Neuroscience 307 (2015) 273–280
DIFFERENCES IN THE USE OF VISION AND PROPRIOCEPTION FOR
POSTURAL CONTROL IN AUTISM SPECTRUM DISORDER
S. L. MORRIS, * C. J. FOSTER R. PARSONS,
M. FALKMER, T. FALKMER AND S. M. ROSALIE
Faculty of Health Sciences, Curtin University, GPO
Box U1987, Perth, Western Australia 6845, Australia
Abstract—Background: People with autism spectrum disor-
ders (ASDs) also have poorer fundamental motor skills. The
development of postural control underlies both social and
motor skills. All three elements are facilitated by the active
use of visual information. This study compares how adults
with ASD and typically developed adults (TDAs) respond
to a postural illusion induced using neck vibration. Adults
with ASD unlike the TDA, were not expected to correct the
illusion using vision. Methods: The study used intermittent
(15off, 5on) posterior neck vibration during 200 s of quiet
stance to induce a postural illusion. In TDAs and only in
the absence of vision this protocol induces a forward body
lean. Participants (12 ASD, 20 TDA) undertook four condi-
tions combining vibration and visual occlusion. Results:
As predicted, TDA were only affected by the postural illusion
when vision was occluded (vibration condition: vision
occluded (n = 1) p = 0.0001; vision available (n = 3)
p > 0.2466). Adults with ASD were affected by the postural
illusion regardless of the availability of vision (all conditions
p < 0.0007). Conclusions: Our findings indicated the adults
with ASD did not use visual information to control standing
posture. In light of existing evidence that vision-for-
perception is processed typically in ASD, our findings sup-
port a specific deficit in vision-for-action. These findings
may explain why individuals with ASD experience difficul-
ties with both social and motor skills since both require
vision-for-action. Further research needs to investigate the
division of these visual learning pathways in order to pro-
vide more specific intervention opportunities in ASD.
Ó 2015 IBRO. Published by Elsevier Ltd. All rights reserved.
Key words: postural control, autism spectrum disorder,
vision-for-action, vibration, sensory integration.
INTRODUCTION
Autism spectrum disorders (ASDs) are
neurodevelopmental disorders present from early
childhood which are characterized by atypical
*Corresponding author. Tel: +61-0892669226.
E-mail address: s.morris@curtin.edu.au (S. L. Morris).
Abbreviations: AS, Asperger syndrome; ASDs, autism spectrum
disorders; HFA, high-functioning autism; TDAs, typically developed
adults.
http://dx.doi.org/10.1016/j.neuroscience.2015.08.040
0306-4522/Ó 2015 IBRO. Published by Elsevier Ltd. All rights reserved.
273
development of social interaction and communication
with restricted or repetitive interests and behaviors (Lai
et al., 2014). ASD is commonly used as an comprehen-
sive umbrella term and includes among others the diag-
noses of Autistic disorder, Asperger syndrome (AS)
(American Psychiatric Association, 2000). The term
high-functioning autism (HFA), usually refers to individu-
als within the spectrum with an IQ score within or above
the normative average range (Volker, 2012). ASD (previ-
ously known as pervasive developmental disorder) is one
of the most common childhood disorders with 1 in 68 chil-
dren being diagnosed (Lord and Bishop, 2015). Despite
being predominantly a social disorder, at least 79% of
individuals with ASD are reported as having difficulties
with fundamental motor skills (Ghaziuddin and Butler,
1998; Pan et al., 2009; Fournier et al., 2010a). Funda-
mental motor skills are important in the development of
skills in play, interaction with others, communication and
language (Gernsbacher et al., 2008; Blaesi and Wilson,
2010; Clearfield, 2011). Therefore, the motor difficulty in
those with ASD may contribute to the significant social dif-
ficulty associated with this disorder (Bhat et al., 2011).
The ability to control posture is critical to the typical
development of fundamental motor skills (Miyahara,
2013; Travers et al., 2013). If individuals with ASD expe-
rience difficulties with postural control at key time points
during their development of fundamental motor skills,
such difficulties may be a causative factor in the atypical
development of motor and social skills characteristic of
ASD (Bhat et al., 2011). Studies have shown that children
and adolescents with ASD demonstrate differences in
postural control with increased anteroposterior and medi-
olateral sway when standing still compared to their typi-
cally developing peers (Molloy et al., 2003; Fournier
et al., 2010b; Memari et al., 2013). Evidence suggests
that underdevelopment of postural control in the setting
of ASD continues into adulthood (Kohen-Raz et al.,
1992; Minshew et al., 2004; Travers et al., 2013); how-
ever, the mechanisms for the differences are not known.
Visual information usually dominates other forms of
sensory information in the control of posture (Peterka
and Benolken, 1995; Nolan et al., 2005). Differences in
visual processing have been commonly reported in ASD
(Simmons, 2009). Studies comparing the effect of a visual
illusion of motion on the posture of young children with
and without ASD have shown that both younger children
(Gepner et al., 1995) and older children (Gepner and
Mestre, 2002) with ASD change their posture less in
response to the perception of motion than typically
274 S. L. Morris et al. / Neuroscience 307 (2015) 273–280
developing children. These findings suggest that children
with ASD are less dependent on the perception of motion
to maintain a balanced posture. Interestingly, in the same
study children with AS as opposed to children with autistic
disorder were shown to be more reactive to the perception
of motion than both children with autistic disorder and typ-
ically developing children (Gepner and Mestre, 2002).
