Exp Brain Res (2004) 159: 487–490
DOI 10.1007/s00221-004-1971-5
RESEARCH ARTICLES
Naznin Virji-Babul . Michelle Brown
Stepping over obstacles: anticipatory modifications in children
with and without Down syndrome
Received: 4 December 2003 / Accepted: 8 April 2004 / Published online: 9 July 2004
# Springer-Verlag 2004
Abstract The purpose of this study was to explore the
mechanism of anticipatory control of gait in relation to the
perception of an obstacle. Typically developing (TD)
children (4–7 years of age) and children with Down
syndrome (5–6 years of age) walked and stepped over
obstacles of two different heights—a “subtle” obstacle that
was placed at a very low distance from the floor (1% of
total body height) and an “obvious” obstacle that was
placed at a much higher distance from the floor (15% of
total body height). Spatial and temporal measures of the
gait cycle were analyzed. TD children showed increased
variability in pre-obstacle step lengths only in response to
the higher obstacle. Children with DS showed a decrease
in variability in response to the higher obstacle and
marked qualitative changes in their gait cycle. Both groups
of children were able to scale toe clearance with obstacle
height. These results show that TD young children can
make task-specific anticipatory adjustments by modulating
step length and toe clearance. Children with DS show
appropriate scaling of toe clearance and are beginning to
show the emergence of anticipatory responses under
specific environmental conditions.
Keywords Obstacle . Locomotion . Children . Down
syndrome
N. Virji-Babul (*)
Centre for Human Movement Analysis, Queen Alexandra
Centre for Children’s Health,
2400 Arbutus Road,
Victoria, B.C., V8 N 1V7, Canada
e-mail: naznin@dsrf.org
Tel.: +1-250-7216849
Fax: +1-250-7216837
N. Virji-Babul
Down Syndrome Research Foundation,
Vancouver, B.C., Canada
M. Brown
School of Physical Education, University of Victoria,
Victoria, B.C., Canada
Introduction
Children with Down syndrome (DS) often exhibit both
motor and perceptual difficulties that impact on motor
development. While specific sensory, motor, cognitive and
perceptual impairments have been widely reported, the
way these localized deficits impact on perceptual-motor
processing and function remains unclear. The interaction
between perception and action in DS has often focused on
examining goal-directed movements of the upper limbs. It
has been suggested that poor performance may be due to
impairments in regulating the temporal aspects of move-
ment (Henderson 1981; Savelsbergh et al. 2000). Charlton
et al. (2000) reported that children with DS have difficulty
in matching both the spatial and temporal aspects of their
grasp in relation to the size of the object or to the task of
the goal. They suggest that difficulties in using extrinsic
information about the properties of an object in relation to
the task may point to a dysfunction in relating information
about limb position with respect to environmental/task
demands.
Anecdotal evidence suggests that individuals with DS
often have trouble avoiding obstacles, leading to increased
frequency of trips and falling. The key question is: are the
deficits in avoidance strategies related to difficulties in the
visual perception of the object, in the execution of
appropriate strategies, or associated with attention-related
processes?
One experimental approach that has been used to study
the relationship between visual perception and motor
output is through anticipatory gait modifications during
obstacle crossing. Walking in the natural environment
often involves anticipatory adaptations to either avoid or
negotiate obstacles. The interaction between the visual
perception of the obstacle and the execution of strategies
to avoid the obstacle has recently received much attention
in adults (Pryde and Patla 1997; Patla and Rietdyk 1993;
McFadyen and Winter 1991). However, few studies have
studied the developmental basis of anticipatory behaviour
during walking (McFadyen et al. 2001), or in individuals
with intellectual disability (Sparrow et al. 1998). To our
488
knowledge there has been no previous research on how
children with DS negotiate obstacles.
The present study investigated the strategies used by
children with and without DS to negotiate obstacles. In
order to evaluate the effect of the visual perception of the
obstacle on the resulting movement strategies, we chose
one obstacle height of a very low distance off the floor to
create a “subtle” obstacle and compared the responses with
those to an obstacle of a much higher distance from the
floor.
