Journal of Gerontology:

MEDICAL SCIENCES

Copyright 2000 by The Gerontological Society of America

2000, Vol. 55A, No. 3, M147–M154

M147

Gait Characteristics of Young and Older Individuals

Negotiating a Raised Surface: Implications for the

Prevention of Falls

Rezaul K. Begg

1

and William A. Sparrow

2

1

Centre for Rehabilitation, Exercise and Sport Science and School of Human Movement, Recreation and Performance,

Victoria University, Melbourne, Australia.

2

School of Health Sciences, Deakin University, Burwood, Victoria, Australia.

Background.

Falls in older individuals are a major public health issue because of the financial cost of surgery and re-

habilitation and the human cost of associated pain and disability. Older individuals are most likely to fall when negotiat-
ing an obstacle or obstruction during locomotion. This research was aimed at investigating lower limb motion while a
subject negotiated a raised surface.

Methods.

The gait of six healthy young (Y) women (mean age 23.1 years) and six healthy older (O) women (mean

age 67.6 years) were analyzed with a PEAK motion analyzer and a dual-force-platform system during unobstructed
walking and when the subjects were stepping on and off a raised surface of 15 cm. The effect of age on foot clearance
and force platform variables was analyzed.

Results.

During stepping on, the young women cleared the step by the lead foot by a significantly greater margin

than the older subjects did (Y

⫽

10.6 cm, O

⫽

9.1 cm;

p

⬍

.05) but trail-foot clearance was not significantly different (Y

⫽

9.4 cm, O

⫽

8.8 cm). Foot clearance in stepping off was low compared with that of ascent, and the older individuals had

a significantly higher lead (Y

⫽

1.5 cm, O

⫽

3.3 cm,

p

⬍

.05) and trail (Y

⫽

1.0 cm, O

⫽

2.1 cm) vertical clearance.

Older individuals positioned both the lead and the trail foot relatively farther from the step edge on ascending a raised
surface, respectively, Y

⫽

87% and O

⫽

93% of the step cycle and Y

⫽

29% and O

⫽

34%. Foot placement in descent

was qualitatively similar for the two groups. The force and the impulse data under the lead and the trail feet confirm
modulations consistent with the foot clearance data.

Conclusion.

In negotiating a raised surface older individuals appear to use a nonoptimal foot placement strategy in

which, compared with that of young subjects, the trail foot is placed a long way from the edge of the step. The older sub-
jects allowed very little correction time and little latitude in foot placement beyond the edge of the step, suggesting that
the approach to the obstacle may be a critical determinant of safety.

ALLS in older individuals are, worldwide, a major pub-
lic health issue because of the financial cost of treatment

and rehabilitation and the human cost of associated pain and
disability. Tinetti and Speechley (1), for example, reported
that 32% of a sample of community-dwelling older persons
fell at least once a year, with 24% of those sustaining seri-
ous injuries. In Australia, the falls-related financial costs of
medical treatment and rehabilitation have been estimated to
be approximately $2.5 billion per year (2), and in response
to this problem an extensive epidemiological literature has
emerged that documents the incidence and cause of falls in
older people. This work has identified associated risk fac-
tors such as visual and attention deficits, the effects of med-
ication, environmental hazards, and age-related declines in
neuromuscular function (3–5). Despite such findings, little
is known about the direct cause of perceptual–motor errors
associated with walking, particularly when a person is nego-
tiating obstacles, that are likely to lead to a fall.

One category of falls-related incidents that has received

little research attention is that associated with negotiating
raised surfaces. In the everyday environment, people are
continually required to accommodate changes in support
surface elevation. In some cases such changes are small, as

in stepping over the edge of a carpet, but in other situations
the changes in elevation are large, as in climbing stairs or
stepping on or off a roadside curb. A large proportion of
falls in public places occurs on steps (6) and, more gener-
ally, in older individuals, falls are most likely to occur when
they are negotiating an obstacle or obstruction.

