Underdevelopment of the postural control

system in autism
Nancy J. Minshew, MD; KiBum Sung, MD, PhD; Bobby L. Jones, PhD; and Joseph M. Furman, MD, PhD

Abstract—Objective: To determine if abnormalities exist in postural control in autism and if they are related to age.
Methods: Dynamic posturography was performed in 79 autistic individuals without mental retardation and 61 healthy
volunteers between ages 5 and 52 years. Both the sensory organization and the movement coordination portions of the test
were performed. Results: The autistic subjects had reduced postural stability (p ! 0.002). Examination of age effects
revealed that the development of postural stability was delayed in the autistic subjects (p " 0.001) and failed to achieve
adult levels (p ! 0.004). Postural stability was reduced under all conditions but was clinically significant only when
somatosensory input was disrupted alone or in combination with other sensory challenges (mean reduction in stability of
2.6 # 1.0 for the first three conditions without somatosensory disruption vs 6.7 # 2.7 for the last three conditions with
somatosensory disruption), indicating problems with multimodality sensory integration. Conclusions: The evidence from
this and studies of the motor system suggests more general involvement of neural circuitry beyond the neural systems for
social behavior, communication, and reasoning, all of which share a high demand on neural integration of information.
NEUROLOGY 2004;63:2056 –2061

Autism is a disorder defined by impairments in so- mentally retarded subjects, showed paradoxically
cial reciprocity, verbal and nonverbal language, and better stability when vision was occluded or somato-
imaginative play and by restricted and repetitive be- sensory input restricted. The postural instability in
havior. There are also associated symptoms that are the autistic subjects was comparable to that of the
not part of this diagnostic constellation but nonethe- mentally retarded nonautistic subjects; it was only
less appear to be neurologically and clinically impor- their paradoxical response to somatosensory and vi-
tant elements of this syndrome. Abnormalities of sual stimuli that was distinguishing. The second
motor coordination, posture, and gait are among quantitative study involved 5 autistic children with
these.1 Numerous studies have documented clumsi- mental retardation, 4 to 7 years old, and 12 normal
ness and gross and fine motor apraxia, which are control subjects of the same age.8 The mentally re-
now considered integral aspects of autism.2-4 Only a tarded autistic children had impaired postural sta-
few studies of postural control have been done. Stud- bility compared with normal children under eyes
ies using clinical tests reported abnormal postural closed and static visual conditions but were more
balance in children with autism.2,5,6 Quantitative stable to visually perceived environmental motion
methods were used in three studies, but in two the than control subjects. It was not clear whether this
results appeared to be largely related to mental re- was due to a general lack of visual attention or to
tardation rather than to autism, and the remaining impaired motion processing. The third quantitative
findings in the mentally retarded subjects were the study involved eight children with ASD ages 5 to 12
opposite of those of typical children and of those in years who had a receptive language level of at least 4
the one small study of high-functioning children with years and thus were not mentally retarded or only
autism spectrum disorder (ASD).7-9 mildly so.9 The ASD children performed as well as
The first quantitative study of postural control in- control subjects with eyes open but had significantly
volved 91 mentally retarded autistic children, ages 6 more difficulty maintaining an upright balance when
to 20 years, compared with 166 normal 4- to 11-year- vision was occluded and when somatosensory cues
old children and 18 mentally retarded children ages were modified. The paradoxical improvement with
7 to 16 years.7 The autistic and mentally retarded sensory challenge reported in the earliest study7 not
nonautistic subjects had significantly lower postural only was not observed in this study, but the subjects
stability than the normal control children, perform- demonstrated deficits in sensory integration.
ing at the level of preschool children even as adoles- Many regions of the brain have been proposed to
cents. The autistic subjects, but not the nonautistic be involved in the pathophysiology of autism. His-

From the Departments of Psychiatry (Dr. Minshew), Neurology (Drs. Minshew and Furman), and Otolaryngology (Dr. Furman), University of Pittsburgh
School of Medicine, and H. John Heinz III School of Public Policy and Management (Dr. Jones), Carnegie Mellon University, Pittsburgh, PA; and Department
of Neurology (Dr. Sung), Soonchunhyang University Hospital, Bucheon, Korea.
Supported by a National Institute of Neurological Disorders and Stroke grant (NS 33355) and by the National Institute of Child Health and Human
Development Collaborative Program of Excellence in Autism (HD U19 HD 35469) (N.J.M.).
Received May 1, 2002. Accepted in final form July 22, 2004.
Address correspondence and reprint requests to Dr. N.J. Minshew, Department of Psychiatry, University of Pittsburgh School of Medicine, 3811 O’Hara St.,
Webster Hall, Suite 300, Pittsburgh, PA 15213; e-mail: [email protected]