Materials and methods
Participants
Six typically developing (TD) children (mean age =5 years; range:
4–7 years, SD:1.1); mean height (1.1 m; SD=.09), with no history of
neurological or musculoskeletal deficits, and five children with DS
(mean age =5.5 years, range: 5–6 years, SD=.63; mean height
=1.0 m, SD=.05) participated in the study. The average mental age
of the children with Down syndrome, as assessed by the Peabody
Picture Vocabulary Test—Revised (PPVT-R), was 46 months
(SD=15 months). Participants were recruited from elementary
schools in Victoria, B.C. Data were collected at the Centre for
Human Movement Analysis at the Queen Alexandra Centre for
Children’s Health in Victoria, B.C. The University of Victoria
Human Ethics Committee, as well as the VIHA Child, Youth and
Maternal Health Program Research Advisory Committee approved
the research protocol. Written informed consent was obtained from
the parents of each participant, as well as verbal assent from each
child.
Apparatus
Lower limb movement was recorded using a 3D VICON motion
analysis system (Oxford Metrics, Oxford, UK) interfaced with a
personal computer, which also ran the data smoothing and analysis
programs. Six infrared cameras mounted on tripods were used with a
sampling frequency of 120 Hz. Reflective markers (14 mm in
diameter) were placed on the right and left sides of the body
(anterior-superior iliac spine, posterior superior iliac spine, thigh,
tibia, estimated knee center, lateral malleolus, heel and over the
second metatarsal head. These markers were used to define different
segments of the pelvis and lower limbs. Sixteen markers were
placed on each side of the body for a total of 32 markers. Markers
were also placed on the obstacle to identify the characteristics of
foot crossing. The obstacle was constructed with a piece of
micropore tape connected to two vertical poles separated by a
distance of 1 m. Micropore tape was used as we felt there was less
chance of the children risking injury if there were to trip over the
obstacle. The obstacle was placed at a distance of 2.5 m from the
start of the walkway.
Procedure
Participants first performed three unobstructed walking trials. This
was followed by walking under two obstructed conditions:
1. Obstacle height of 1% of total height
2. Obstacle height of 15% of total height
The micropore tape was moved to create obstacles of specific
heights for each child. A total of nine walking trials were collected.
The order of presentation of obstacle heights was counterbalanced
across participants. One of the investigators demonstrated the task
prior to the start of data collection. All participants were instructed
to walk at a comfortable speed.
Data analysis
The VICON Plug-In-gait software was used to obtain information
about the following variables: step length (distance between two
successive heel contacts of the opposite legs), step length variability
(i.e. SD of the step length), stride length (distance between
consecutive heel contacts for the same leg), stride duration, velocity
and toe clearance (distance between the lead limb toe marker and the
maximal height of the obstacle at the time of obstacle crossing).
Stride length was normalized to leg length for each subject.
Averaged data from the three unobstructed walking trials were
analyzed using a one-way ANOVA. For the obstacle conditions, a 2
(group) by 3 (condition) ANOVA with repeated measures was used
to compare groups and conditions. Only the first trial of each
obstacle condition was used to analyze toe clearance and only the
pre-obstacle analysis for the lead foot is reported here.
Results
There was no significant effect of condition on normalized
stride length or stride velocity. Only normalized stride
length was significantly smaller in the children with DS
(F=6.6, P=.033). There was no significant effect for group
and no significant interaction effects were found for these
measures.
Step length variability
The major significant effects were found in step length
variability for the pre-obstacle steps. We examined
individual step lengths for the first three steps before the
step over the obstacle. Figure 1 shows the mean variability
for the 1 and 15% obstacle conditions for both groups.
Note that average step length variability increased in the
15% condition for the TD children. In contrast, the
children with DS showed an overall decrease in step
length variability in the same condition. The main effect of
condition was significant (F=10.6, P=.02). There was also
a significant interaction effect of Group*Condition
(F=106, P<.05).
