Dyspraxia in autism:
valuable insights into developmental disorders of the brain.1
association with
motor, social, and
communicative deficits
M A Dziuk, University of Texas Health Science Center at San
Antonio, Texas;
J C Gidley Larson;
A Apostu;
E M Mahone;
M B Denckla;
S H Mostofsky*, Kennedy Krieger Institute, Departments of
Neurology and Psychiatry, Johns Hopkins University School
of Medicine, Baltimore, MD, USA.
*Correspondence to last author at Kennedy Krieger
Institute, 707 N Broadway, Baltimore, MD 21205, USA.
E-mail: mostofsky@kennedykrieger.org
Impaired performance of skilled gestures, referred to as
dyspraxia, is consistently reported in children with autism;
however, its neurological basis is not well understood. Basic
motor skill deficits are also observed in children with autism
and it is unclear whether dyspraxia observed in children with
autism can be accounted for by problems with motor skills.
Forty-seven high-functioning children with an autism spectrum
disorder (ASD), autism, or Asperger syndrome (43 males, four
females; mean age 10y 7m [SD 1y 10m], mean Full-scale IQ
(FSIQ) 99.4 [SD 15.9]), and 47 typically developing (TD)
controls (41 males, six females; mean age 10y 6m [SD 1y 5m],
mean FSIQ 113.8 [SD 12.3], age range 84y) completed: (1) the
Physical and Neurological Assessment of Subtle Signs, an
examination of basic motor skills standardized for children, and
(2) a praxis examination that included gestures to command, to
imitation, and with tool-use. Hierarchical regression was used to
examine the association between basic motor skill performance
(i.e. times to complete repetitive limb movements) and praxis
performance (total praxis errors). After controlling for age and
IQ, basic motor skill was a significant predictor of performance
on praxis examination. Nevertheless, the ASD group continued
to show significantly poorer praxis than controls after
accounting for basic motor skill. Furthermore, praxis
performance was a strong predictor of the defining features of
autism, measured using the Autism Diagnostic Observation
Schedule, and this correlation remained significant after
accounting for basic motor skill. Results indicate that dyspraxia
in autism cannot be entirely accounted for by impairments in
basic motor skills, suggesting the presence of additional
contributory factors. Furthermore, praxis in children with
autism is strongly correlated with the social, communicative,
and behavioral impairments that define the disorder, suggesting
that dyspraxia may be a core feature of autism or a marker of
the neurological abnormalities underlying the disorder.
See end of paper for list of abbreviations.
734 Developmental Medicine & Child Neurology 2007, 49: 734739
Abnormalities on motor examination have often afforded
Such abnormalities have been well documented in individuals
with autism, and date back to some of the original descriptions
of the disorder (see Gidley Larson and Mostofsky2 for review).
Increased insight into brain mechanisms underlying autism
can be gained from careful consideration of these motor
signs.3 By using tests of motor function for which the neuro-
logical basis is well mapped out, it is possible to gain an under-
standing of the neural circuits impaired in autism. Motor signs
can also serve as markers for deficits in parallel brain systems
which are important for control of social and communication
skills, impairments which characterize autism.
Motor deficits associated with autism appear to be evident
as early as infancy, manifesting as problems in sequencing
movements to crawl or walk.4 In older children with autism,
problems have been observed in basic motor control, includ-
ing poor coordination (both axial and limb) and postural
control, slow response speed, clumsy gait, and low tone.57
Difficulty with performance of skilled motor gestures has
also been observed, and is one of the most consistently
reported motor findings in children with autism.8,9 Im-
pairments in imitation of skilled motor gestures have been
particularly emphasized, leading some to suggest that
impaired imitation may be a core feature of autism, con-
tributing to abnormal development of functions critical to
social and communicative development, such as empathy,
joint attention, and theory of mind.10 Others have since
hypothesized that autism may be related to abnormalities in
the mirror neuron system critical to imitation, and associated
deficits in selfother mapping.11
Comprehensive investigation of skilled motor gestures
using a traditional praxis examination, however, reveals that
children with autism show deficits in performance not only
during imitation, but also in response to command and with
tool-use.8,9 These findings suggest that autism may be associ-
ated with a generalized praxis deficit, rather than a deficit
specific to imitation. In a developmental context, impaired
performance of skilled gestures, including those involving
imitation, may, therefore, be secondary to abnormalities in
frontal/parietal-subcortical circuits important for acquisition
of sensory representations of movement and/or the motor
sequence programs necessary to execute them.