2056 Copyright © 2004 by AAN Enterprises, Inc.

topathologic studies have revealed abnormalities of Table 1 Sensory conditions
the cerebellum, raising the possibility of motor dys- Term Condition
function as the cause of postural instability in au-
tism. However, emerging evidence suggests it is the Sensory condition 1 Normal vision and normal support
development of the connectivity between brain struc- surface
tures to support integrative functions that is key.10 Sensory condition 2 Eyes closed with normal support surface
We evaluated a large sample of non–mentally re- Sensory condition 3 Sway-referenced visual input with normal
tarded individuals with autism across the age span support surface
to determine if there were impairments in postural Sensory condition 4 Normal vision and sway-referenced
stability and if there were age effects. The advantage support surface
of studying nonretarded individuals with autism is Sensory condition 5 Eyes closed and sway-referenced support
that they have all the defining features of autism surface
without the nonspecific confounding features of men- Sensory condition 6 Sway-referenced visual and support
tal retardation. The methods used in this study were surface
more challenging than prior studies, in that none of
the prior studies used sway referencing for vision or
challenging perturbations of the floor. sensory conditions 1, 2, and 3 and two or three trials for sensory
conditions 4, 5, and 6 as described in table 1. For each 15-second
Methods. Subjects. The subjects for this study consisted of 79 trial, an equilibrium score was derived by relating the anterior–
high-functioning autistic individuals and 61 healthy control sub- posterior peak-to-peak sway to a maximal sway of 12°. A score of 100
jects between ages 5 and 52 years. The autistic subjects were indicated no sway, and 0 indicated a fall. A weighted average of the
recruited as consecutive referrals meeting the eligibility criteria individual trial scores was then used to develop a derived equilib-
for this study. Control subjects were recruited from community rium measure (see Data Analysis). The ability of subjects to adapt to
volunteers responding to advertisements. All subjects had Full- repeated platform rotations was performed using both toes-up and
Scale and Verbal IQ scores above 70. Subjects were group toes-down rotations. Rotations lasted for 400 milliseconds with am-
matched with equal variances across groups. The study was ap- plitude of 8°. The response to each rotation was determined by mea-
proved by the University of Pittsburgh Institutional Review suring the rate of change of the force exerted on the platform. An
Board. Procedures were fully explained to the subjects or their adaptation ratio was obtained by dividing the magnitude of the re-
guardians, and written informed consent was obtained. sponse to the fifth rotation by the magnitude of the second rotation.
Potential autistic subjects were excluded if found to have an All posturography was performed on the same day.
associated neurologic, genetic, infectious, or metabolic disorder, a Data analysis. The data analysis was geared toward address-
seizure disorder, or they were taking medication known to affect ing the hypothesis that autistic subjects will appear developmen-
the measurements under study. The diagnosis of autism was es- tally delayed in their postural physiology relative to healthy
tablished through expert clinical evaluation, the Autism Diagnos- control subjects. The best scores from the 15-second trials for each
tic Interview (ADI), and the Autism Diagnostic Observation subject were used in the analysis. This was necessary as many
Schedule (ADOS).11-14 Diagnostic accuracy was documented trials were falls. Variability within successful trials was small.
through rescoring of 20% of the audiotaped ADIs and videotaped The autistic individuals and control subjects were initially com-
ADOSs by a developer of these instruments or a member of her pared using two-sample t-tests on their performance on the equi-
research group. librium tests for sensory conditions 1 through 6, the principal
Potential control subjects were excluded if they had evidence of components analysis (PCA)– derived equilibrium measure (see be-
birth or developmental abnormalities, acquired brain injury, a low), the adaptation ratio: toes-up, and the adaptation ratio:
learning or language disability, a current or past history of psychi- toes-down.
atric or neurologic disorder, a medical disorder with implications Multivariate analysis of covariance (MANCOVA) was used to
for the CNS or requiring regular medication use, or a family investigate potential gender, autism, age, and Full-Scale IQ ef-
history in first-degree relatives of developmental cognitive disor- fects on the performance measures jointly. Last, a PCA-derived