Toe clearance
Figure 2 shows the toe clearance for the lead foot for both
groups of children in all three conditions. There is a
significant effect of condition (F=21.95, P=.001). Toe
clearance increased with increased obstacle height. The
group effect was not significant (F=.108, P=.75). Howev-
er, it is worth noting that in both the unobstructed and 1%
obstacle conditions, the children in the DS group, on
average, had lower toe clearance than the children in the
TD group. The average obstacle height for the 1%
condition was 10 mm for the DS group and 11 mm for
the TD group. Thus, the TD children had an average

Fig. 1 Variability of step lengths in the 1% and 15% obstacle
conditions for both typically-developing (TD) and Down syndrome
(DS) participants
margin of 7 cm to clear the obstacle, whereas the margin
for the DS group was only 5 cm.
Qualitative differences
Figure 3 shows the stick figure and trajectory of the lower
limb segment for one representative TD child (top) and
one child with DS (bottom) for the higher obstacle
condition. A number of qualitative differences were
observed. TD children generally showed a smooth
progression from walking to stepping over the obstacle.
In contrast, 4 out of 5 children with DS stopped in front of
the obstacle for various lengths of time ranging from .4 s
to 4 s before stepping over the obstacle. This was not
observed when stepping over the subtle obstacle.
Fig. 2 Mean and standard deviation of toe clearance for the first
trial under all three conditions for TD and DS participants
489
Fig. 3 Stick figure and trajectory of the lower limb segment for the
15% obstacle condition. Top figure shows the trajectory from one
representative TD participant. Bottom figure shows the trajectory
from one representative DS participant
Discussion
Our goal was to determine if children with and without DS
between the ages of 4 and 7 years are able to make
anticipatory adjustments to their gait in response to
different obstacle heights. We chose the 1 and 15%
obstacle heights to evaluate if and how gait is modulated
in response to subtle and obvious changes in the
environment. We expected that if anticipatory behaviour
is task or experience based, modifications would be
observed that are specifically matched to the environ-
mental constraint.
Anticipatory adjustments in TD young children
In our study, the distance to the obstacle remained the
same in both conditions. The main difference was the
change in obstacle height. This difference led to specific
changes in step length and toe clearance that were
dependent on the obstacle height, suggesting that visual
information about obstacle height and the distance to the
obstacle, is extracted as soon as, or at the initiation of, gait
(Patla and Vickers 1997). Regulation of step length has
been shown to be critical for task-specific modulation in
adults (Pryde and Patla 1997; Patla et al. 1989). Using
computer simulations Taga (1998) has also shown that
modulating step length and modifying the trajectory of the
swing limbs’ obstacle are two strategies used to clear
obstacles. Whether children are able to use the same
strategies, and the developmental age at which such
strategies emerge, have not previously been demonstrated.
McFadyen et al. (2001) reported that children between 7
and 9 years are capable of clearance obstacles of moderate
height, but that anticipatory strategies are still maturing.
490
Ledebt et al. (1998) compared shifts in the centre of
pressure as an index of anticipatory postural adjustments
during gait initiation in children between the ages of 2.5
and 8 years of age. They reported that anticipatory
behaviour was observed in the youngest age group, and
systematic behaviour was consistently observed in chil-
dren from the 6-year-old age group onwards.
The inconsistency in terms of the developmental age at
which anticipatory changes begin to emerge is in part
related to the types of measures used to study this
behaviour and the experimental task conditions. Interest-
ingly, our results showed that overall stride length and
stride velocity remained consistent across conditions. This
may suggest that the nervous system maintains a
consistent, overall steady state gait pattern and that
modulation of gait in response to the environment occurs
within these global parameters.
