because of the number of comparisons
conducted, only p levels ⬍.01 were considered signif-
icant. Table III summarizes the characteristics of both
groups with regard to age of first symptoms, early lan-
guage development, and ADI-R algorithm scores.
The age at which parents first noted symptoms of
abnormal development was somewhat earlier in the
autism group but the level of significance was marginal
(see Table III). However, there was a significant dif-
ference in parental reports of the types of symptoms
that first gave rise to concern. In the autism group, the
most commonly reported early signs were language
problems (32%); problems or delays in motor devel-
opment (23.5%); social abnormalities (23.5%), and
general behavior difficulties (15%). Only one parent in
this group cited repetitive and stereotyped behavior pat-
terns as being the initial cause for concern. In the As-
perger group, parents’ initial concerns focused on
general behavior problems (29%); repetitive and stereo-
typed behaviours/interests (21%); motor difficulties/
delays (19%); social abnormalities (19%), and language
(12%). (X
2
⫽11.82; p⫽.037). Information on first
symptoms was missing for one individual in the Autism
group who had been adopted.
Although all informants were able to recollect
whether or not the individual had shown language de-
lays (i.e., words by 2 years and phrases by 3 years) not
all were able to give the actual age of speech acquisi-
tion. For those on whom this information was avail-
able, the average age of first words in the autism group
was 38 months compared to 15 months in the Asperger
group. The average ages for first phrases were 52 and
26 months, respectively. (see Table III). All partici-
pants had some speech when assessed as adults, but in
the autism group the latest age at which words were ac-
quired was 72 months; for phrases, the latest age was
84 months. In the Asperger group the latest ages of
acquisition were 24 months for words and 30 months
for phrases.
ADI-R algorithm scores confirmed that, despite
the late age at which the diagnostic assessment was
conducted, all individuals met criteria for communica-
tion and social impairments and for ritualistic/stereo-
typed behaviors. There were no significant differences
between the groups, either in their total ADI-R algo-
rithm score or in their algorithm scores on the individ-
ual domains (see Table III). ADI (“ever”) ratings for
problems in gross and fine motor skills did not differ
between groups. (Median: autism group ⫽0; Asperger
group ⫽1.0; Mann Whitney Z 1.43; p⫽.15).
Differences in present functioning were also ex-
amined, using standardized tests, “Current”
1
ratings
from the ADI algorithm and information pertaining to
jobs, relationships, and independent living. As is ap-
parent from Table IV, there were no group differences
in nonverbal IQ, and scores on the ADI-R domains
were almost identical.
From Table IV it can also be seen that there were
no significant group differences in the numbers of peo-
8Howlin
ple living independently, in ratings of current friend-
ships, or in employment levels, although the propor-
tion of individuals with Asperger syndrome in some
form of employment (sheltered or independent) was
higher than in the autism group (38% vs. 52%). The
one social variable for which a significant difference
was found was educational qualifications. Significantly
more individuals in the Asperger group had obtained
at least A level qualifications (i.e., the level required for
university entry) than in the autism group (24% vs.
52%). Median scores for current ratings of clumsiness
(fine and gross) did not differ.
Ten individuals in the autism group were described
by their referring agency as having additional psychi-
atric problems (8 depression, one alcohol dependency,
and one agoraphobia). No further mental state examina-
tion was conducted during the present assessment and
hence psychiatric diagnoses could not be confirmed.
However, for one of these individuals with a depressive
disorder the “Current” section of the ADI-R could not
be completed because his depression was so severe as to
have significantly altered his typical pattern of func-
tioning. In the Asperger group, 7 individuals were de-
scribed as suffering from depression (one with co-morbid
obsessional compulsive disorder [OCD] and one with co-
morbid anorexia); one man had a diagnosis of OCD, and
one had been treated for a psychotic illness in the past.
On tests of current linguistic functioning, age-
equivalent scores for language comprehension and ex-
pression were slightly higher in the Asperger group but
the difference was not significant. Only on ratings for
language expression (based on overall age band reached)
was there a marginally significant difference, in favor
of the Asperger group (see Table IV).