This raises the possibility that children with the AS are
more, rather than less, dependent on the perception of
motion to maintain a balanced posture than typically
developing children. Conversely, spatial and motion inte-
gration tests that have compared the abilities of children
with autism and typically developing children to verbally
respond to the perception of motion have shown no differ-
ences (Del Viva et al., 2006).
Although visual information is the dominant source of
sensory information used in the control of posture, it is not
the only source of information. Information from the
somatosensory and vestibular systems also contributes
to the maintenance of an optimal body position (Peterka
and Benolken, 1995; Nolan et al., 2005). Thus, either pri-
mary deficits in somatosensory and/or vestibular informa-
tion (Weimer, 2001), or an inability to integrate visual
information with somatosensory and vestibular informa-
tion (Baranek, 2002; Cascio et al., 2012) could lead to
the difficulties in motor control experienced by individuals
with ASD. Evidence for both primary somatosensory and/
or vestibular deficits and deficits in integration is inconsis-
tent. It has been reported; however, that children with
ASD have difficulties in standing on one leg with their
vision occluded and this has been attributed to a proprio-
ceptive deficit that caused an over-reliance on visual infor-
mation to balance (Weimer, 2001). It has also been
reported that adolescents with ASD do not have a primary
proprioceptive deficit (Fuentes et al., 2011) and in fact are
proprioceptive learners (Haswell et al., 2009). Con-
versely, children with ASD have been shown to experi-
ence greater postural sway than typically developing
children when balance is perturbed by standing on an
unstable surface (i.e., foam) both with and without the
availability of visual information. This greater sway has
been attributed to impaired integration of information from
the visual, somatosensory and vestibular systems (Molloy
et al., 2003). Though informative, in the foregoing study
postural sway was examined under conditions where
visual control of posture was directly inhibited through
blindfolding the participants; whereas, somatosensory
control of posture was inhibited through manipulation of
the environment conditions. That is, balance was per-
turbed via changes in the surface (platform versus foam).
To our knowledge researchers are yet to examine differ-
ences in the integration of visual and somatosensory
information for the control of posture between individuals
with and without ASD in a single study that directly and
independently inhibits sensory input from both the visual
and somatosensory systems. Thus, evidence regarding
differences between individuals with and without ASD in
the integration of visual and somatosensory information
for the control of posture on a stable surface is incom-
plete. Furthermore, previous studies of differences in pos-
tural control between individuals with and without ASD
have focused on children and adolescents. Findings of
studies including adults with ASD are mixed. Evidence
both supports a deficit (Kohen-Raz et al., 1992;
Minshew et al., 2004) and no deficit (Greffou et al.,
2012; Travers et al., 2013) in the use of visual information
for postural control in adults with ASD. Thus, it is unclear
whether the difficulties that children and adolescents with
ASD experience in postural control persist into adulthood,
which has implications with respect to learning and perfor-
mance of vocational skills.
Bove and colleagues (2009) recently pioneered a
method of examining the integration of visual and
somatosensory information for posture control in quiet
standing. This method uses sequential periods of vibra-
tion of posterior neck muscles interspersed with
vibration-free periods to produce a transient propriocep-
tive illusion of movement of the head which is interrupted
as a backward lean of the truck due to constant vestibular
input (Bove et al., 2009). This illusion results in a ‘‘correc-
tive” transient forward movement of the center of pressure
coinciding with periods of vibration. Posture is typically
normalized once the illusion is removed. When partici-
pants were instructed to close their eyes across succes-
sive periods of vibration, thereby denying themselves a
visual reference for the position of the head, the forward
movement of the center of pressure was not only
observed to increase but participants were unable to cor-
rect their posture in the absence of vibration until vision
was restored (Bove et al., 2009). This result highlights
the importance of visual information for typically develop-
ing individuals in setting the postural reference point and
returning to a neutral position after posture is perturbed
(Bove et al., 2009). The method used by Bove et al.
(2009) required the participants to open or close their
eyes in response to a verbal command, which meant that
the availability of visual information was controlled by the
participant and not the experimenter. This limitation can
be overcome by using liquid crystal spectacles that allow
the precise control of the availability of vision across time
(Rosalie and Müller, 2013). These spectacles are particu-
larly useful for controlling vision in individuals with ASD
whose difficulties with communication and social interac-
tion may make it problematic to follow a sequence of com-
mands to open and close their eyes. By combining neck
muscle vibration with visual occlusion, it is therefore pos-
sible to have tight experimental control of both the
somatosensory and visual information available for postu-
ral control in individuals with ASD.
The purpose of this study was thus to determine
whether adults with ASD use and integrate visual and
somatosensory information to control posture during
quiet standing differently from TDAs. Based on existing
evidence (Weimer, 2001; Molloy et al., 2003; Fuentes
et al., 2011) it was hypothesized that: (i) proprioception
is processed similarly in typically developed adult (TDA)
and ASD such that when vision is occluded, both groups
will respond to the postural illusion by leaning forward; (ii)
vision is processed differently in TDA and ASD such that
TDA, but not adults with ASD, will normalize their posture
when vision is restored during the postural illusion; and
(iii) integration of vision and proprioception is different in
 S. L. Morris et al. / Neuroscience 307 (2015) 273–280
adults with ASD and TDA so the groups would respond
differently in conditions when the availability of vision
was varied between periods of postural illusion and no
postural illusion compared to when the availability of
vision was constant across periods of postural illusion
and no postural illusion.