Given these findings, it is important to consider the associa-
tion between basic motor skill deficits and impaired per-
formance of skilled motor gestures. In traditional adult models,
the terms apraxia and dyspraxia are reserved for individuals
who demonstrate impaired ability to perform skilled motor
tasks despite relatively normal motor dexterity.12 In contradis-
tinction to the adult literature, findings from typically develop-
ing (TD) children and those with developmental coord-
ination disorder reveal impaired performance on praxis exami-
nation and are associated with deficits in motor coordination.13
Given that children with autism spectrum disorders (ASD),
including high-functioning autism and Asperger syndrome,
also show deficits in basic aspects of motor execution,6 a ques-
tion remains regarding the association between basic motor
coordination and dyspraxia in autism: whether dyspraxic
errors made by children with ASD are attributable to these basic
motor impairments or whether they represent a distinct
impairment of gesture. The finding that body-part-for-tool
(BPT) errors account for a larger proportion of praxis errors in
children with ASD than in TD controls suggests a distinct
impairment of gesture that is unlikely to be due to impaired
motor execution. These errors occur when the participant
uses a body part as the tool rather than using the body part to
hold the tool (e.g. when asked to show how to cut with scis-
sors, the participant moves two fingers in a scissoring motion
rather than pretending to actually hold the scissors).9
There has been limited prior examination of the direct asso-
ciation between basic motor skill deficits and dyspraxia in
autism. To understand this association better, the current
authors examined the correlation between performance on a
traditional praxis examination and a measure of basic motor
skill1,14 in children with ASD and a group of TD children. We
hypothesized that children with ASD would show impairments
on praxis performance above what could be accounted for by
basic motor skill deficits as compared with controls.
The association between basic motor skill deficits and dys-
praxia in autism was further investigated by examining correla-
tions with the Autism Diagnostic Observation ScheduleG
(ADOS-G)15 scores. While basic motor skill deficits and dys-
praxia have been reported in several previous studies, none
has directly examined the relationship between these findings
and the social, communicative, and behavioral impairments
that define the disorder. It was hypothesized that for children
with autism, worse praxis performance would correlate with
higher ADOS-G scores and that this correlation would remain
significant after accounting for basic motor skill performance.
Further, we hypothesized that higher ADOS-G scores would
correlate not only with performance of gestures to imitation,
but also with gestures to command and with tool use on the
praxis examination.
Method
PARTICIPANTS
Ninety-four children, aged 8 to 14 years, participated. Forty-
seven participants were children with ASD (43 males, four
females; mean age 10y 7mo [SD 1y 10mo]) and 47 were TD
children (41 males, six females; mean age 10y 6mo [SD 1y
5mo]). Children with ASD met Diagnostic and Statistical
Manual of Mental DisordersIV16 criteria for either autism
(n=22) or Asperger syndrome (n=25). Children with high-
functioning autism and Asperger syndrome have been found
to show similar degrees of impairment on the measures used
in this study to assess praxis and basic motor skills;6,9 they
were, therefore, collapsed into a single ASD group. ASD diag-
noses were confirmed using the Autism Diagnostic Inter-
viewRevised (ADI-R)17 and the ADOS-G, Module 3.15 Eight
of the children with autism did not receive the ADOS; for five
of those participants, diagnosis was confirmed using the
ADIR and clinical impression, and for three of those partici-
pants, diagnosis was based on clinical impression.
ASD participants were recruited from outpatient clinics at
the Kennedy Krieger Institute in Baltimore, local Autism
Society of America chapters, flyers at schools, social skills
groups, pediatricians offices, and by word of mouth. None
of the participants had a history of seizures, traumatic brain
injury, or mental retardation.*
TD controls had no history of neurological or psychiatric
diagnoses and no immediate family member with a pervasive
developmental disorder. Psychiatric diagnoses were ruled-out
*UK usage: learning disability.
using a structured parent interview: the Diagnostic Interview
for Children and AdolescentsIV.16
Intellectual functioning was assessed using the Wechsler
Intelligence Scale for Children 3rd edition (WISC-III)18 or
the 4th edition (WISC-IV).19 Full-scale IQ (FSIQ) scores from
the WISC-III were transposed to the WISC-IV scale as derived
from the WISC-IV technical report number 2.20 Participants
had an FSIQ above 80 with the exception of three children.