Table III. Comparison Between Asperger and High-Functioning Autism Groups on Early Childhood Measures
Autism (n ⫽34) Asperger (n ⫽42)
Group Mean SD Mean SD t pvalue
Age first signs 15.01 months 13.21 21.27 months 12.9 2.03 .046
Age first words
a
37.82 months 14.01 14.63 months 6.16 6.77 p ⬍.001
Age first phrases 52.07 months 16.04 26.21 months 11.53 5.98 p ⬍.001
ADI-R algorithm scores
b
ADI total 40.29 8.38 37.83 8.21 1.28 .20
ADI social 18.79 5.06 18.38 4.82 .36 .72
ADI communication 15.64 3.22 14.19 4.10 1.69 .09
ADI stereotypies 5.85 1.59 5.26 1.53 1.64 .11
a
Good estimates of age for first words/phrases available for 29 in the autism group, 19 in the Asperger group;
t values given for unequal variances.
b
For all ADI ratings, the higher the score the more abnormal the behavior.
1
These are similar to the items included in the diagnostic algorithm
but exclude items for which there are only childhood ratings, that
is, direct gaze, spontaneous imitation, play (imaginative, imitative
and group), interest in/response to other children.
DISCUSSION
The findings of this comparison between 34 adults
with autism and 42 designated as having “Asperger syn-
drome,” matched for age and nonverbal IQ, indicate that
parents do report certain differences between the groups
when they are younger. Because participants were as-
signed to the two groups on the basis of their early lan-
guage development, the fact that there were significant
differences in the age at which they first used words or
phrases is a consequence of group selection. This method
of sample selection would also explain why parents of
children with autism were more concerned about lan-
guage delays initially. However, there were also several
other differences.
Children with Asperger syndrome tended to be
older when parents first noted abnormalities in their de-
velopment (average age 21 months as compared to an
average of 15 months in the autism group) and their
parents’ first concerns were split fairly evenly between
general behavior problems, ritualistic and stereotyped
behaviors/interests, and motor delays/difficulties. In the
autism group, social difficulties and motor delays were
(after language delays) the problems most likely to give
Implications for the Differentiation Between Autism and Asperger Syndrome 9
rise to early concerns. Only one parent in this group
remembered ritualistic/stereotyped behaviors as being
the most prominent symptoms in the first 3 years (cf.
Howlin & Asgharian, 1998).
However, ADI-R algorithm scores for social and
communication difficulties, which are based largely
(although not exclusively) on children’s difficulties in
these areas when they are somewhat older (4–5 years)
suggested that these early differences may have de-
creased over time. Thus, no significant group differences
were found on the ADI-R algorithm scores for impair-
ments in social or communication skills. Moreover, rat-
ings for ritualistic/stereotyped patterns of behavior or
interests (which are based on whether the behavior has
ever occurred) did not differ between the groups. The
suggestion that early differences may decrease with age
is also supported by the findings of recent studies
(Gilchrist et al., 2001; Ozonoff et al., 2000; Szatmari
et al., 1995). In the Gilchrist study, for example (al-
though participants differed in age and full-scale IQ),
the significant group differences in ADI scores at age
4–5 years had disappeared by adolescence or early adult-
hood.
Motor clumsiness (based on ADI “ever” and “cur-
Table IV. Comparison Between Asperger and High-Functioning Autism Groups on Measures of Adult Func-
tioning
Autism (n ⫽34) Asperger (n ⫽42)
Group Mean SD Mean SD t pvalue
Nonverbal IQ 100.06 16.79 101.50 18.38 ⫺.11 .91
Current ADI ratings
ADI total 24.68 7.25 23.58 5.91 1.29 .20
ADI social 11.09 3.71 11.51 3.75 .36 .72
ADI communication 8.77 3.47 7.98 2.62 1.69 .09
ADI stereotypies 4.81 1.82 4.09 1.48 1.64 .11
Social functioning
Education: n with A level 8 22 6.55 .01
qualifications or above
Jobs: n in employment 13 23 1.25 .26
(independent/sheltered)
a
Living: n semi/independent 11 15 .09 .76
Friendships: n with some friends 13 17 .04 .84
Language scores
BPVS age equivalent
b
15.16 yr 4.11 16.09 yr 4.17 ⫺.91 .37
BPVS rating
c
1.59 1.37 1.40 1.52 .51 .61
EOWPVT age equivalent 15.40 yr 3.52 16.60 yr 2.63 ⫺1.54 .13
EOWPVT rating 1.38 1.24 .82 .79 2.15 .04
a
Excludes 6 individuals in the autism group and 4 in the Asperger group who were still at college.
b
BPVS/EOWPVT age scores based on 29 individuals in the autism group and 35 in the Asperger group.
c
BPVS/EOWPVT ratings based on age bands on tests; see text for details. Lower ratings ⫽ higher level.
rent” ratings) failed to show any difference between the
groups in the present study, although this is one area that
has frequently been claimed to characterize individuals
with Asperger syndrome (cf. Volkmar & Klin, 2000).