der, learning disability, mood disorder, anxiety disorder, or autism equilibrium measure was created using the first principal compo-
nent from the PCA. The goal of this analysis was to come to a
in first-, second-, or third-degree relatives.
one-dimensional description of the postural control measures, fa-
Control subjects were recruited as individual matches for the
cilitating understanding of the relationship of the performance
autistic subjects but compared on a group basis at analysis be-
measures to autism status and subject age.
cause fewer control subjects were identified than autistic subjects
The PCA calculations were performed using the function prin-
using this recruiting procedure. The age difference for an autism–
comp in the free statistical package R (Foundation for Statistical
control pair was no more than 6 months for subjects age 17 or
Computing, Vienna, Austria). After removing the mean value of
younger and no more than 12 months for subjects older than age
the equilibrium scores, the data form a p $ n matrix X, where n is
17. Wechsler Full-Scale IQ scores for a pair could differ by no
the number of subjects and p is the number of scores (6). PCA
more than 5 points. Socioeconomic status of the family of origin consists of finding the eigenvalues and eigenvectors of the (p $ p)
was matched at a group level.15 covariance matrix of the data matrix X. The eigenvalues yield the
Posturography. Dynamic posturography was performed using percentage of variance explained by the corresponding eigenvec-
EquiTest (NeuroCom International, Clackamas, OR). Both the tor. The first principal component is the result of projecting the
“sensory organization” and the “movement coordination” portions mean centered data observation onto the first eigenvector. This
of the test were performed. Sensory organization testing consisted result of this projection of the original variables yields a univari-
of recording the position of the center of force during 15-second ate weighted measure that will maximize the resultant explained
trials wherein the subject stood on either a fixed platform or a variability in the equilibrium scores. In our case, about 64% of the
“sway-referenced” platform that rotated to null the angle between variability in the six equilibrium scores was explained by the first
the foot and the lower leg. Subjects had eyes closed or viewed a principal component, our weighted equilibrium measure. As the
fixed visual surround or a “sway-referenced” visual surround. scores of the pooling of autistic subjects and healthy control sub-
There were six conditions: condition 1 ! fixed platform, eyes open jects were centered to have a mean zero, both negative and posi-
with fixed visual surround; condition 2 ! fixed platform, eyes closed; tive first component scores were obtained.
condition 3 ! fixed platform, eyes open with sway-referenced visual
surround; condition 4 ! sway-referenced platform, eyes open with
fixed visual surround; condition 5 ! sway-referenced platform, eyes Results. Means and SD for Verbal, Performance, and
closed; condition 6 ! sway-referenced platform, eyes open with sway- Full-Scale IQ, age, and gender distribution for the autistic
referenced visual surround. Subjects performed one or two trials for individuals and control subjects appear in table 2. Verbal
December (1 of 2) 2004 NEUROLOGY 63 2057
Table 2 Demographic information for autistic individuals and isons. The PCA measure was different in the two groups
control subjects (p ! 0.005).
As the equilibrium scores are correlated within the sub-
Autistic Control
individuals, subjects,
jects, a multivariate analysis, MANCOVA, was performed
Characteristic n ! 79 n ! 61 using the equilibrium scores for sensory conditions 1
through 6, the adaptation ratio: toes-up, and the adapta-
Mean age, y 17.0 # 10.4 16.7 # 10.5 tion ratio: toes-down as dependent measures. The multi-
Mean Verbal IQ 96.0 # 17.2 100.0 # 14.3 variate analysis revealed differences in the eight
dependent measures by autism status (p ! 0.02), age (p "
Performance IQ 94.2 # 13.5 98.4 # 12.6
0.001), age squared (p " 0.001), Full-Scale IQ (p ! 0.001),
Full-Scale IQ 94.8 # 15.0 99.2 # 13.9 and gender (p ! 0.03). The effects of age and autism status
Male/female (% male) 71:8 (90) 55:6 (90) were present on all of the eight dependent measures. How-
ever, the effect of gender was seen only in the adaptation
Values are means # SD except where indicated. ratio: toes-down measure. The effect of Full-Scale IQ was
seen only on the adaptation ratio: toes-up and toes-down,
IQs ranged from 70 to 148 for autistic subjects and from 71 sensory condition 1, and sensory condition 6 measures.
to 136 for control subjects. Performance IQs ranged from Having found a relationship between the multivariate
70 to 126 for autistic subjects and from 69 to 129 for response, age, Full-Scale IQ, gender, and autism status,