Previous experimental studies of potentially hazardous

modifications to lower limb trajectories have primarily con-
cerned changes to the walking gait when a person is step-
ping over an obstacle such as a metal rod (7,8) or a wooden
block (9). There has also been limited research on gait adap-
tations to stair walking (10,11) but there have been no previ-
ous attempts to show why negotiating a raised surface might
pose a particular hazard to older people. This may be appli-
cable to situations in which individuals confront a roadside
curb, a porch step, or other contingencies demanding an
abrupt change in limb elevation.

Previous research has suggested that foot placement be-

fore and after an obstacle may play a critical role in the suc-
cess of obstacle negotiation. Chen and colleagues (12), for
example, showed that older participants not only crossed
obstacles more slowly than young controls but also that
their foot placement was such that the obstacle was crossed

F

M148

BEGG AND SPARROW

farther forward (by 10% of step length) in the swing phase
of the crossing cycle. They argued that this strategy could
be interpreted as a safety mechanism in providing greater
toe clearance over the obstacle. One drawback of this prac-
tice is that, in the event of contact with the obstacle, older
individuals would have less time to recover their gait. Chen
and colleagues (12) did not, however, reveal how individu-
als’ foot placement would be affected in accommodating
obstacles requiring a change in whole body elevation, such
as stepping onto a raised surface.

Foot clearance over an obstacle, measured vertically from

the obstacle to either the toe or heel, has been reported in the
literature, with the assumption that a large clearance repre-
sents greater safety. Chen and colleagues (12) showed lead-
foot clearance of up to 12 cm when subjects were stepping
over a 15-cm obstacle, and Patla and Rietdyk (9) measured
lead-toe clearance at 10 cm independent of obstacle height
(4 cm to 26 cm). Less commonly, foot clearance when sub-
jects were stepping across an obstacle has been investigated,
with clearance measured from either the toe or the heel to
the boundary of the area to be stepped across. Sparrow and
colleagues (11) measured heel clearance when subjects
were stepping across obstacles, when the obstacle to be
stepped across was formed by two parallel strips of tape on
the floor, essentially an obstacle of zero height. Sparrow
and colleagues (11) found that heel clearance of the lead
foot was approximately 7 cm beyond the tape defining the
farther boundary of the area to step across. In the present
study the heel was used as the anatomical marker for lead-
foot clearance and the surface–toe distance was the clear-
ance measure for the trail foot. This was done consistently
with our earlier work, showing the toe to be most frequently
the lowest point on the trail foot in stepping over an obstacle
and, on most trials, the heel was shown to be the lowest
point of the lead foot (7).

Collectively these results suggest that when a person is

stepping over an obstacle that projects vertically, with ef-
fectively zero constraint on limb position horizontally, and
is stepping across obstacles, young individuals invariably
maintain clearance within safe bounds. It has not previously
been shown, however, how foot clearance is affected in ac-
commodating a raised surface. In this case there is a require-
ment to clear the obstacle vertically and horizontally, in
order to position the foot safely beyond the edge of the step.
In the experiment reported here, therefore, both foot place-
ment horizontally and foot clearance vertically were investi-
gated for stepping on and off a raised surface. The aim of
the experiment was therefore to show the effects of age-related
declines in gait on this type of potentially hazardous obsta-
cle negotiation task.

Our related work on stepping over obstacles has revealed

that the lead foot (which goes over the obstacle first) and the
trail foot not only have different trajectories (7) but also dif-
ferent ground reaction-force (kinetic) characteristics (8).
The kinetics of the lead and the trail foot during step negoti-
ation were therefore also of interest, and we used a dual-
force-platform apparatus to measure simultaneously the
force/time characteristics of the trail foot before the raised
surface and the lead-foot kinetics to show the forces exerted
on the platform itself. Patla and colleagues (13) proposed

that during obstructed gait a higher-order parameter such as
impulse is modulated in adapting the gait pattern. To present a
concise picture of the force/time characteristics in raised
surface negotiation, we also selected foot–ground force im-
pulses to highlight possible falls-risk–related differences
between the young and the older individuals. Impulse is the
integral of the force/time curve and represents both the
magnitude of force and the duration of force application.
The capacity to propel the body vertically and horizontally
(in the anterior-posterior direction) in order to negotiate an
obstacle is dependent on the impulses generated in these
two directions of motion. This fundamental observation has
been recognized in previous work (8,13). It was hypothe-
sized that obstacle-crossing trajectories and associated im-
pulses would highlight age-related declines in gait that
might predispose older individuals to falls in the raised sur-
face task.