Assessments of current functioning also failed to
reveal major differences between the two groups. The
number of individuals with an additional psychiatric di-
agnosis was similar in both groups, and although these
diagnoses were not independently confirmed, depressive
disorders appeared to be the most common. “Current”
ADI-R ratings for social functioning, communication,
and ritualistic/stereotyped behaviors were remarkably
similar, and outcome in terms of friendships, employ-
ment, and independent living did not differ. Only two
individuals in the Asperger group and three in the autism
group were reported to have developed friendships that
involved closeness and sharing of emotions, although a
higher number (10 with autism, 15 with Asperger syn-
drome) did have acquaintances whom they met outside
home or work. Three men (2 autistic, 1 Asperger) were
married or living with a partner, and one of the mar-
ried men in the autism group had children. One woman
in the Asperger group was divorced. The majority in both
groups (19 with autism, 26 with Asperger syndrome) still
lived with their parents and of those who did live inde-
pendently most (7/11 in the Autism group; 9/15 in the
Asperger group) continued to require support either
from their families or social services. Three people with
autism and one with Asperger syndrome were in spe-
cialist residential provision, and one man with autism
was in a extended-care psychiatric hospital. In terms of
employment, 15 individuals in each group had never
had a job, and only 2 in each group had relatively well-
paid, permanent employment. Three individuals in each
group were in sheltered placements or worked with a
family firm; the remainder was in short-term, low-pay,
or voluntary posts that did not provide them with suf-
ficient money to live independently. The only area in
which a significant group difference emerged was in
academic attainments. Although the numbers without
any formal qualifications were similar (8 in the As-
perger group, 9 in the autism group) eight individuals
in the Asperger group had a university degree (4 of
these at postgraduate level) and a further 14 had ob-
tained some formal qualifications at sixth form level (A
levels or equivalent). In the autism group only 2 people
had a degree (none at postgraduate level) and 6 others
had A levels or equivalent. However, this advantage in
terms of academic attainments did not seem to have re-
sulted in higher levels of achievement in later life.
Adult assessments of language comprehension and
expression also showed some superiority for the As-
10 Howlin
perger group, although the differences were generally
small and only (marginally) significant for the ratings
based on their scores on the Expressive One Word
Picture Vocabulary (EOWPVT). The marked delays in
language experienced by those in the autism group as
children (several of whom did not begin to use words
until they were over 4 years or phrases until they were
5 or older) may mean that they remained at a dis-
advantage in their ability to “catch up” linguistically.
Thus the small differences found here may not represent
a true diagnostic “splitting” between the groups.
Rather, they could be due principally to the initial
severity of language problems, which then leads, as
Szatmari (2000) suggests, to the individuals in each
of the two groups following somewhat different tra-
jectories over time.
The language test results also raise further problems
in terms of the validity of the distinction between autism
and Asperger syndrome. Thus, although the latter con-
dition is, by definition, not associated with early lan-
guage delays, the present study indicated that language
skills in adulthood were often far from age appropriate.
The low age ceilings of the EOWPVT (19 years) and
the British Picture Vocabulary Scale (BPVS) (18 years)
give rise to obvious problems when calculating the ex-
tent of the language delay in older individuals. Never-
theless, 57% of the Asperger group still scored below
ceiling on the BPVS (a measure of receptive vocabu-
lary) and 62% scored below ceiling on the EOWPVT
(a measure of expressive vocabulary). In the autism
group 76% were below the age ceiling on the expres-
sive test and 69% on the comprehension test (X
2
non-
significant for each comparison between groups).