EXPERIMENTAL PROCEDURES
Participants
Twenty TDAs 19–35 years old (Mean 23.4 ± SD 5.1) and
twelve adults with a diagnosis of HFA or Asperger
syndrome (ASD) from 19–40 (Age: Mean 23.6 ± SD
7.9) with a Body Mass Index (BMI) between 20 and 42
participated in this experimental study. Groups were
balanced for sex with two females in each group. The
participants were recruited through advertising on social
media. Inclusion criteria were that the participant was
aged over 18 years of age and had been diagnosed with
HFA or AS. Exclusion criteria were an inability to
communicate effectively, inability to understand
instructions, the presence of co-morbid neurological or
movement disorders other than developmental
coordination disorder or epilepsy, any therapeutic
intervention received within the last 2 weeks and
inability to stand with vision occluded for 2 min safely.
Formal diagnosis of ASD was ascertained from the
participant as diagnosis by authorized and qualified
professionals in independent clinics, hence a re-
evaluation of their diagnosis including excessive re-
testing in order to participate in the current study was
considered as unethical (Falkmer et al., 2013). Partici-
pants with ASD reported their diagnostic process as pedi-
atrician only (3), pediatrician and clinical psychologist (3),
clinical psychologist only (4) and three did not indicate.
Four participants with ASD also reported comorbidities
which included dysgraphia (1), epilepsy (2), ADD or
ADHD (2), restless leg syndrome and/or anxiety (1). Five
patients reported taking medication which included Ritalin
(2), Sifrol (1), anti-anxiety medication not specified (1),
epilepsy medication not specified (1) and/or medication
not specified (2). Each participant provided informed writ-
ten consent prior to participation. This study was com-
pleted according to the guidelines of the Helsinki
declaration and was approved by the Curtin University
Human Research Ethics Committee (Approval No:
PT250/2013).
Experimental design and equipment
The procedure used for this study followed closely the
methodology developed by Bove and colleagues (2009).
The participants’ task was to stand on a portable force
platform (AMTI, Watertown, MA, USA) at a set distance
of 1.5 m from a white wall and look at a black horizontal
line running at a height of 165 cm from the floor. The
sagittal plane of the participant was aligned with the y axis
of the force plate. Force and moment data from the force
plate were collected at 1000 Hz and decimated to a fre-
quency of 100 Hz. Post data collection the trajectory of
the center of foot pressure was calculated and the sagittal
275
plane data (i.e., forward and backward movement on the
plate) was retained for further analysis (CFP). Leaning
forward would be indicated by an increase in the CFP
and leaning backward would be indicated by a reduction
in CFP. The postural illusion was induced using a vibrator
(VB 115, Techno Concept, Cereste, France) fixed bilater-
ally to the dorsal neck muscles at C7 using an elastic
strap around the neck and shoulders in the same manner
as Bove and colleagues (2009). Vision was manipulated
using liquid crystal spectacles that could be changed from
translucent to opaque (PLATO, Model P1, Translucent
Technologies Inc., Toronto, ON, Canada) under the com-
mand of the experimenter. Each of these systems was
controlled via a custom-made computer program, which
allowed the accurate control of each of the trial conditions.
The reasoning behind retaining only the sagittal plane
data for analysis was because the bilateral neck vibration
illusion targeted movement in the anterior posterior direc-
tion (i.e. forward lean). Postural lean in medio-lateral
direction due to bilateral neck vibration was not expected
and has not been reported.
Procedure
The visual occlusion and vibration apparatus was placed
on the participant and a period of familiarization
provided. On commencing the experiment the
participant was asked to stand (wearing shoes) with
both feet in the middle of the force plate in a
comfortable standing position with arms hanging by the
side of the body and foot position was marked by
drawing an outline of the entire foot on paper over the
plate. A researcher was close by for support. Between
conditions the participant stepped off the plate. With
each new condition the participant was directed to stand
in the same position and as still as possible. Four trials
representing a combination of visual occlusion/vibration
experimental conditions were undertaken in a
randomized order. Each trial was 200 s long and
consisted of 10 cycles of; 15 s of no vibration
(henceforth known as rest periods) followed by; 5 s of
vibration ((15 + 5)  10 = 200 s). The four visual
occlusion conditions were: (1) vision available during
both the rest and vibration periods (VAVA), (2) vision
available during rest periods but vision occluded during
vibration (VAVO), (3) vision occluded in the rest period
and available during vibration (VOVA) and (4) vision
occluded throughout the full sequence (VOVO) of which
an example of a TDA trial is presented in Fig. 1. After
each trial the participants dismounted the force plate
and a five-minute rest period was provided between trials.