While these three participants had an FSIQ lower than 80,
either their verbal or perceptual index scores were greater
than 90. TD children had an average FSIQ of 113.8 (SD 12.3)
and children with ASD an average FSIQ of 99.4 (SD 15.9).
The Johns Hopkins Medicine Institutional Review Board
approved the study. Written consent was obtained from parents
of all participants and the participants gave written assent.
PRAXIS EXAMINATION
A version of the Florida Apraxia Screening Test (Revised),21
adapted for children,9 was used to examine the performance of
purposeful, skilled movements. The examination consists of
three sections: (1) gestures to command (GTC); (2) gestures to
imitation (GTI); and (3) gestures with tool use (GTU). The GTC
and GTI sections include both transitive gestures (those that
act on or with an object, e.g. hammering a nail) and intransitive
gestures (those that do not act on or with an object, e.g. waving
goodbye); the GTU section contains only transitive gestures.
The GTI section also includes meaningless gestures. A detailed
description of the praxis battery (including types of gestures) is
outlined in Mostofsky et al.9
Participants were videotaped and later scored by two inde-
pendent raters blind to participant diagnosis. Gestures were
first scored as correct or incorrect. If a gesture was incorrect, it
was evaluated for five categories of errors: (1) spatial (external
configuration, internal configuration, amplitude, movement);
(2) temporal (sequencing, temporal, occurrence); (3) con-
tent/concretization (perseverative, related, non-related, hand,
concretization); (4) BPT; or (5) other (no response, unrecog-
nizable response). Each gesture could have more than one
type of error. The total number of correct gestures on the entire
examination was calculated and the number of errors in each
of the three sections was totaled for each of the five error cate-
gories (i.e. total spatial errors for GTI, total BPT errors for GTC,
etc.). Total raw scores in each of the three sections were aver-
aged between the two raters. Acceptable rates of reliability have
been established for these measures. Interrater reliability sta-
tistics and further details regarding scoring of the praxis exami-
nation are outlined in Mostofsky et al.9
BASIC MOTOR SKILL EXAMINATION
Each participant completed the Physical and Neurological
Assessment of Subtle Signs (PANESS), a neurological examina-
tion for children used to evaluate subtle neurological signs,
which is standardized for age, sex, and handedness.1,14 The
revised PANESS1 includes a lateral preference assessment,
measures of axial motor skill, and measures of appendicular
(limb) motor skill. Axial tasks include gaits on heels, toes, and
sides of feet; tandem gait forward/backward; standing/hopping
on one foot; standing heel-to-toe with eyes closed; and stand-
ing both feet together, arms outstretched with eyes closed.
Limb motor skill assessment comprises timed measures of
repetitive and sequential movements of the hands and feet,
including toe-tapping, alternating heel-toe tapping, repetitive
Dyspraxia Associations in Autism M A Dziuk et al. 735
hand patting, hand pronation-supination, repetitive finger
tapping, and finger sequencing each of which was per-
formed on the right and left sides.
The PANESS has been found to have adequate testretest
reliability,22 interrater reliability, and internal consistency.23
The PANESS was chosen because it was developed to mini-
mize the need for equipment, to eliminate time-consuming
and less reliable sensory tasks, and to be completed in only
15 to 20 minutes. These parameters were considered impor-
tant factors when studying behavior in children with ASD.
Given that praxis involves purposeful limb movements, a
measure of timed repetitive movements of the hands and feet
was used to examine the effect of basic motor skill on praxis
performance. We did not survey all motor elements, choosing to
focus on rapid distal limb control reflecting frontal and frontal-
subcortical elements basic to incorporation of motor skill into
items assessed on the praxis battery. For each repetitive
movement (finger tapping, hand patting, and toe-tapping on
right and left sides), the time to complete 20 movements was
recorded; these times were summed to provide a total mea-
sure of speed of timed repetitive movements.
STATISTICAL ANALYSES
Univariate analysis of variance was used to evaluate for differ-
ences between TD children and children with ASD in relation
to age, FSIQ, PANESS scores, and praxis examination total.
Sex distributions between groups were examined using 2
analyses. Throughout the analyses, a significance level of p<0.05
(two-tailed) was used. Total praxis errors score was negatively
skewed in the control group, but not the ASD group. Therefore,
for analyses comparing groups on praxis performance, a log
transformation of the total praxis errors scores was used. For
analyses using the autism group alone, the variable was not
transformed. All other variables examined were normally
distributed.