Amongst the 21 individuals with Asperger syndrome
whose scores were below the ceiling age of the test (i.e.,
those for whom an exact language age score rather than
an extrapolation was available), the average gap be-
tween chronological age and language comprehension
age was 12.62 years (SD 6.29 years) and the gap be-
tween chronological age and expressive language age
was 10.20 years (SD 6.10 years). In the autism group the
gap between chronological age and comprehension and
expressive ages for those scoring below the age ceiling
(n ⫽22) was similar (difference between comprehen-
sion age and CA ⫽13.33 years, SD 9.53 years; differ-
ence between expressive age and CA ⫽11.67 years,
SD 9.34 years). There were no significant group dif-
ferences on either measure. Moreover, although as
young children, individuals in the Asperger group were
reported not to have shown clinically significant de-
lays in language, in adulthood either their compre-
hension and/or vocabulary use remained well below
chronological age in the majority of cases.
The present study suffers from a number of limi-
tations, which should be borne in mind when drawing
conclusions about possible differences between these
two diagnostic groups. First, and perhaps most impor-
tant, the group designated as having Asperger syndrome
would not meet this diagnosis using strict DSM-IV
(TR) criteria. However, in common with other recent
research (Eisenmajer et al., 1996; Manjiviona & Prior,
1995; Mayes & Calhoun 2001; Miller & Ozonoff, 2000;
Szatmari et al.,1995) so few cases could be identified
from the patient database who met strict criteria, that
the modified criteria used in many of the studies cited
in Tables I and II were applied instead. There is no in-
dication that participants in either the autism or “As-
perger” groups in this study were markedly different
in terms of cognitive level than the participants in com-
parable investigations of autism versus Asperger syn-
drome (reported mean performance IQs in the studies
reported in Tables I and II typically range from around
the mid 80s to just over 100). Thus the groups appear
to be fairly representative of the samples involved in
similar research. Although these findings might not be
replicated if individuals meeting strict DSM-IV crite-
ria were to be included, a recent study by Manjiviona
and Prior (1999) found that the results of group com-
parisons remained the same whether strict or modified
DSM-IV criteria were used for classification.
Secondly, information on early development was
based on retrospective accounts from parents rather than
on contemporaneous data. In particular, the classifica-
tion of individuals as having/not having early language
delays was dependent on parental memory, not on for-
mal assessments in childhood. IQ scores, too, were
based on current performance and on nonverbal tests
only; data on childhood IQ were generally not avail-
able. However, recent research (Howlin, Goode, Hul-
ton & Rutter, in press) indicates that, particularly among
high-functioning individuals with autism/Asperger syn-
drome, IQ scores remain remarkably stable from child
to adulthood, and hence it is likely that childhood IQ
scores would have been within a similar range.
Thirdly, because ADI and other data were obtained
in the course of routine clinical assessments it was not
possible to make audio or video recordings that would
allow interviews to be coded by independent raters.
A further problem, and one affecting almost all
studies of linguistic competence in adulthood, is the
lack of standardized language measures that extend be-
yond the age of 18 or 19 years. Thus, assessments of
functioning in this area tend to be limited in the range
of language skills they measure and ceiling effects
Implications for the Differentiation Between Autism and Asperger Syndrome 11
mean that scoring is invariably prone to some degree
of error.
Finally, assessments of clumsiness, a variable that
has been said to distinguish between individuals with
Asperger syndrome and those with autism, were lim-
ited to ADI-R ratings; no direct measures were used.
Nevertheless, the results of the study are based on
considerably larger samples than those involved in most
previous comparisons of matched IQ groups; age and
gender ratios were similar in both groups, and outcome
measures examined functioning across a range of dif-
ferent domains. Unlike most previous studies in this
area, which have focused on children, the participants
were all over 18 years of age, thus allowing an exam-
ination of differences amongst adults. Overall, the find-
ings indicate that there may be some group differences
in the early years (i.e., in the symptoms reported by
parents by the age of 3). ADI-R algorithm scores, how-
ever, suggested that such differences may decrease with
age, and in adulthood there were no marked differences
between the groups on “Current” ADI-R scores, other
ratings of social outcome or standardized tests.
SUMMARY
In summary, the findings of this study and of the
accompanying literature review suggest that compar-
isons between high-functioning individuals with autism
and Asperger syndrome can only produce meaningful
conclusions if the two groups are appropriately matched
for intellectual level. Moreover, the age at which data
are collected may well influence results, with differences
between the two groups probably becoming less evident
as they grow older (Gilchrist et al., 2001; Ozonoff et al.,
2000; Szatmari et. al., 1995). Other authors (Szatmari,
2000; Wing, 2000) have reported that young children
with autism may “shift” to follow an “Asperger-type”
pathway subsequently, especially if they develop flu-
ent language skills. The slight superiority of individu-
als with Asperger syndrome with regard to language
expression has been replicated in a number of previous
comparative studies. However, the fact that individu-
als with a diagnosis of autism are so much further be-
hind in their language development from the start may
also be the prime reason for the relative severity of their
language difficulties in later life (see also Eisenmajer
et al.,1996; Szatmari, et al., 2000).