control subjects. Full-Scale IQs ranged from 70 to 136 for we further investigated the relationship by performing a
autistic subjects and from 72 to 136 for control subjects. regression analysis of the PCA-derived equilibrium mea-
Ages ranged from 5 years 4 months to 51 years 10 months sure. The regression revealed differences between the au-
for autistic subjects (median age 15 years 2 months) and tistic subjects and control subjects (p ! 0.002) as well as
from 5 years 10 months to 52 years 3 months (median age an age effect (p " 0.001) on the PCA measure. The PCA
14 years 9 months) for control subjects. Mean equilibrium measure was not affected by gender (p ! 0.487) or Full-
scores and SD for sensory conditions 1 through 6, PCA- Scale IQ (p ! 0.099). The results of the multivariate anal-
derived equilibrium measure, adaptation ratio: toes-up, ysis (MANCOVA) and univariate analysis (regression)
and adaptation ratio: toes-down for the autistic individuals together give evidence that the relationship is nonlinear in
and control subjects appear in table 3. age, depends on autism status, and (from the multivariate
Preliminary two-sample univariate tests on the original analysis) also depends on Full-Scale IQ and gender.
measures (sensory conditions 1 to 6 and the two adapta- To illustrate the age effects and differences between
tion ratios) did not reveal significant differences between autistic subjects and healthy control subjects, scatterplots
the autistic individuals and control subjects when applying vs subject age of the adaptation ratio: toes-up, adaptation
a Bonferroni correction to account for the multiple compar- ratio: toes-down, and equilibrium scores for sensory condi-
tions 1 through 6 with locally smoothed curves (function
supsmu in the statistical package R) appear in figures 1
Table 3 Posturography information for autistic individuals and and 2. A scatterplot of the PCA-derived equilibrium mea-
control subjects
sure vs subject age appears in figure 3. Only the differ-
Autistic Control ences in sensory conditions 4 through 6 appeared clinically
individuals, subjects, significant, with an average difference in sway measure of
Characteristic n ! 79 n ! 61 p Value* %5. Thus, overall, the autistic individuals swayed more
than control subjects but particularly in those conditions
Equilibrium score: 88.2 # 7.4 90.9 # 5.0 0.011 that included a sway-referenced support surface, namely,
condition 1
sensory conditions 4, 5, and 6 (regressions on the individ-
Equilibrium score: 84.4 # 8.0 87.1 # 5.9 0.024 ual measures yielded a mean reduction in stability of 2.6 #
condition 2 0.9 for sensory conditions 1, 2, and 3 vs 6.7 # 2.7 for
Equilibrium score: 84.3 # 9.4 86.9 # 6.7 0.056 sensory conditions 4, 5, and 6, when the sway-referenced
condition 3 support surface was included). In figure 2, it can be seen
Equilibrium score: 72.0 # 15.6 77.4 # 14.9 0.039 that the difficulty emerges in the last three conditions (4 to
condition 4 6) as platform motion is introduced. It is apparent from
these three panels that there is a delay in development of
Equilibrium score: 51.4 # 20.3 59.4 # 15.7 0.010
condition 5 postural control in autism. They did not begin to improve
until about age 12 and failed to reach adult levels. In
Equilibrium score: 51.8 # 21.6 59.1 # 18.0 0.030
contrast, postural control in the control subjects improves
condition 6
steadily from ages 5 to 15 when it plateaus.
PCA-derived equilibrium &5.6 # 28.8 7.2 # 24.0 0.005
measure
Adaptation ratio: toes-up 0.97 # 0.47 0.90 # 0.21 0.266
Discussion. This study evaluated postural stabil-
ity in 79 non–mentally retarded individuals with au-
Adaptation ratio: toes- 0.91 # 0.19 0.93 # 0.18 0.445
tism between ages 5 and 52 and 61 healthy control
down
subjects. The autistic subjects were found to have
Values are means # SD. reduced postural stability, especially for the condi-
tions in which somatosensory input was disrupted,
* Two sided.
which is seen in some patients who are unable to use
PCA ! principal components analysis. reliable visual or vestibular information to maintain
2058 NEUROLOGY 63 December (1 of 2) 2004
 Figure 2. Measures for autistic subjects (circles) and con-
trol subjects (crosses) and locally smoothed curves (solid
line for autistic subjects, broken line for control subjects).
Figure 1. Measures for autistic subjects (circles) and con- R2 for fits: 0.198 (sensory condition 3 [SC3]), 0.164 (SC4),
trol subjects (crosses) and locally smoothed curves (solid 0.175 (SC5), and 0.170 (SC6). In the last three panels
line for autistic subjects, broken line for control subjects). (SC4 to SC6), the difficulty emerges as platform motion is
R2 for fits: 0.112 (adaptation ratio: toes-up), 0.158 (adap- introduced. These panels demonstrate delayed develop-
tation ratio: toes-down), 0.323 (sensory condition 1 [SC1]), ment and a failure of the autism group to achieve adult
and 0.247 (SC2). levels.