A hierarchical regression analysis was used to examine the
association of basic motor skill performance and diagnosis
with praxis performance, after accounting for age and FSIQ.
For the main regression analysis, assumptions of linear regres-
sion were examined. Residuals for each of the predictor vari-
ables were consistent across values, showing homoscedasticity.
Error terms were normally distributed and were without signif-
icant autocorrelation (DurbinWatson statistic for n=94 with
four independent variables=1.01). Log transformation of the
total errors score on the praxis examination was used as the
dependent variable. Age was entered as the first independent
variable, followed by FSIQ. Age can be a strong predictor of
Table I: Group performance on examinationsa
Examination Autism Control p 2
type Mean SD Mean SD
Total praxis errors 36.9 21.9 17.2 10.1 <0.001
Log transformation
(total praxis errors) 1.5 0.3 1.2 0.2 <0.001
PANESS repetitive total (s) 38.2 6.1 33.5 5.7 <0.001
PANESS total score 35.3 11.2 18.3 10.1 <0.001
aComparisons are using t-tests (two-tailed). 2, effect size (eta squared);
PANESS, Physical and Neurological Assessment of Subtle Signs.
736 Developmental Medicine & Child Neurology 2007, 49: 734739
both repetitive movement speed and praxis performance in
children, although both reach relative plateaus by 8 years of
age.1,9 Nevertheless, age was entered as the first predictor in
the regression model to ensure that any effects of age were
controlled. FSIQ has also been shown to correlate with
motor skill performance,24 and the autism group had a sig-
nificantly lower FSIQ than the control group. Therefore,
FSIQ was included as a second predictor in the hierarchical
regression model. The measure of basic motor skill (raw
score of total timed repetitive movements from the PANESS)
was entered third, followed by primary diagnosis as the
fourth predictor. The examination of diagnosis as the final
predictor variable provided a means of determining whether
children with autism continue to show significantly worse
praxis performance than controls, even after accounting for
age, IQ, and basic motor skill deficits.
Within the ASD group, separate linear regression analyses
were used to examine the association of both basic motor
skill (raw score of total timed repetitive movements from the
PANESS) and praxis performance (total praxis errors) with
ADOS-G measures of the social, communicative, behavioral
deficits that define autism. In addition, this study examined:
(1) the correlation between component scores from the
praxis examination (i.e. total errors on GTC, GTI, and GTU
sections) and total ADOS-G scores which provided a means
of addressing the hypothesis that ADOS-G scores in children
with autism would correlate not only with performance of
GTI, but also with GTC and GTU; and (2) a hierarchical
regression examining the correlation between praxis total
errors score and total ADOS-G score, after accounting for
basic motor skill performance from the PANESS; this provid-
ed a means of addressing the hypothesis that the correlation
between worse praxis performance and higher ADOS-G
scores in children with ASD would remain significant after
accounting for deficits in basic motor skill performance.
Results
PRELIMINARY ANALYSES
Comparisons of the ASD and TD groups revealed no significant
group difference in age (F[1,92]=0.10, p=0.75) or sex distribu-
tion (2=0.004; p=0.7). Mean FSIQ was significantly lower for
the autism group (F[1,90]=23.79, p<0.001). For children
given the WISC-IV (22 TD controls, 28 ASD), there was no sig-
nificant difference between groups on the Verbal Compre-
hension Index (F(1,48)=1.82, p=0.184) or Perceptual
Reasoning Index (F(1,48)=0.26, p=0.613). In contrast, for the
children given the WISC-III (24 TD controls, 17 ASD), controls
had significantly higher Verbal IQ (VIQ; (F(1,39)=26.38,
p<0.001) and Performance IQ (PIQ; F(1,39)=14.68, p<0.001)
than the ASD group. The discrepancy between findings from
the WISC-III and WISC-IV may be due to VIQ and PIQ, but not
Verbal Comprehension and Perceptual Reasoning Indices,
including subtests that relate to processing speed and working
memory.
0.25 Children with ASD showed significantly poorer perfor-
mance than TD controls on the praxis examination (total
0.30 errors score, F[1,92]=31.06, p<0.001; log transformed errors
0.14
score F[1,92]=39.85, p<0.001) as well as PANESS (total score,
0.39
F[1,82]=52.86, p<0.001), with time to complete repetitive
movements being significantly slower in children with ASD
than TD controls (F[1,86]=13.94, p=0.001; Table I).