The present results do not support the view that
HFA and Asperger syndrome (at least as defined by the
modified criteria adopted here) are distinct conditions.
They also indicate that it may be misleading to suggest
(as do DSM-IV/ICD-10) that children with Asperger
syndrome do not have significantly delayed language
skills, as in adulthood their level of language (as well
as their social use of language) is in fact often markedly
impaired. These findings are of clinical importance
since they not only highlight the lack of any substantial
difference between the two diagnostic groups in adult-
hood, but also illustrate the poor prognosis, even for
high-functioning individuals within the autistic spec-
trum. Despite having IQ scores well within the normal
range (and, sometimes, reaching quite high academic
levels) the majority of individuals in both groups had no
close friends, they remained highly dependent on their
families for support, and employment status was low.
Larger-scale prospective studies are still required
in order to clarify the extent and nature of the possi-
ble differences between these two relatively high-
functioning groups. However, for the present, the
weight of evidence seems to suggest that any distinc-
tions that do exist in early childhood may become less
marked with age. There is little to support the view that
individuals with a diagnosis of Asperger syndrome
should be deprived of the support and services that are
available for those with a diagnosis of autism, or that
educational and management programs should differ in
any way. Instead, it should be recognized that for all
high-functioning individuals with an autistic spectrum
disorder there is a need for much improved services
throughout childhood and adulthood, if the long-term
outcome is to be significantly enhanced.
REFERENCES
American Psychiatric Association. (1994). Diagnostic and statisti-
cal manual of mental disorders (4th Ed.), Washington, DC:
Author.
American Psychiatric Association. (2000). Diagnostic and statisti-
cal manual of mental disorders (4th ed.—Text Revision), Wash-
ington, DC: Author.
Department of Health. (2001). Valuing people: A new strategy for
learning disability in the 21st century. London: Department Of
Health White Paper.
Dunn, L., Dunn, L., Whetton, C., & Burley, J. (1997). British Pic-
ture Vocabulary Scale: Revised. Slough, Bucks: NFER-Nelson
Publishing Co.
Dunn, L., Dunn, L., Whetton, C., & Pintillie D. (1982). British Pic-
ture Vocabulary Scale: Slough, Bucks: NFER-Nelson Publish-
ing Co.
Ehlers, S., Nyden, A., Gillberg, C., Sandberg, A. D., Dahlgren, S. O.,
Hjelmquist, E., & Oden, A. (1997). Asperger syndrome, autism
and attention disorders: A comparative study of the cognitive
profiles of 120 children. Journal of Child Psychology and Psy-
chiatry, 38, 207–217.
Eisenmajer, R., Prior. M., Leekam, S., Wing, L., Gould, J., Welham,
M., & Ong, B. (1996). Comparison of clinical symptoms in
autism and Asperger’s disorder. Journal of the American Acad-
emy of Child and Adolescent Psychiatry, 35, 1523–1531.
12 Howlin
Gardner, M. (1982). Expressive One Word Picture Vocabulary Test
(Upper Extension). Los Angeles: Western Psychological
Services.
Ghaziuddin, M., & Butler, E. (1998). Clumsiness in autism and
Asperger syndrome: A further report. Journal of Intellectual
Disabilities Research, 42, 43–48.
Ghaziuddin, M., Butler, E., Tsai, L., & Ghaziuddin, N. (1994). Is
clumsiness a marker for Asperger’s syndrome? Journal of
Intellectual Disabilities Research, 38, 519–527.
Ghaziuddin, M., & Gerstein, L. (1996). Pedantic speaking style dif-
ferentiates Asperger syndrome from high-functioning autism.
Journal of Autism and Developmental Disorders, 26, 585–595.
Ghaziuddin, M., Leininger, L., & Tsai, L. (1995). Brief report:
Thought disorder in Asperger syndrome—comparison with high-
functioning autism. Journal of Autism and Developmental Dis-
orders, 25, 311–317.