Strategies used by children with DS
Children with DS also showed a consistent overall stride
length and stride velocity across conditions. Stride length
was shorter than that found in TD children and is likely to
be related to the anthropometric differences between the
two groups. Interestingly, children with DS showed a
robust scaling of toe elevation to obstacle height, similar to
the TD children. This implies that these children were able
to successfully extract information about obstacle height,
and appropriately match this information to their move-
ments. However, the question is whether this information
is used in a feed-forward or on-line mode of control. Our
results seem to indicate that visual information about the
obstacle is not used consistently to modulate movements
early in the gait cycle. First, greater step length variability
was observed in response to the subtle obstacle, suggest-
ing that some form of anticipatory adjustments were being
made. In contrast, there was very little variability observed
in response to the higher obstacle. This finding, in
combination with the observations that children with DS
stopped in front of the higher obstacle, indicates that
anticipatory adjustments could not be made effectively and
that the children adopted the gait pattern of the
unobstructed condition. This would suggest that these
children were unable to use advance visual information
about the obstacle and had to wait until they reached the
obstacle to extract the relevant information to modulate
their motor output, i.e. stepping over the obstacle.
Another point worth noting is that the children with DS
had a lower toe clearance compared with the TD children.
This was particularly true for the unobstructed and
minimal obstacle conditions. The lower toe clearance
means that the margin of error for clearing minimal
obstructions found in the natural environment is quite low
and increases the possibility of hitting the obstacle, leading
to increased instances of tripping and stumbling. The
results from this study suggest that children with DS have
deficits in both visual perception and attention-related
processes. Further study is currently underway to examine
the trajectories of the hip and knee joints to determine
whether any changes can be observed in the execution of
the movement during obstacle crossing.
Acknowledgements The authors are grateful to the children and
their parents for their participation. This work was supported by the
Queen Alexandra Foundation.
References
Charlton J, Ibsen E, Lavelle BM (2000) Control of manual skills in
children with Down Syndrome. In: Weeks DJ, Chua R, Elliott
D (eds) Perceptual-motor behavior in Down syndrome. Human
kinetics, Champaign, IL, pp 151–174
Henderson SE, Morris J, Frith U (1981) The motor deficit of
Down’s syndrome children: a problem of timing? J Child
Psychol Psychiatry 22:233–245
Ledebt A, Bril B, Breniere Y (1998) The build-up of anticipatory
behaviour. An analysis of the development of gait initiation in
children. Exp Brain Res 120:9–17
McFadyen BJ, Winter DA (1991) Anticipatory locomotor adjust-
ments during obstructed human walking. Neurosci Res Com-
mun 9:37–44
McFadyen BJ, Malouin, F, Dumas, F (2001) Anticipatory locomotor
control for obstacle avoidance in mid-childhood aged children.
Gait Posture 13:7–16
Patla AE, Rietdyk S (1993) Visual control of limb trajectory over
obstacles during locomotion: effect of obstacle height and
width. Gait Posture 1:45–60
Patla AE, Vickers JN (1997) Where and when do we look as we
approach and step over an obstacle in the travel path?
Neuroreport 8:3661–3665
Patla AE, Robinson C, Samways M, Armstrong CJ (1989) Visual
control of step length during overground locomotion: task-
specific modulation of the locomotor synergy. J Exp Psychol
Hum Percept Perform 15:603–617
Pryde KM, Roy EA, Patla AE (1997) Age-related trends in
locomotor ability and obstacle avoidance. Hum Mov Sci
16:507–516
Savelsbergh G, van der Kamp J, Ledebt A, Planinsek T (2000)
Information-movement coupling in children with Down syn-
drome. In Weeks DJ, Chua R, Elliott D (eds) Perceptual-motor
behavior in Down syndrome. Human Kinetics, Champaign, IL,
pp 151–174
Sparrow WA, Shinkfield AJ, Summers JJ (1998) Gait characteristics
in individuals with mental retardation: unobstructed level-
walking, negotiating obstacles, and stair climbing. Hum Mov
Sci 17:167–187
Taga G (1998) A model of the neuro-musculo-skeletal system for
anticipatory adjustment of human locomotion during obstacle
avoidance. Biol Cybern 78:9–17