M

ETHOD

Subjects

Six healthy young women (age 21.2 years, standard devi-

ation [SD] 1.3 years; height: 172.2 cm, SD 8.9 cm; body
mass: 73.1 kg, SD 14.2 kg; leg length: 87.7 cm, SD 5.0 cm),
and six healthy older women (age: 67.6 years, SD 4.8 years;
height: 158.0 cm, SD 4.7 cm; body mass: 66.1 kg, SD 10.1
kg; leg length: 80.3 cm, SD 2.8 cm) volunteered to partici-
pate in the experiment. The young adults were recruited
from the academic community of Victoria University and
the older subjects were community-living individuals drawn
from our subject pool. An in-house questionnaire was used
to screen for musculoskeletal and visual impairments that
might affect normal locomotion. All participants completed
informed consent procedures approved by the Victoria Uni-
versity Research Ethics Committee.

Equipment and Experimental Setup

The experimental setup is shown in the top panel of Fig-

ure 1. Lower limb movement was recorded with a PEAK
(Peak Technologies Inc.) two-dimensional motion analysis
system interfaced to a personal computer, which also ran the
data smoothing, and analysis programs. The raised surface
was a solid wooden construction 0.15 m high, 1 m wide, and
5 m long. The Peak video camera was positioned approxi-
mately 8 m from the platform edge perpendicular to the
plane of motion. The experimental setup also incorporated a
dual-force-platform system with which to simultaneously
record the lead- and the trail-foot ground reaction forces.

Procedure

Reflective markers were attached to the big toe, heel, an-

kle, knee, and hip joints of the subjects’ left and right legs.
Markers were also placed on the raised surface to determine
foot clearance over the step and also to identify the distance
and time characteristics of foot crossing, as shown at the top
of Figure 1. Subjects first undertook trials of unobstructed
walking followed in counterbalanced order by conditions
involving stepping on or off the raised surface. For each
condition there were practice trials followed by 10 experi-
mental trials, and in all conditions instructions were to walk

NEGOTIATING A RAISED SURFACE

M149

at a normal comfortable speed and negotiate the raised sur-
face (if present) normally as in the natural environment,
“...such as when stepping on or off the kerb.” Subjects were
advised to rest between trials if required and a researcher
was stationed alongside the raised surface to assist the sub-
ject in the event of unusual unsteadiness.

Data Analysis

For each of the conditions, raw horizontal and vertical

marker coordinates were automatically digitized and smoothed
with a low-pass digital filter with a cutoff frequency of 6 Hz
for one complete lead- and trail-foot stride for 5 trials ran-
domly selected in each condition. The marker trajectories
were calibrated with the Peak software and calculated with
respect to absolute time and as a percentage of the step cy-
cle, i.e., from preobstacle toe off to postobstacle foot con-
tact of the lead and the trail feet. Key kinematic variables
including vertical and horizontal clearances and time/dis-
tance parameters of the crossing step were calculated from
the smoothed marker displacement data for both lead and
trail feet. These dependent variables are also illustrated in
Figure 1.

In addition to the kinematic variables, impulse variables

extracted from the ground reaction-force data for both the
lead and the trail feet were used in the analysis. Vertical and
anterior-posterior impulses during the braking (Brake) and

propulsive (Prop) phases of the gait cycle were calculated
from the corresponding force–time curves, as shown at the
bottom of Figure 1. To eliminate between-subject differ-
ences, the impulse values were normalized by body mass (in
Newton seconds per kilogram). The values of the above ki-
nematic and kinetic variables for the five trials for individual
subjects in each condition (unobstructed walking, stepping on,
and stepping off) were computed for statistical analysis. The
effect of age on the dependent variables was determined
with one-way analysis of variance (SPSS Inc.) procedures.
Because the condition effect on the dependent variables is
not a major issue for this report, aging effects were deter-
mined for each of the conditions separately. For those vari-
ables specific to the raised surface, such as foot clearance,
the analysis was run for two conditions, on and off. The .05
probability level was used as the criterion for statistically
significant differences between group means for each of the
dependent measures.