Statistical analysis
Data for two TD participants in the VAVA conditions were
not collected due to hardware issues. All other data were
complete. For each trial the mean CFP position was
calculated for each of the 20 periods in the 200 s of
standing. CFP is reported in millimeters. This created 10
CFP values representing the rest periods and 10 CFP
values representing the vibration periods. CFP for each
of the 19 subsequent periods was normalized to the
276 S. L. Morris et al. / Neuroscience 307 (2015) 273–280

Fig. 1. Trace of a CFP in the sagittal plane from the typically developed adult during the VO-VO trial where vision was occluded during the entire
trial and vibration occurred in 5-s cycles preceded by 15 s rest over the 200 s of the trial.

CFP position inPthe first rest period of each trial significant interaction was observed between group and
((CFP = CFP ( CFP0–15)/15)) (now called CFP visual occlusion condition F(3, 2355) = 10.11,
position) so that each participant started each trial with p < 0.0001. Post-hoc pairwise comparisons revealed
a position of zero. A positive CFP then indicates a that the adults with ASD group had a significantly
leaning position forward from baseline and a negative greater change in CFP position compared to TDA when
CFP indicates a leaning position backward from baseline. vision was either fully or partially available (VAVA,
A linear mixed model with participant as a random t = 2.64, p < 0.01; VAVO t = 2.99, p < 0.01; VOVA
effect (intercept) was used to investigate the effect of t = 2.69 p < 0.01); whereas, there was no difference in
visual occlusion condition (VOVO, VAVA, VOVA, CFP position between groups when vision was occluded
VAVO), vibration (rest period [15 s]) or vibration period (VOVO, t = 0.63, p = 0.527) (Table 1) (Fig. 2).
[5 s]) and group (ASD or TDA) on CFP position. A
second linear mixed model analysis was undertaken
using the standard deviation of CFP position (CFP-SD) Mean CFP position – within-group differences in the
within each period (n = 10 rest and n = 10 vibration for use of vision
each participant) to determine if the variance within Within-group pairwise comparisons revealed no
periods was affected by the above-listed factors. For difference in the CFP position of adults with ASD
both analyses interactions at all levels (visual occlusion regardless of when vision was occluded across periods
condition*group, visual occlusion condition*vibration, of postural illusion (VAVA vs. VAVO, t = 1.74,
vibration group*visual occlusion condition) were p = 0.0822; VAVA vs. VOVA, t = 1.47, p = 0.1417;
investigated and significant interactions retained in the VAVO vs. VOVA, t = 0.27, p = 0.7881; VAVA vs.
analysis. Within-group differences were examined post VOVO, t = 0.39, p = 0.6973; VAVO vs. VOVO,
hoc using least squares means of fixed effects and least t = 2.13, p = 0.0335; VOVA vs. VOVO, t = 1.86,
squares mean differences of fixed effects. The SAS p = 0.0632) (Table 1) (Fig. 2). In contrast, a significant
version 9.2 software was used to perform these difference was observed in the CFP position of TDA
analyses (SAS Institute Inc., Cary, NC, USA 2008). when comparing if vision was available across periods
Following convention, a p-value <.05 was taken to of postural illusion (VAVA vs. VOVO, t = 6.12,
indicate a statistically significant difference in all tests. A p < 0.0001; VAVO vs. VOVO, t = 5.25, p < 0.0001;
more stringent p value of .01 was used for the pairwise
comparisons to reduce the chance of a type 1 error.
Table 1. Means and standard deviations in the displacement of the
Centre of Foot Pressure in the sagittal plane (CFP) for adults with
RESULTS autism spectrum (ASD) vs. typically developing (TDA) adults in each of
the four trial conditions (vibration period only): no occlusion (VAVA); no
Mean CFP position – between-group differences occlusion no vibration/occlusion during vibration (VAVO); occlusion no
vibration/no occlusion during vibration (VOVA); and occlusion (VOVO)
No statistically significant difference was observed
between TDA and adults with ASD in age (ASD 23.58 Condition TDA CFP (mm) ASD CFP (mm)
± 7.9 years; TDA 23.35 ± 5.05 years) or mass (ASD Mean SD Mean SD
92.62 ± 23.4 kg; TDA 79.58 ± 12.71 kg). Significant
VAVA 3.2 7.8 8.7 11.3
effects for CFP were observed for group (TDA vs.
VAVO 4.3 11.5 10.2 9.3
ASD), F(1, 2355) = 5.50, p < 0.05, postural illusion
VOVA 3.6 9.9 9.4 11.3
condition F(1, 2355) = 232.26, p < 0.0001, and visual VOVO 7.1 14.6 7.9 8.8
occlusion condition F(3, 2355) = 4.22, p < 0.01, and a
 S. L. Morris et al. / Neuroscience 307 (2015) 273–280
Rest periods Vibration periods
14 (15 seconds) (5 seconds)
CFP position (mm)
12
10
8 significantly smaller CFP-STD in the VAVA vs. VOVO
6 (t = 3.53, p = 0.0004) indicating that the loss of vision
4 in the rest period increased the within-period variability.