balance.16 This is indicative of a problem with multi- Another finding of this study is the significant
modality sensory integration. Examination of age ef- effect of age on postural control in autism. Postural
fects revealed that the development of postural control did not begin to improve in the autistic sub-
stability was delayed in the autistic subjects and did jects until age 12 and never achieved adult levels,
not achieve adult levels.
Our findings of reduced postural stability for con-
ditions of a stable floor and normal visual input are
consistent with two previous studies: one of 91 men-
tally retarded autistic individuals age 6 to 20 and
another of 5 mentally retarded autistic children age
4 to 7.7,8 Our study, however, demonstrates that the
reduced stability is directly related to autism and not
to mental retardation, which the prior two studies
could not do. Our findings of reduced postural stabil-
ity with occlusion of vision and reduction in proprio-
ceptive input are consistent with the study of eight
ASD children age 6 to 12.9
The results of the current study do not support the
paradoxical improvement in balance with occlusion
of vision reported in one study.7 The reason for this
difference is not apparent. The individuals in the
prior study were mentally retarded, and it is possible
that they were more prone to environmental distrac-
tions or stimulation and that this contributed to
instability when their eyes were open. This explana-
tion is supported by the absence of the paradoxical Figure 3. Principal components analysis (PCA)– derived
effect and the reduced susceptibility to environmen- equilibrium measure for autistic subjects (circles) and con-
tal motion in the study of ASD children who were trol subjects (crosses) and locally smoothed curves with
higher functioning than the mentally retarded approximate 95% confidence bands (solid line for autistic
children.9 subjects, broken line for control subjects). R2 ! 0.246.
December (1 of 2) 2004 NEUROLOGY 63 2059
whereas in control subjects, it improved steadily ities responsible for autism are not restricted to the
from age 5 years to 15 to 20, when it plateaued at neural systems involved in social, language, and rea-
adult levels. This profile provides evidence of both a soning abilities. The presence of these motor and
delay in development and an underdevelopment of sensory deficits suggests more general involvement
the postural control system in autism. of neural circuitry, perhaps related to disruption of a
No significant differences were found between the general cytoarchitectural feature of brain organiza-
autism and control groups on the adaptation ratios. tion required for the higher levels of integration of
These measurements are considered to be primarily information.4 The motor system is better delineated
dependent on motor control function. The absence of and can be more easily probed than systems that
differences for these measurements suggests that underlie complex cognitive functions. Thus, investi-
the motor control system is not primarily responsible gation of the motor system may provide an opportu-
for the postural instability found in the autistic sub- nity for understanding the more widespread
jects. Thus, although reduction in Purkinje cell num- abnormalities in neural connectivity underlying
ber has been universally observed in autopsy autism.
studies,17,18 it does not appear to have functional con-
sequences with regard to postural control in autism.
Acknowledgment
Consistent with this, the predominant histopatho-
logic changes in autism have been found in the pos- The authors thank the volunteers for their participation.
terolateral regions of the cerebellar hemispheres, not
the vermis.18 References
There was a small but significant effect of IQ on 1. Damasio AR, Mauer RG. A neurological model for childhood autism.
the postural deficits in this study. Within our autis- Arch Neurol 1978;35:777–786.
tic population, lower IQ corresponded to more im- 2. Ghaziuddin M, Butler E. Clumsiness in autism and Asperger syndrome:
a further report. J Intellect Disabil Res 1998;42:43– 48.
pairment in postural control. In the normal 3. Burgoine E, Wing L. Identical triplets with Asperger’s syndrome. Br J
population, postural control does not relate to IQ. In Psychiatry 1983;143:261–265.
4. Minshew NJ, Golstein G, Siegel DJ. Neuropsychologic functioning in
the autism population, IQ, particularly Verbal IQ, is autism: profile of a complex information processing disorder. J Int Neu-
a general index of the severity of autism.19 The rela- ropsychol Soc 1997;3:303–316.