R

ESULTS

Stride Length and Duration and Gait Cycle
Phase Modulations

The most straightforward measures of gait adaptations

associated with negotiating a raised surface are the modifi-
cations to length and duration of the lead-foot and the trail-
foot stride. The data presented in Table 1 show the gait ad-
aptations by young and older individuals in stride length
and duration for unobstructed walking and stepping on and
off the raised surface. For unobstructed walking the older
individuals had significantly shorter stride length for both
the lead (

F

[1,10]

⫽

14.3,

p

⬍

.01) and trail (

F

[1,10]

⫽

12.2,

p

⬍

.01) foot. Stride length is influenced by many factors

including age, stature, and disability. The older group was
significantly shorter (

p

⬍

.01) in stature and also had shorter

(

p

⬍

.01) leg length than their younger counterparts. These

differences in anthropometric characteristics may have con-

Figure 1. A, Experimental setup illustrating the distance/time and

foot clearance variables as subjects were stepping on a raised surface,
and B, Typical force–time curves (vertical and anterior-posterior)
during gait obtained by the force platforms. The area under the curve
represents impulse. The total impulse has been divided into braking
and propulsive components.

Table 1. Mean Values for Stride Length, Stride Duration and

Stride Velocity (Stride Length/Stride Duration) for Young and

Older Subjects During Unobstructed Walking and in Stepping On

and Stepping Off a 15-cm Raised Surface

Lead Foot

Trail Foot

Parameter

Young

Old

Young

Old

Unobstructed

Stride length, m

1.18 (.10)

0.97 (.10)**

1.18 (.10)

1.02 (.05)**

Stride duration, s

0.91 (.08)

0.86 (.05)

0.89 (.02)

0.88 (.05)

Stride velocity, m/s

1.29

1.13

1.32

1.16

Stepping on

Stride length, m

1.13 (.14)

0.93 (.08)*

1.11 (.11)

0.95 (.06)*

Stride duration, s

0.98 (.05)

0.92 (.06)

0.99 (.08)

1.04 (.06)

Stride velocity, m/s

1.15

1.01

1.11

.91

Stepping off

Stride length, m

1.16 (.12)

0.92 (.09)**

1.18 (.15)

1.02 (.06)*

Stride duration, s

1.02 (.05)

0.98 (.08)

0.88 (.03)

0.80 (.04)

Stride velocity, m/s

1.14

0.94

1.34

1.28

Notes:

Standard deviation is given in parentheses. Differences between

groups significant at *

p

⬍

.05, **

p

⬍

.01.

M150

BEGG AND SPARROW

tributed to some of the between-group differences in stride
length. In normal unobstructed gait there should be no dis-
tinction between lead and trail foot, such that the stride
parameters for both feet would be expected to be approxi-
mately equal. For the older subjects the trail-foot stride was,
however, 5 cm longer than that of the lead foot, suggesting
an asymmetry associated with visually guiding the lead foot
onto the further force platform and, in so doing, shortening
the lead foot stride.

The data in Table 1 also indicate that the young subjects

had significantly greater lead (

F

[1,10]

⫽

8.7,

p

⬍

.05) and

trail (

F

[1,10]

⫽

8.2,

p

⬍

.05) stride lengths in the stepping-

on condition. In stepping off, lead (

F

[1,10]

⫽

15.4,

p

⬍

.01)

and trail (

F

[1,10]

⫽

6.4,

p

⬍

.05) stride lengths were also

significantly greater for the young group. When the be-
tween-group differences in stride duration were taken into
account, the combined effect of modifications to length and
duration was that the older participants tended to walk more
slowly, as reflected in the slower stride velocities. In addi-
tion, there was a systematic increase in lead-foot stride du-
ration across the three conditions with the effect of the
raised surface being to slow the stride, and, interestingly,
stepping off (down) revealed a longer lead stride than step-
ping on. Trail-foot duration increased for both groups, rela-
tive to unobstructed walking, in stepping on, but decreased
in the stepping-off condition. It is interesting to note that the
shortest stride duration for the older subjects was as they
were stepping off the raised surface.