2
0
-2
-4
VAVA VAVO VOVA VOVO VAVA VAVO VOVA VOVO
Fig. 2. Mean and 95% confidence interval of the average position of
the center of foot pressure (CFP) during no vibration periods (n = 9)
and vibration periods (n = 10) relative to the initial CFP position at
the start of each trial (n = 1) for participants with ASD (circles)
(n = 12) and TDA (triangles) (n = 20) across the four trial types
(vision available throughout trial VAVA; vision occluded throughout
trial VOVO; vision available only during the pre-vibration periods and
vision occluded during the vibration periods VAVO; and vision
occluded during the pre-vibration periods and vision available during
the vibration periods VOVA).
VOVA vs. VOVO, t = 5.26, p < 0.0001), but not when
comparing when vision was available across periods of
postural illusion (VAVA vs. VAVO, t = 1.05, p = 0.2932;
VAVA vs. VOVA condition, t = 1.68, p = 0.0932; VAVO
vs. VOVA, t = 0.65, p = 0.5153) (Table 1) (Fig. 2).
Mean CFP position – within group differences in the
effect of the postural illusion
Post-hoc pairwise comparisons revealed that TDA were
only affected by the postural illusion when vision was
occluded throughout the postural illusion (VAVA
t = 0.32, p = 0.7496; VAVO t = 0.85, p = 0.3977;
VOVA t = 1.16, p = 0.2466, VOVO t = 3.37,
p < 0.0001); whereas the adults with ASD group were
significantly affected regardless of the availability of
vision (VAVA t = 3.60, p < 0.0001; VAVO t = 4.44,
p < 0.0001; VOVA t = 4.31, p < 0.0001, VOVO
t = 3.41, p = 0.0007) (Table 1) (Fig. 2).
CFP-STD – between-group differences
A significant 3-way interaction for visual occlusion
condition*group*vibration indicated breaking the analysis
into parts for ease of interpretation. Data were split into
rest periods and vibration periods.
CFP-STD – within group differences in the use of
vision
Significant group*visual occlusion condition interactions
were evident during both the vibration (F(3,1222)
p < 0.0001) and rest (F(3,1222) p < 0.0001) periods for
CFP-SD. During vibration, the TDA demonstrated
significantly smaller CFP-STD in the conditions in VAVA
vs. VOVO (t = 5.66, p < 0.0001) and VAVA vs. VAVO
(t = 4.79, p < 0.0001); VOVA vs. VAVO (t = 4. 97,
p < 0.0001) and VOVA vs. VOVO (t = 5.87,
p 6 0.0001) indicating that the loss of vision during the
vibration period increased variability during that period
(underlines indicate the vibration period). No significant
difference was evident for the CFD-STD for the ASD
277
group or between the two groups (p > 0.01) during the
vibration period.
During the rest periods TDA demonstrated
During the rest periods TDA also demonstrated
significantly smaller CFP-STD in the VAVA vs. VAVO
(t = 2.8, p = 0.0052).
Significant within-group differences for adults with
ASD were demonstrated only in the rest periods (not
vibration periods). For the ASD group there was a
significantly greater CFP-STD in the VAVA vs. VOVO
conditions (t = 2.99, p = 0.0029) and the VOVA vs.
VOVO (t = 3.41, p = 0.0007) conditions.
DISCUSSION
The purpose of this experiment was to determine whether
adults with ASD use and integrate visual and
somatosensory information to control CFP in the same
way as typically developing individuals. This was
achieved by comparing the CFP responses of adults
with ASD and TDAs to the illusion of neck extension
(due to posterior neck muscle vibration) when vision
was available and when vision was occluded. The
results showed that adults with ASD leaned forward in
response to the postural illusion regardless of whether
vision was available or occluded. In contrast, TDAs only
leaned forward when vision was occluded. Therefore,
TDA were able to use visual information to override
incongruous information provided by the somatosensory
system during the postural illusion and reset their
standing posture. Surprisingly, adults with ASD relied
entirely on somatosensory information to determine
posture and ignored the visual information even when
the somatosensory information was clearly anomalous.
This result suggests that adults with ASD do not use
vision to control posture but rather appear to be
‘‘selectively blind” to the optic flow information specifying
self-motion. When adults with ASD were presented with
aberrant somatosensory information which induced
them to shift their CFP position forward they were
unable to use visual information to detect and correct
their self-motion.
These results of the vibration study are consistent with
our first hypothesis that proprioception is processed
similarly in TDA and ASD since both typical adults and
adults with ASD responded to the postural illusion by
leaning forward. They also support our second
hypothesis, that vision is processed differently in ASD
because only the TDA group, not adults with ASD, used
visual information to normalize their postural position.