5. Manjiviona J, Prior M. Comparison of Asperger syndrome and high-
tionship between IQ and postural control in this non- functioning autistic children on a test of motor impairment. J Autism
retarded autism group could reflect increasing Dev Disord 1995;25:23–29.
problems with sensory integration with increasing 6. Miyahara M, Tsuji M, Hori M, Nakanishi K, Kageyama H, Sugiyama T.
Brief report: motor incoordination in children with Asperger syndrome
severity of the disorder. There was also a small but and learning disabilities. J Autism Dev Disord 1997;27:595– 603.
significant effect of gender. Gender is not known to 7. Kohen-Raz R, Volkmar FR, Cohen DJ. Postural control in children with
autism. J Autism Dev Disord 1992;22:419 – 432.
influence postural control normally, but in autism, 8. Gepner B, Mestre D, Masson G, de Schonen S. Postural effects of
the female gender is associated with greater autism motion vision in young autistic children. Neuroreport 1995;6:1211–
1214.
severity.19 9. Molloy CA, Dietrich KN, Bhattacharya A. Postural stability in children
The findings of this study do not, therefore, sup- with autism spectrum disorder. J Autism Dev Disord 2003;33:643– 652.
10. Just MA, Cherkassky VL, Keller TA, Minshew NJ. Cortical activation and
port localized cerebellar or vestibular dysfunction as synchronization during sentence comprehension in high-functioning au-
proposed by other studies of postural control in au- tism: evidence of underconnectivity. Brain 2004;127:1811–1821.
tism and some neurobehavioral models.7-9,20,21 In 11. Minshew NJ, Meyer J, Dunn M. Pervasive development disorders: au-
tism and related conditions. In: Segalowitz S, Rapin I, eds. Handbook of
adults, this type of generalized postural dysfunction neuropsychology. 2nd ed., vol. 9, pt II. Elsevier Science B.V., London,
can be seen with Parkinson disease,22,23 progressive 2003:863– 896.
12. Le Couteur A, Rutter M, Lord C, et al. Autism diagnostic interview: a
supranuclear palsy,22 dysequilibrium of aging,24-26 standardized investigator-based instrument. J Autism Dev Disord
and Huntington disease.27 The similarity of the pos- 1989;19:363–387.
13. Lord C, Rutter M, LeCouteur AL. Autism diagnostic interview-revised.
tural findings in autism to those in these other disor- A revised version of a diagnostic interview for caregivers of individuals
ders suggests involvement of basal ganglia, with possible pervasive developmental disorders. J Autism Dev Disord
supplementary motor, and anterior cingulate regions 1994;24:659 – 685.
14. Lord C, Rutter M, Goode S, et al. Autism Diagnostic Observation
to which it is reciprocally connected and subcortical Schedule: a standardized observation of communicative and social be-
connections more generally. This localization is con- havior. J Autism Dev Disord 1989;19:185–212.
15. Hollingshead AB. Four factor index of social status. New Haven: Yale
sistent with the reports in autism of parkinsonian University Department of Sociology, 1975.
faces, the problems in the initiation of motor move- 16. Jacob RG, Furman JM, Durrant JD, Turner SM. Surface dependence: a
balance control strategy in panic disorder with agoraphobia. Psychosom
ments,28 increased volume of the caudate,29,30 and re- Med 1997;59:323–330.
duced volume of the anterior cingulate.31 17. Bailey A, Luthert P, Dean A, et al. A clinicopathological study of au-
tism. Brain 1998;121:889 –905.
The decreased postural stability found in this 18. Kemper TL, Bauman M. Neuropathology of infantile autism. J Neuro-
study is part of a more pervasive impairment in pathol Exp Neurol 1998;57:645– 652.
movement that ranges from fine and gross motor 19. Liss M, Harel B, Fein D, et al. Predictors and correlates of adaptive
functioning in children with developmental disorders. J Autism Dev
apraxia to the planning and execution of skilled mo- Disord 2001;31:219 –230.
tor sequences.4,5,32-35 These impairments all place 20. Ornitz EM. Childhood autism: a review of the clinical and experimental
literature. Calif Med 1973;118:21– 47.
high demands on neural integration.4 In addition, 21. Courchesne E, Yeung-Courchesne BA, Press GA, Hesselink JR, Jerni-
the evidence that deficits in skilled motor sequences gan TL. Hypoplasia of cerebellar vermal lobules VI and VII in autism.
N Engl J Med 1988;318:1349 –1354.
and now sensory integration are integral parts of the 22. Ondo W, Warrior D, Overby A, et al. Computerized posturography
autism syndrome suggests that the neural abnormal- analysis of progressive supranuclear palsy: a case-control comparison