Vertical Clearance

Vertical clearances (as illustrated in Figure 1) for step-

ping on and off are shown in Figure 2. The older partici-
pants had significantly lower lead-foot (heel) clearance
(

F

[1,10]

⫽

7.9,

p

⬍

.05) whereas the toe of the trail foot

showed no significant clearance difference. For both groups
clearance during stepping off was low compared with the
stepping-on condition, ranging from

ⵑ

1.0 cm to 3.3 cm. It

is interesting that, in stepping off, the older individuals
cleared the step by a significantly greater margin with the
lead (

F

[1,10]

⫽

6.6,

p

⬍

.05) foot; the trail-foot clearance

was also greater, but not significantly.

Horizontal Clearance and Foot Placement

Figure 3 is a plan view of the mean foot placement posi-

tions both before the step edge and on the raised surface it-
self for stepping on and off, with significant between-group

Figure 2. Mean (

⫾ standard deviation) lead-foot (heel) and trail-

foot (toe) vertical clearances during stepping on and stepping off a 15-cm
raised surface for young and older individuals. Significant differences
between the young and the older groups are starred, *p

⬍ .05.

Figure 3. Plan view of lead- and trail-foot placement prestep and poststep negotiation. Significant young–old differences are starred, *p

⬍

.01, **p

⬍ .001.

NEGOTIATING A RAISED SURFACE

M151

differences in foot position starred. In stepping off, there ap-
peared to be little qualitative difference in foot position be-
tween the young and the older participants. The stepping-off
foot positions on the left of Figure 3 simply reflect the abso-
lute differences in step length between the two groups con-
sistent with the stride length data presented in Table 1. In
contrast, the foot placements of the two subject groups for
stepping on show two important differences. First, note that
on approaching the raised surface the trail foot of the young
and the older individuals was placed approximately the
same distance from the step edge. Given the differences in
absolute step length this implies that the relative position of
the trail foot for the older individuals is farther away (as a
proportion of step length) from the step edge. The second
important observation on the stepping-on data is that the
older person’s lead foot (heel) contacted the raised surface
very close to the step edge compared with that of the young
subjects (

F

[1,10]

⫽

57.5,

p

⬍

.001), on average with only

approximately 6-cm horizontal clearance. SDs of lead-foot
horizontal clearance of 9.1 and 2.9 cm were calculated for
the young and the older groups, respectively, suggesting
that the older people with such a small horizontal clearance
were restricted to placing their foot within a narrow margin.
These two observations suggest a different strategy for step-
ping on for the young and the older walkers based on funda-
mental differences in lead- and trail-foot placement before
they negotiated the raised surface.

The different step negotiation characteristics of the young

and the older people suggested in the data presented in Fig-
ure 3 are highlighted when foot placements’ prestep and
poststep edge are expressed qualitatively as percentages of

step time and step duration, as in Figure 4. The bottom
panel of Figure 3 confirms the above observation that in
stepping on the older participants positioned their trail foot
relatively farther away from the step edge, at 34% of their
normalized step length compared with 29% for the young
walkers. The older individuals also had significantly greater
lead-foot percentage distance from the step edge (

F

[1,10]

⫽

54.9,

p

⬍

.001) before step negotiation and, possibly as a

consequence, placed their lead foot on the step such that
only approximately 7% of step length and 5% of step dura-
tion remained before contact with the step surface. In con-
trast, the younger walkers allowed considerably more time
(approximately 24% of the step cycle) and almost double
the distance (13% of the step cycle) in which to position the
lead foot after crossing the step edge. It is interesting to
note, however, that during stepping off the percentage of
duration and time of the lead-foot events shown in Figure 4
were not significantly different between the two groups. In
summary, the data suggest that, compared with young sub-
jects, the older individuals position the lead and the trail
foot at a greater distance from the step edge on ascending a
raised surface, whereas foot position in descent is qualita-
tively similar to that of young people.

Lead- and Trail-Foot Kinetics

The data presented in Figure 5 are the vertical and hori-

zontal impulses for stepping on and off the raised surface as
measured with the dual-force-platform system. Impulses for
the stepping-on and the stepping-off conditions were di-
vided by the unobstructed condition values to calculate im-
pulse ratios for the two groups. A ratio of unity reflects

Figure 4. Lead- and trail-foot prestep and poststep crossing percentage distances and times.