Our findings support the hypothesis that TDAs
preferentially use optic flow information regarding the
motion of the head over somatosensory information
regarding the position of the head in controlling posture
(Bove et al., 2009). This is consistent with direct percep-
tion theory which predicts that the perception of egocen-
tric motion, or self-motion with respect to the
environment, is facilitated by the visual perception of
278 S. L. Morris et al. / Neuroscience 307 (2015) 273–280
global changes in optic flow information (Gibson, 1966). In
contrast, we found that adults with ASD did not use optic
flow information to control postural position. Instead,
when we perturbed the posture of adults with ASD with
the postural illusion we found that they were not only visu-
ally blind to the fact that they were not leaning backward,
but are also visually blind to the fact that they were lean-
ing forward. Thus, in adults with ASD, somatosensory
information regarding the position of the head clearly
dominates over optic flow information regarding the
motion of the head. Analysis of the within-period variation
(CFP-STD) supported the finding for CFP mean position
in that larger variation was observed in the TDA during
visual occlusion compared with vision available – sug-
gesting a stabilization of posture. This stabilization was
not evidence in the adults with ASD where variation was
similar with and without vision. In fact the adults with
ASD in our study may have experienced visual informa-
tion as disturbing to postural control when vibration was
absent. When standing without vibration, unlike the TDA
the adults with ASD demonstrated a larger variation in
postural position when vision was available compared
with when vision was occluded. Kohen-Raz et al. (1992)
described a similar ‘‘paradoxical postural stress
response” in their study of autistic children and adoles-
cents where they observed an unexpected better stability
with vision occluded and standing on foam. If we consider
people with ASD to be proprioceptive rather than visual
learners as suggested by Haswell and colleagues
(2009), then it may not be surprising that they perform
better on foam and without vision. The findings of our
study certainly suggest that the difficulties that children
with ASD have in using optic flow information to perceive
motion (Gepner et al., 1995; Gepner and Mestre, 2002)
extend into adulthood.
Researchers have suggested that the motor deficits
experienced by individuals with ASD are due to difficulty
integrating visual and somatosensory information
(Molloy et al., 2003). Our results are not consistent with
this position which is a statement of our fourth hypothesis.
Instead we found no difference in the postural response of
adults with ASD across the three experimental conditions
involving some visual occlusion (VOVA, VAVO, VOVO).
Therefore, the ability to integrate visual information with
somatosensory either before or during the postural illu-
sion did not affect postural position compared to the situ-
ation when only somatosensory information was
provided. As far as we understand this is the first study
that attempted to test the integration vs. visual deficit
hypothesis directly. This finding reinforces our contention
that the adults with ASD in our study selectively ignored
the visual set point and used somatosensory information
to control posture regardless of whether the somatosen-
sory information was accurate or inaccurate. Further
research is necessary to confirm these findings including
exploring specific visual deficits in ASD in the absence of
the requirement for integration.
Our findings may explain many of the motor deficits
experienced by individuals with ASD such as motor
incoordination (Fournier et al., 2010b; Gowen and
Hamilton, 2013). Somatosensory input is less accurate
than visual input (Forssberg and Nashner, 1982;
Rodrigues et al., 2010). Therefore over reliance on
somatosensory information for motor control is likely to
result in greater postural error which will be most apparent
in tasks requiring greater postural accuracy. Individuals
with ASD exhibit difficulties in controlling posture during
tasks that require accurate postural control such as sport-
ing activities (Forssberg and Nashner, 1982; Rodrigues
et al., 2010).
Our findings have important clinical implications for
the diagnosis and treatment of ASD. As well as
providing insight into factors that may contribute to the
motor difficulties individuals with ASD experience, the
study also suggests possible avenues for developing
motor interventions. Visual processing is often referred
to as an area of strengths in individuals with ASD
(Dakin and Frith, 2005; Mottron et al., 2006). As a conse-
quence, visual strategies have been successfully used to
promote skill acquisition in individuals with ASD (Tissot
and Evans, 2003). One of the visual techniques, video
modeling has been increasingly popular and recognized
as a promising strategy and has been used to teach indi-
viduals with ASD a range of skills including daily living
skills, social and academic skills and in order to enhance
independence in individuals with ASD (Charlop-Christy
et al., 2000; Shipley-Benamou et al., 2002; D’Ateno
et al., 2003; Ayres and Langone, 2005; Buggey, 2005;
Bellini and Akullian, 2007; Delano, 2007; McCoy and
Hermansen, 2007; Sansosti and Powell-Smith, 2008).
Although these techniques show promising results in
many areas of skill development for individuals with
ASD (McCoy and Hermansen, 2007), the results from
the current study indicate that a more somatosensory
approach to intervention may need to be investigated in
regard to motor control. Finally, our study adds to current
research which has shown that static visual information (i.
e., a visual cue that is not in motion) can guide an action
response (in this case postural adjustment) (Rosalie and
Müller, 2013) which is consistent with a Gibsonian
account of visual perception (Gibson, 1979). Moreover,
our findings suggest that individuals with ASD process
visual information for the control of action atypically. The
dorsal visual stream is thought to dominate in the pro-
cessing of visual information for the control of action
(vision-for-action) (Goodale et al., 1991). Previous
research has reported that individuals with ASD process
visual information for perception typically (Del Viva
et al., 2006). The processing of visual information for per-
ception (vision-for-perception) is thought to be a function
of the ventral visual stream (Goodale et al., 1991). In light
of these findings it appears that individuals with ASD
experience deficits in vision-for-action (dorsal stream)
but not vision-for-perception (ventral stream). It would
be valuable; however, to examine differences in the con-
tribution of the dorsal and ventral visual streams to visual
perception between individuals with ASD and typically
developing individuals in a single experiment.