2060 NEUROLOGY 63 December (1 of 2) 2004

 with Parkinson’s disease and healthy controls. Arch Neurol 2000;57: 30. Sears LL, Vest C, Mohamed S, Bailey J, Ranson BJ, Piven J. An MRI
1464 –1469. study of the basal ganglia in autism. Prog Neuropsychopharmacol Biol
23. Waterston JA, Hawken MB, Tanyeri S, Jantti P, Kennard C. Influence Psychiatry 1999;23:613– 624.
of sensory manipulation on postural control in Parkinson’s disease. 31. Haznedar MM, Buchsbaum MS, Wei TC, et al. Limbic circuitry in
J Neurol Neurosurg Psychiatry 1993;56:1276 –1281. patients with autism spectrum disorders studied with positron emission
24. Baloh RW, Kerber KA, Enrietto JA, Jacobson KM. Disequilibrium of tomography and magnetic resonance imaging. Am J Psychiatry 2000;
aging. Neurology 1998;50(suppl 4):A8. 157:1994 –2001.
25. Kerber KA, Enrietto JA, Jacobson KM, Baloh RW. Disequilibrium in 32. Ghaziuddin M, Tsai LY, Ghaziuddin N. Brief report: a reappraisal of
clumsiness as a diagnostic feature of Asperger syndrome. J Autism Dev
older people: a prospective study. Neurology 1998;51:574 –580.
Disord 1992;22:651– 656.
26. Wolfson L, Whipple R, Derby CA, et al. A dynamic posturography study
33. Hughes C, Russell J. Autistic children’s difficulty with mental disen-
of balance in healthy elderly. Neurology 1992;42:2069 –2075. gagement from an object: its implications for theories on autism. Dev
27. Tian J, Herdman SJ, Zee DF, Folstein SE. Postural stability in patients Psychol 1993;29:498 –510.
with Huntington’s disease. Neurology 1992;42:1232–1238. 34. Leary MR, Hill DA. Moving on: autism and movement disturbance.
28. Hughes C. Brief report: planning problems in autism at the level of Ment Retard 1996;34:39 –53.
motor control. J Autism Dev Disord 1996;26:99 –107. 35. Rinehart NJ, Bradshaw JL, Brereton AV, Tonge BJ. Movement prepa-
29. Piven J, Arndt S, Bailey J, Andreasen N. Regional brain enlargement ration in high-functioning autism and Asperger disorder: a serial choice
in autism: a magnetic imaging study. J Am Acad Child Adolesc Psychi- reaction time task involving motor reprogramming. J Autism Dev Dis-
atry 1996;35:530 –536. ord 2001;31:79 – 88.

WWW.NEUROLOGY.ORG OFFERS IMPORTANT INFORMATION

TO PATIENTS AND THEIR FAMILIES
The Neurology Patient Page provides:
— a critical review of ground-breaking discoveries in neurologic research that are written especially for patients and
their families
— up-to-date patient information about many neurologic diseases
— links to additional information resources for neurologic patients.

All Neurology Patient Page articles can be easily downloaded and printed, and may be reproduced to distribute for
educational purposes. Click on the Patient Page icon on the home page (www.neurology.org) for a complete index of
Patient Pages.

December (1 of 2) 2004 NEUROLOGY 63 2061