Ghaziuddin, M., Shakal, J., & Tsai, L. (1995). Obstetric factors in
Asperger syndrome: Comparison with high functioning autism.
Journal of Intellectual Disabilities Research, 39, 538–543.
Ghaziuddin, M., Tsai, L. Y. & Ghaziuddin, N. (1992). Brief report:
A reappraisal of clumsiness as a diagnostic feature of Asperger
syndrome. Journal of Autism and Developmental Disorders, 22,
651–656.
Gilchrist, A., Green, J., Cox, A., Rutter, M., & Le Couteur, A. (2001).
Development and current functioning in adolescents with
Asperger syndrome: A comparative study. Journal of Child
Psychology and Psychiatry, 42, 227–240.
Gillberg, C., & Gillberg, I. C. (1989). Asperger syndrome: Some epi-
demiological aspects—a research note. Journal of Child Psy-
chology and Psychiatry, 30, 631–638.
Howlin, P., and Asgharian, A. (1999). The diagnosis of autism and
Asperger syndrome: Findings from a survey of 770 families.
Developmental Medicine and Child Neurology, 41, 834–839.
Howlin, P., Goode, S., Hulton, J., & Rutter, M. (in press). A cogni-
tive and behavioural study of outcome in young adults with
autism. Journal of Child Psychology and Psychiatry.
Iwanaga, R., Kawasaki, C., & Tsuchida, R. (2000). Comparison of
sensory-motor and cognitive function between autism and As-
perger syndrome in pre-school children. Journal of Autism and
Developmental Disorders, 30, 16–174.
Klin, A. (2000). Attributing social meaning to ambiguous social stim-
uli in higher-functioning autism and Asperger syndrome: The
Social Attribution Task. Journal of Child Psychology and Psy-
chiatry, 41, 831–846.
Kim, J. A., Szatmari, P., Bryson, S. E., Streiner, D. L., and Wilson,
F. J. (2000). The prevalence of anxiety and mood problems
among children with autism and Asperger syndrome. Autism:
International Journal of Research and Practice, 4, 117–132.
Klin, A., & Volkmar, F. R. (2000). Treatment and intervention
guidelines for individuals with Asperger syndrome. In A. Klin,
F. K. Volkmar, & Sara S. Sparrow (Eds.), Asperger Syndrome
(pp. 340–366). New York: The Guildfold Press.
Klin, A., Volkmar, F. R., & Sparrow, S. S. (Eds.) (2002). Asperger
syndrome. New York: Guildford Press.
Klin, A., Volkmar, F. R., Sparrow, S. S., Cicchetti, D. V., & Rourke,
B. P. (1995). Validity and neuropsychological characterization
of Asperger syndrome: convergence with nonverbal learning dis-
abilities syndrome. Journal of Child Psychology and Psychiatry
36, 1127–1140.
Kugler, B. (1998). The differentiation between autism and Asperger
syndrome. Autism: International Journal of Research and Prac-
tice, 2, 11–32.
Kurita, H. (1997). A comparative study of Asperger syndrome with
high-functioning atypical autism. Psychiatry and Clinical Neuro-
science, 51, 67–70.
Leekham, S., Libby, S., Wing, L., Gould, J., & Gillberg, C. (2000).
Comparison of ICD-10 and Gillberg’s criteria for Asperger syn-
drome. Autism: International Journal of Research and Practice,
4, 11–28.
Lord, C., Rutter, M., & Le Couteur, A. (1994). Autism Diagnostic
Interview—Revised: A revised version of a diagnostic interview
for caregivers of individuals with possible pervasive develop-
mental disorders. Journal of Autism and Developmental Disor-
ders, 24, 659–685.
Manjiviona, J., & Prior, M. (1995). Comparison of Asperger syndrome
and high functioning autistic children on a test of motor impair-
ment. Journal of Autism and Developmental Disorders, 25, 23–29.
Manjiviona, J., & Prior, M. (1999). Neuropsychological profiles of
children with Asperger syndrome and autism. Autism: Interna-
tional Journal of Research and Practice, 3, 327–354.
Mawhood L. M., Howlin, P., & Rutter M. (2000). Autism and de-
velopmental receptive language disorder: A follow-up compar-
ison in early adult life. I: Cognitive and language outcomes.