This study represents a small sample of adults with
ASD and in order to reduce the burden on participants,
diagnosis was not re-evaluated during the study. The
differences between groups in outcomes in this study
 S. L. Morris et al. / Neuroscience 307 (2015) 273–280
should be interpreted as the differences between adults
with HFA and adults without HFA as the findings may
not extend to all sub types of ASD. In addition the
magnitudes of the difference in CFP movement between
groups were small. It should be noted that this small
average difference between groups represented the
average across 10 successive periods of vibration. Bove
and colleagues demonstrated that the size of response
in all conditions of vibration increased across trials
(Bove et al., 2009) so the average likely underestimates
the cumulative effect of vibration. The CFP position
reflects the control of the center of mass in the sagittal
plane. The magnitude of the difference in CFP position
between groups in the vision available condition is the
average error in estimation of center of mass position of
the participants with ASD. Small differences in estimating
average postural position can translate to large differ-
ences in body movement during activity since postural
control underlies the development of motor skill.
CONCLUSION
This experimental study sought to extend the
understanding our how individuals with ASD use and
integrate visual and somatosensory information to
control posture during quiet standing. Through the use
of visual occlusion in combination with a postural illusion
it was determined that adults with ASD rely more on
somatosensory information to control posture than TDAs
and are visually blind to changes in posture during quiet
standing. These findings have furthered the
understanding of visual perception in the setting of ASD
and have important implications in diagnosis and
treatment.
FINANCIAL DISCLOSURES
This study was funded by Curtin University. All authors
except Christopher Foster are employees of Curtin
University. Christopher Foster was a student at Curtin
University. The authors have nothing else to declare in
terms of conflict of interest or financial interests.
Acknowledgments—The authors would like to acknowledge the
inputs of Paul Davey in developing the software and integrating
the hardware for the study. We would also like to acknowledge
the Curtin Autism Research Group for continued inputs and
support of our research.
REFERENCES
American Psychiatric Association (2000) Diagnostic and statistical
manual of mental disorders, 4th edition, text revision (DSM-IV-
TR). Washington: American Psychiatric Publishing.
Ayres KM, Langone J (2005) Intervention and instruction with video
for students with autism: a review of the literature. Educ Training
Dev Disab 40:183–196.
Baranek GT (2002) Efficacy of sensory and motor interventions for
children with autism. J Autism Dev Disord 32:397–422.
Bellini S, Akullian J (2007) A meta-analysis of video modeling and
video self-modeling interventions for children and adolescents
with autism spectrum disorders. Except Children 73:264–287.
279
Bhat AN, Landa RJ, Galloway JC (2011) Current perspectives on
motor functioning in infants, children, and adults with autism
spectrum disorders. Phys Ther 91:1116–1129.
Blaesi S, Wilson M (2010) The mirror reflects both ways: action
influences perception of others. Brain Cogn 72:306–309.
Bove M, Fenoggio C, Tacchino A, Pelosin E, Schieppati M (2009)
Interaction between vision and neck proprioception in the control
of stance. Neuroscience 164:1601–1608.
Buggey T (2005) Video self-modeling applications with students with
autism spectrum disorder in a small private school setting. Focus
Autism Other Dev Disab 20:52–63.
Cascio CJ, Foss-Feig JH, Burnette CP, Heacock JL, Cosby AA
(2012) The rubber hand illusion in children with autism spectrum
disorders: delayed influence of combined tactile and visual input
on proprioception. Autism 16:406–419.
Charlop-Christy M, Le L, Freeman K (2000) A comparison of video
modeling with in vivo modeling for teaching children with autism. J
Autism Dev Disord 30:537–552.
Clearfield MW (2011) Learning to walk changes infants’ social
interactions. Infant Behav Dev 34:15–25.
Dakin S, Frith U (2005) Vagaries of visual perception in autism.
Neuron 48:497–507.
D’Ateno P, Mangiapanello K, Taylor BA (2003) Using video modeling
to teach complex play sequences to a preschooler with autism. J
Positive Behav Interv 5:5–11.
Del Viva MM, Igliozzi R, Tancredi R, Brizzolara D (2006) Spatial and
motion integration in children with autism. Vision Res
46:1242–1252.
Delano ME (2007) Video modeling interventions for individuals with
autism. Remedial Special Educ 28:33–42.
Falkmer T, Anderson K, Falkmer M, Horlin C (2013) Diagnostic
procedures in autism spectrum disorders: a systematic literature
review. Eur Child Adolesc Psychiatry 22:329–340.
Forssberg H, Nashner LM (1982) Ontogenetic development of
postural control in man: adaptation to altered support and visual
conditions during stance. J Neurosci 2:545–552.
Fournier KA, Hass CJ, Naik SK, Lodha N, Cauraugh JH (2010a)
Motor coordination in autism spectrum disorders: a synthesis and
meta-analysis. J Autism Dev Disord 40:1227–1240.
Fournier KA, Kimberg CI, Radonovich KJ, Tillman MD, Chow JW,
Lewis MH, Bodfish JW, Hass CJ (2010b) Decreased static and
dynamic postural control in children with autism spectrum
disorders. Gait Posture 32:6–9.
Fuentes C, Mostofsky S, Bastian A (2011) No proprioceptive deficits
in autism despite movement-related sensory and execution
impairments. J Autism Dev Disord 41:1352–1361.
Gepner B, Mestre DR (2002) Brief report: postural reactivity to fast
visual motion differentiates autistic from children with Asperger
syndrome. J Autism Dev Disord 32:231–238.