Effects of basic motor skills and diagnosis on praxis
performance
Table II shows results of the hierarchical regression analysis.
Age and FSIQ were both significant indicators of praxis per-
formance (log transformed total praxis errors score). After
accounting for age and FSIQ, total timed repetitive move-
ments on the PANESS was a significant predictor of praxis
performance (R2=0.073, p=0.004). Further, after account-
ing for age, FSIQ, and basic motor performance, diagnosis
(ASD vs control) added a significant proportion of unique vari-
ance in prediction of praxis performance (R2=0.128,
p<0.001), with controls having significantly higher mean
praxis scores than children with ASD.
Analysis of association between praxis performance and
ADOS scores
Linear regression was used to evaluate the association between
praxis performance and ADOS scores that reflect social, com-
municative, and behavioral impairments diagnostic of autism.
Total errors from the praxis examination was a significant pre-
dictor of total ADOS-G score (R2=0.31, p<0.001; Fig. 1).
Further, the number of errors on praxis examination signifi-
cantly predicted each of the ADOS-G component scores: recip-
rocal social interaction (R2=0.13, p=0.026), communication
(R2=0.25, p=0.001), and stereotyped/repetitive behaviors
(R2=0.26, p=0.001). Furthermore, there was a significant cor-
relation between total ADOS-G scores and each of the praxis
subsections: GTC errors (R2=0.28, p<0.001), GTI (R2=0.24,
p=0.002), and GTU (R2=0.35, p<0.001).
A separate hierarchical regression was used in the ASD
group to predict ADOS-G scores from praxis performance (total
errors), after controlling for basic motor skill performance.
Basic motor skill performance was not a significant predictor
of total ADOS-G score (R2=0.07, p=0.12); after accounting
for basic motor skill, praxis performance was still a significant
predictor of total ADOS-G score (R2=0.237, p=0.002).
Discussion
After taking into account age and FSIQ, there was a significant
effect of basic motor skill (total timed repetitive movements on
the PANESS) on praxis performance. However, there remained
a significant effect of diagnosis on praxis after accounting for
motor skill performance, with the ASD group continuing to
show poor praxis performance compared with TD controls.
The findings suggest that dyspraxia in the autism group can-
not be entirely accounted for by basic motor skill deficits.
The observed association between basic motor skill and
praxis performance is not surprising, as the praxis examina-
tion is, in part, assessing limb motor function. This may indi-
Table II: Hierarchical regression predicting log (total praxis errors)a
Predictor entered 95% confidence interval for
Age 0.048
FSIQ 0.007
PANESS repetitive total 0.013
Group (autism vs control) 0.234
aAnalyses indicate hierarchical regression with age entered first, followed by FSIQ, PANESS total repetitive timed score, and group. FSIQ, Full-
scale IQ; PANESS, Physical and Neurological Assessment of Subtle Signs.
cate that impaired performance of skilled gestures in autism
is, at least partially, due to problems with basic motor coordi-
nation. Alternatively, the association may be an indication
that impairments in basic motor skills and impaired praxis
performance in autism are epiphenomena, or secondary
symptoms, i.e. that a common neurological abnormality con-
tributes to both basic motor skill deficits and to the dyspraxia
observed in autism.
Findings also showed that performance on praxis examina-
tion is worse than what would be expected from poor basic
motor skill performance, or clumsiness, alone, and suggest
that additional factors are contributing to dyspraxia observed
in autism.
Findings from studies of acquired dyspraxia suggest that a
neural network within the left hemisphere is important for
control of goal-directed skilled movements, including pari-
etal regions (supramarginal and angular gyri) important for
storage of learned timespace movement representations
and premotor regions (in particular the supplementary
motor area) important for transcoding of praxic representa-
tions into motor programs.12,25 Movement formulas are
stored in the inferior parietal lobe in three-dimensional
form. These formulas must be translated into motor plans
for movement, and either the storage or the translation
could be a contributing factor to the dyspraxia seen in
autism. Autism is associated with abnormalities in percep-
tion of biological motion26 and, more generally, in dorsal
stream processing involved in detection and analysis of
motion.27 It is possible that these deficits in visual processing
may also contribute to impaired development of complex
skills and gestures assessed on praxis examination. Examining
correlations between these measures of visual impairment
and praxis performance is an important area of future study.