Journal of Child Psychology and Psychiatry, 41, 547–559.
Mayes S. D., & Calhoun, S. L. (2001). Non-significance of early
speech delay in children with autism and normal intelligence
and implications for DSM-IV Asperger’s disorder. Autism:
International Journal of Research and Practice, 5, 81–94.
Miller, J. N., & Ozonoff, S. (2000). The external validity of Asperger
disorder: Lack of evidence from the domain of neuropsychol-
ogy. Journal of Abnormal Psychology, 109, 227–238.
Ozonoff, S., Rogers, S. J., & Pennington, B. F. (1991). Asperger’s
syndrome: Evidence of an empirical distinction from high func-
tioning autism. Journal of Child Psychology and Psychiatry, 32,
1107–1122.
Ozonoff, S., South, M., & Miller, J. N. (2000). DSM-IV defined As-
perger syndrome: Cognitive, behavioural and early history dif-
ferentiation from high-functioning autism. Autism: International
Journal of Research and Practice, 4, 29–46.
Pomeroy, J. C. (1998). Subtyping of pervasive developmental disor-
ders: Issues of validity and implications for child psychiatric di-
agnosis. In E. Schopler, G. B. Mesibov, & L. J. Kunce. Asperger
syndrome or high functioning autism? New York: Plenum.
Raven, J. (1956). Standard Progressive Matrices. London: H. K.
Lewis and Co. Ltd.
Rinehart, N. J., Bradshaw, J. L., Moss S. A., Brereton, A. V., &
Tonge, B. J. (2001). A deficit in shifting attention present in
high-functioning autism but not Asperger’s disorder. Autism:
International Journal of Research and Practice, 5, 67–80.
Implications for the Differentiation Between Autism and Asperger Syndrome 13
Schopler, E., Mesibov G. B., & Kunce L.J. (1998). Asperger syn-
drome or high functioning autism? New York: Plenum.
Szatmari, P. (2000). Perspectives on the classification of Asperger
syndrome. In A. Klin, F. R., Volkmar, & S. S. Sparrow (Eds.).
Asperger syndrome (pp. 403–417). New York, Guildford Press.
Szatmari, P., Archer, L., Fisman, S., Streiner, D., and Wilson, F.
(1995). Asperger’s syndrome and autism: Differences in be-
haviour, cognition and adaptive functioning. Journal of the
American Academy of Child and Adolescent Psychiatry, 34,
1662–1670.
Szatmari, P., Bartolucci, G., & Bremner, R. (1989). Asperger’s syn-
drome and autism: Comparison of early history and outcome.
Developmental Medicine and Child Neurology, 31, 709–720.
Szatmari, P., Tuff, L., Finlayson, M. A. J., & Bartolucci, G. (1990).
Asperger’s syndrome and autism: Neurocognitive aspects. Jour-
nal of the American Academy of Child and Adolescent Psychi-
atry, 29, 130–136.
Tonge, B., Brereton, A. V., Gray, K. M., and Einfeld, S. (1999). Be-
havioural and emotional disturbance in high functioning autism
and Asperger syndrome. Autism: International Journal of Re-
search and Practice, 3, 117–130.
Volkmar, F. R., & Klin, A. (2000). Diagnostic issues in Asperger
syndrome. In A. Klin, F. K. Volkmar, & S. S. Sparrow (Eds.),
Asperger syndrome (pp. 25–71). New York: The Guildford
Press.
Wechsler, D. (1981). Wechsler Adult Intelligence Scale—Revised.
San Antonio, TX: The Psychological Corporation.
Wechsler, D. (1997). Wechsler Adult Intelligence Scale—Third edi-
tion. San Antonio, TX: The Psychological Corporation.
Wechsler, D. (1999). Wechsler Abbreviated Scale of Intelligence.
San Antonio, TX: The Psychological Corporation.
Wing, L. (1981). Asperger’s syndrome: A clinical account. Psycho-
logical Medicine, 11, 115–129.
Wing, L. (2000). Past and Future of Research on Asperger Syn-
drome. In A. Klin, F. K. Volkmar, & S. S. Sparrow (Eds.), As-
perger syndrome (pp. 418–432). New York: The Guildford
Press.
World Health Organization (1992). The ICD-10 Classification of
Mental and Behavioural Disorders: Clinical Descriptions and
Diagnostic Guidelines. Geneva: World Health Organization.