Gepner B, Mestre D, Masson G, de Schonen S (1995) Postural
effects of motion in young autistic children. NeuroReport
6:1211–1214.
Gernsbacher MA, Sauer EA, Geye HM, Schweigert EK, Hill
Goldsmith H (2008) Infant and toddler oral- and manual-motor
skills predict later speech fluency in autism. J Child Psychol
Psychiatry 49:43–50.
Ghaziuddin M, Butler E (1998) Clumsiness in autism and Asperger
syndrome: a further report. J Intellect Disab Res 42(Pt 1):43–48.
Gibson JJ (1966) The senses considered as perceptual
systems. Boston: Houghton Mifflin.
Gibson JJ (1979) The ecological approach to visual
perception. Boston, MA: Houghton-Mifflin.
Goodale MA, Milner AD, Jakobson LS, Carey DP (1991) A
neurological dissociation between perceiving objects and
grasping them. Nature 349:154–156.
Gowen E, Hamilton A (2013) Motor abilities in autism: a review using
a computational context. J Autism Dev Disord 43:323–344.
Greffou S, Bertone A, Hahler EM, Hanssens JM, Mottron L, Faubert J
(2012) Postural hypo-reactivity in autism is contingent on
development and visual environment: a fully immersive virtual
reality study. J Autism Dev Disord 42:961–970.
280 S. L. Morris et al. / Neuroscience 307 (2015) 273–280
Haswell CC, Izawa J, Dowell LR, Mostofsky SH, Shadmehr R (2009)
Representation of internal models of action in the autistic brain.
Nat Neurosci 12:970–972.
Kohen-Raz R, Volkmar FR, Cohen DJ (1992) Postural control in
children. J Autism Dev Disord 22:419–431.
Lai M-C, Lombardo MV, Baron-Cohen S (2014) Autism. Lancet
383:896–910.
Lord C, Bishop SL (2015) Recent advances in autism research as
reflected in DSM-5 criteria for autism spectrum disorder. Ann Rev
Clin Psychol 11:53–70.
McCoy K, Hermansen E (2007) Video modeling for individuals with
autism: a review of model types and effects. Educ Treat Child
30:183–213.
Memari AH, Ghanouni P, Gharibzadeh S, Eghlidi J, Ziaee V,
Moshayedi P (2013) Postural sway patterns in children with
autism spectrum disorder compared with typically developing
children. Res Autism Spect Disord 7:325–332.
Minshew NJ, Sung K, Jones BL, Furman JM (2004)
Underdevelopment of the postural control system in autism.
Neurology 63:2056–2061.
Miyahara M (2013) Meta review of systematic and meta analytic
reviews on movement differences, effect of movement based
interventions, and the underlying neural mechanisms in autism
spectrum disorder. Front Integr neurosci 7:16.
Molloy CA, Dietrich KN, Bhattacharya A (2003) Postural stability in
children with autism spectrum disorder. J Autism Dev Disord
33:643–652.
Mottron L, Dawson M, Soulières I, Hubert B, Burack J (2006)
Enhanced perceptual functioning in autism: an update, and eight
principles of autistic perception. J Autism Dev Disord 36:27–43.
Nolan L, Grigorenko A, Thorstensson A (2005) Balance control: sex
and age differences in 9- to 16-year-olds. Dev Med Child Neurol
47:449–454.
(Accepted 18 August 2015)
(Available online 24 August 2015)
Pan C-Y, Tsai C-L, Chu C-H (2009) Fundamental movement skills in
children diagnosed with autism spectrum disorders and attention
deficit hyperactivity disorder. J Autism Dev Disord 39:1694–1705.
Peterka R, Benolken M (1995) Role of somatosensory and vestibular
cues in attenuating visually induced human postural sway. Exp
Brain Res 105:101–110.
Rodrigues EC, Lemos T, Gouvea B, Volchan E, Imbiriba LA, Vargas
CD (2010) Kinesthetic motor imagery modulates body sway.
Neuroscience 169:743–750.
Rosalie SM, Müller S (2013) Timing of in situ visual information pick-
up that differentiates expert and near-expert anticipation in a
complex motor skill. Q J Exp Psychol 66:1951–1962.
Sansosti FJ, Powell-Smith KA (2008) Using computer-presented
social stories and video models to increase the social
communication skills of children with high-functioning autism
spectrum disorders. J Positive Behav Interv 10:162–178.
Shipley-Benamou R, Lutzker JR, Taubman M (2002) Teaching daily
living skills to children with autism through instructional video
modeling. J Positive Behav Interv 4:166–177.
Simmons D (2009) Vision in autism spectrum disorders. Vision Res
49:2705–2739.
Tissot C, Evans R (2003) Visual teaching strategies for children with
autism. Early Child Dev Care 173:425–433.
Travers B, Powell P, Klinger L, Klinger M (2013) Motor difficulties in
autism spectrum disorder: linking symptom severity and postural
stability. J Autism Dev Disord 43:1568–1583.
Volker MA (2012) Introduction to the special issue: high-functioning
autism spectrum disorders in the schools. Psychol Sch
49:911–916.
Weimer AMY (2001) ‘‘Motor” impairment in Asperger syndrome:
evidence for a deficit in proprioception. J Dev Behav Pediatr
22:92–101.