However, there are limits to how well models suggested
by studies of adults with acquired lesions can inform us about
the brain basis of dyspraxia in autism. Autism is a develop-
mental disorder; therefore, dyspraxia in autism is unlikely to
be due to loss of already acquired skills, but rather it is an
impairment in acquisition (learning) of motor sequences
involved in performance of skilled gestures. Impaired motor
sequence learning has been reported in children with
autism,28 and it is important to consider abnormalities in
neural systems involved in motor learning.
Motor sequence learning relies on a broad neural network
principally involving connections between frontal and parietal
cortices and subcortical regions: the basal ganglia and cerebel-
lum.29,30 Abnormalities in these regions have been reported in
several imaging studies of ASD.31 Particularly compelling is that
decreased Purkinje cell count in the cerebellum is the most
R2 F p
0.088 to 0.008 0.062 5.62 0.020
0.010 to 0.004 0.158 16.95 <0.001
0.004 to 0.023 0.073 8.60 0.004
0.125 to 0.344 0.128 18.04 <0.001
Dyspraxia Associations in Autism M A Dziuk et al. 737
consistent finding in post-mortem studies of autism,32 which
prompts speculation that abnormalities in the cerebellum and/or
connections between the cerebellum and frontal/parietal regions
may contribute to impaired development of motor skills.
ASSOCIATION BETWEEN PRAXIS PERFORMANCE AND DIAGNOSTIC
FEATURES OF AUTISM
In children with autism, the level of impairment on praxis
examination significantly correlated with total ADOS-G score.
This correlation remained significant after controlling for basic
motor skill performance, providing further evidence that the
association between dyspraxia and autism cannot be entirely
accounted for by basic motor skill deficits.
Development of social communication and interaction
involves learning complex motor sequences that parallel those
on the praxis examination. The finding of a significant associa-
tion between praxis performance and ADOS-G scores suggests
that the impaired performance of skilled gestures (including
social gestures) may contribute to impaired social interaction
and communication in autism. More likely, dyspraxia and
impaired social interaction/communication seen in autism
may be epiphenomena, sharing a common neurological basis.
Each individual section of the praxis examination (GTC,
GTI, GTU) was significantly associated with total ADOS-G score.
This finding provides further evidence that performance of
complex skilled gestures in children with autism is not limited
only to imitation,9 but rather that autism is associated with a
generalized impairment in performance of skilled complex
movements, consistent with dyspraxia.
Finally, each ADOS-G subsection (communication, social
interactions, and stereotyped/repetitive behaviors) was sig-
nificantly correlated with performance on the praxis exami-
nation, indicating that impaired performance of skilled
gestures is broadly associated with the social, communicative,
and repetitive behavioral impairments that define autism.
This finding suggests that dyspraxia may be a core feature of
autism, or that it is a marker of the neurological deficits that
underlie the broad features of the disorder.
LIMITATIONS AND FUTURE DIRECTIONS
The praxis examination was adapted for children from an adult
100
80
Total praxis errors
60
40
20
0 3. Mostofsky SH, Burgess MP, Gidley Larson JC. Increased motor
5 10 15 20
ADOS-G score (total)
Figure 1: Plot showing a significant correlation (R2=0.311,
p<0.001) between performance on praxis examination
(total errors scores) and total scores from Autism
Diagnostic Observation ScheduleGeneric (ADOS-G) in 39
children with autism.
738 Developmental Medicine & Child Neurology 2007, 49: 734739
apraxia battery.21 There is no standard evaluation of praxis in
children, and an examination tailored for children would be
helpful in the future. Such an examination could also be a valu-
able tool for clinicians. The age range of the children was limited
to 8 to 14 years. This allowed for a more homogenous group to
evaluate. However, a broader age range would evaluate dysprax-
ia more comprehensively throughout childhood and adoles-
cence. Studies that may quantify dyspraxia in younger children
may also be helpful in detecting deviations from the norm earli-
er in a clinical setting. The amount of experience the children
had in performing these complex gestures is difficult to ascer-
tain. Their individual experiences contribute to their ability to
perform meaningful gestures (e.g. using a hammer or waving
goodbye), which could increase their number of correctly per-
formed gestures. Nevertheless, difficulty with imitation is the
most commonly reported area of dyspraxia in children with
autism, and imitation of gestures is the least reliant on prior
knowledge of and experience with the gestures.
Conclusion
Results indicate that a generalized praxis deficit in high-func-
tioning children with ASD cannot be accounted for by imp-
airments in basic motor skills alone. This suggests that
abnormalities in neural systems, outside of those that would
account for basic motor skill deficits, contribute to dyspraxia in
children with ASD. These abnormalities in neural systems
would likely involve regions (or connections between regions)
within neural networks critical for acquisition of movement
patterns necessary for development of skilled tool use and
social/communicative gestures. This may include posterior
parietal regions important for acquisition and storage of spatial
representations of movement or subcortical (cerebellum,
basal ganglia) regions central to motor learning. Results also
show that the impairment in praxis performance, but not basic
motor skill performance, in children with ASD is broadly asso-
ciated with ADOS-G measures of the social, communicative,
and repetitive behavioral impairments that define ASD. This
suggests that dyspraxia may be a core feature of autism, or a
marker of the neurologic deficits that underlie the disorder.
Accepted for publication 22nd May 2007.
Acknowledgements
This research was supported by NIH Grants R01 NS048527 (SHM),
K02 NS044850 (SHM), M01 RR00052 (Johns Hopkins General Clinical
Research Center), P30 HD 24061 (Mental Retardation and
Developmental Disabilities Research Center), and a grant from the
National Alliance for Autism Research/Autism Speaks (SHM).
References
1. Denckla MB. (1985) Revised Neurological Examination for Subtle
Signs. Psychopharmacol Bull 21: 773800.
2. Gidley Larson JC, Mostofsky SH. (2006) Motor deficits in autism.
In: Tuchman R, Rapin I, editors. Autism: A Neurological Disorder
of Early Brain Development. Internationa Review of Child
Neurology Series. London: Mac Keith Press. p 231247.
25 30 cortex white matter volume predicts motor dysfunction in
autism. Brain 130: 21172122.
4. Teitelbaum P, Teitelbaum O, Nye J, Fryman J, Maurer RG. (1998)
Movement analysis in infancy may be useful for early diagnosis of
autism. Proc Natl Acad Sci USA 95: 1398213987.
5. Hallett M, Lebiedowska MK, Thomas SL, Stanhope SJ, Denckla MB,
Rumsey J. (1993) Locomotion of autistic adults. Arch Neurol
50: 13041308.
6. Jansiewicz EM, Goldberg MC, Newschaffer CJ, Denckla MB,
 Landa RJ, Mostofsky SH. (2006) Motor signs distinguish children
with high functioning autism and Aspergers syndrome from
controls. J Autism Dev Disord 36: 613621.
7. Noterdaeme M, Mildenberger K, Minow F, Amorosa H. (2002)
Evaluation of neuromotor deficits in children with autism and
children with a specific speech and language disorder. Eur Child
Adolesc Psychiatry 11: 219225.
8. Rogers SJ, Bennetto L, McEvoy R, Pennington BF. (1996)
Imitation and pantomime in high-functioning adolescents with
autism spectrum disorders. Child Dev 67: 20602073.
9. Mostofsky SH, Dubey P, Jerath VK, Jansiewicz EM, Goldberg MC,
Denckla MB. (2006) Developmental dyspraxia is not limited to
imitation in children with autism spectrum disorders. J Int
Neuropsychol Soc 12: 314326.
10. Rogers SJ, Ozonoff S, Maslin-Cole C. (1991) A comparative study
of attachment behavior in young children with autism or other
psychiatric disorders. J Am Acad Child Adolesc Psychiatry
30: 483488.
11. Williams JH, Whiten A, Singh T. (2004) A systematic review of
action imitation in autistic spectrum disorder. J Autism Dev
Disord 34: 285299.
12. Heilman KM, Gonzalez Rothi LJ. (2003) Apraxia. In: Heilman KM,
Valenstein E, editors. Clinical Neuropsychology. 4th edn. New
York: Oxford University Press. p 215235.
13. Dewey D. (1993) Error analysis of limb and orofacial praxis
in children with developmental motor deficits. Brain Cogn
23: 203221.
14. Denkla MB. (1973) Development of speed in repetitive and
successive finger-movements in normal children. Dev Med Child
Neurol 15: 635645.
15. Lord C, Risi S, Lambrecht L, Cook EH, Leventhal BL, DiLavore PC,
Pickles A, Rutter M. (2000) The autism diagnostic observation
schedule-generic: a standard measure of social and
communication deficits associated with the spectrum of autism.
J Autism Dev Disord 30: 205223.
16. American Psychiatric Association. (1994) Diagnostic and
Statistical Manual of Mental Disorders. 4th edn. Washington
DC, American Psychiatric Association.
17. 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 developmental
disorders. J Autism Dev Disord 24: 659685.
18. Weschler D. (1991) Weschler Intelligence Scale for Children. 3rd
edn. San Antonio, TX: The Psychological Corporation.
19. Weschler D. (2003) Weschler Intelligence Scale for Children. 4th
edn. San Antonio, TX: The Psychological Corporation.
20. Williams PE, Weiss LG, Rolfhus EL. (2003) WISC-IV Technical
Report #2. Psychometric properties. San Antonio, TX:
The Psychological Corporation.
21. Rothi GLJ, Raymer AM, Ochipa C, Maher LM, Greenwald ML,
Heilman KM. (2003) Florida Apraxia Battery-Revised.
22. Holden EW, Tarnowski KJ, Prinz RJ. (1982) Reliability of
neurological soft signs in children: reevaluation of the PANESS.
J Abnorm Child Psychol 10: 163172.
23. Vitiello B, Ricciuti AJ, Stoff DM, Behar D, Denckla MB. (1989)
Reliability of subtle (soft) neurological signs in children. J Am
Acad Child Adolesc Psychiatry 28: 749753.
24. Ghaziuddin M, Butler E. (1998) Clumsiness in autism and
Asperger syndrome: a further report. J Intellect Disabil Res
42(Pt 1): 4348.
25. Geschwind N. (1965) Disconnection syndromes in animals and
man. Brain 88: 237294, 585644.
26. Blake R, Turner LM, Smoski MJ, Pozdol SL, Stone WL. (2003)
Visual recognition of biological motion is impaired in children
with autism. Psychological Science 14: 151157.
27. Spencer J, OBrien J, Riggs K, Braddick O, Atkinson J, Wattam-
Bell J. (2000) Motion processing in autism: evidence for a dorsal
stream deficiency. Neuroreport 11: 27652767.
28. Mostofsky SH, Goldberg MC, Landa RJ, Denckla MB. (2000)
Evidence for a deficit in procedural learning in children and
adolescents with autism: implications for cerebellar
contribution. J Int Neuropsychol Soc 6: 752759.
29. Doyon J, Penhune V, Ungerleider LG. (2003) Distinct
contribution of the cortico-striatal and cortico-cerebellar
systems to motor skill learning. Neuropsychologia 41: 252262.
30. Ito M. (2000) Mechanisms of motor learning in the cerebellum.
Brain Res 886: 237245.
31. Penn HE. (2006) Neurobiological correlates of autism: a review
of recent research. Child Neuropsychol 12: 5779.
32. Bailey A, Luthert P, Dean A, Harding B, Janota I, Montgomery M,
Rutter M, Lantos P. (1998) A clinicopathological study of autism.
Brain 121(Pt 5): 889905.
List of abbreviations
ADOS-G Autism Diagnostic Observation ScheduleG
ASD Autism spectrum disorder
BPT Body-part-for-tool
GTC Gestures to command
GTI Gestures to imitation
GTU Gestures with tool use
PANESS Physical and Neurological Assessment of Subtle Signs

Become Tomorrows Pediatric Rehabilitation Researcher!

Join the Rehabilitation Medicine Scientist Training Program
Where is the evidence base for rehabilitation treatments?
How can we support the value of rehabilitation with payors and policy-makers?

Rehabilitation research is the key, and well-trained rehabilitation researchers are needed for the job! We offer NIH-
funded research training fellowships at competitive salaries to selected individuals to study with a nationally
prominent mentor of their choice for up to 3 years. The goal of the program is to train a cohort of physiatric
researchers focusing on both adult and pediatric rehabilitation topics who can compete successfully for NIH
and other research funds, and who can contribute original research to the advancement of the field.
PGY2 and PGY3 residents are invited to attend a research training workshop held at the annual meeting of the
Association of Academic Physiatrists (AAP), to help them prepare the fellowship application at the completion of
their residency. Physiatrists up to 5 years post-training are eligible to apply as well.
Applications for the research training workshop are due December 1 2007.
Applications for the NIH-funded fellowships are due September 1 of each year.
For more information visit the AAP website, www.physiatry.org, or contact the RMSTP Program
Coordinator, Mary Czerniak, at meczerni@einstein.edu or +001 (215) 456-5965

Dyspraxia Associations in Autism M A Dziuk et al. 739