M152

BEGG AND SPARROW

Figure 5. Force impulse ratios (relative to unobstructed condition) under the trail and the lead feet during stepping-on and stepping-off con-

ditions. The data in the bar graphs indicate group mean

⫾ standard deviation. These impulse ratios have been shown during the braking and the

propulsive phases of the support phase (see Figure 1). Note different scaling in anterior-posterior impulse values for the stepping-off condition
that are due to large mean and SD values.

NEGOTIATING A RAISED SURFACE

M153

precisely the same impulses for the unobstructed and the
raised surface conditions whereas values greater than unity
indicate that either stepping on or off produced a greater im-
pulse than the unobstructed condition per kilogram of body
mass. To link these findings to the foot placement data de-
scribed above, impulse ratios were calculated for both the
braking and the propulsive phases of foot-ground contact
(see Figure 1) and the lead- and the trail-foot data are also
considered separately. The results for stepping on and step-
ping off are presented in separate subsections below.

Stepping On

The vertical impulses as shown in the top panel of Figure

5 under the trail and the lead feet showed relatively higher
braking impulse in the older group whereas in the propul-
sive phase age-related changes were minimal. However,
none of these differences were significant. In the anterior-
posterior direction, the older individuals showed more brak-
ing impulse and less propulsive impulse for the trail foot.
But young–old differences in anterior-posterior impulses
were minimal under the lead foot.

Stepping Off

Figure 5 also shows (bottom panels) that in stepping off,

the trail-foot braking vertical impulse for the older individu-
als was considerably greater than for the young subjects by
approximately 13%. The anterior-posterior braking impulse
was, again, also considerably higher in the older group com-
pared with that of their younger counterparts. Both the the
vertical braking and vertical propulsive impulses were rela-
tively lower in the older group under the lead foot. It is also
interesting to note that in the anterior-posterior direction the
young subjects decreased braking impulse, whereas the
older individuals increased the braking impulse relative to
the unobstructed impulses under the lead foot. The propul-
sive impulse under the lead foot increased in both groups,
but changes were greater for the young.

D

ISCUSSION

Vertical foot clearance over a step is clearly an important

parameter for safe transit. For stepping onto a raised surface
of 15 cm, we found clearances of 8.8 to 10.6 cm, approxi-
mately the same amplitude as reported previously for step-
ping over an obstacle of approximately the same height
(9,12). This suggests some similarity in gait adaptation for
stepping on the raised surface and for stepping over obstacles.
For stepping off a raised surface, foot clearance was low
compared with that of stepping on, less than 3 cm, and was
particularly low for the trail foot. Such low clearances would
appear to demand very precise foot trajectory control (partic-
ularly by the trail foot) during stepping off a raised surface.
Relatively higher clearances for the older participants may re-
flect a safety strategy in stepping off a raised surface.

The most important finding here was that in approaching

the raised surface the older individuals appear to have used
a nonoptimal foot placement strategy in which, compared
with that of young subjects, the trail foot was placed farther
away from the edge of the step as a proportion of stride dis-
tance. One consequence of trail-foot placement far from the
surface edge is that, in the event of disturbance to the lead-

foot trajectory, the older subjects would have very little cor-
rection time (5% of step time) and little latitude in foot
placement beyond the step edge. This observation implies
that foot placement on the approach to a raised surface may
be a critical determinant of safety. Previous findings (4,5)
have emphasized muscular strength and declines in neuro-
muscular function as dominant factors in the capacity to ne-
gotiate obstructions safely. In light of the present findings,
however, it appears that older people’s foot placement strat-
egy is nonoptimal and, on first consideration, such a strat-
egy would not appear to be directly associated with declines
in neuromuscular capacities such as strength and flexibility.

In stepping on and off raised surfaces and in other obsta-

cle avoidance tasks, the trail foot is positioned first, and vi-
sual guidance of foot position relative to the obstacle is the
critical process underlying this ability. The general principle
of gait control that we advance here is that optimal place-
ment of the trail foot relative to obstructions is essential for
safe traverse. Once the trail limb has been positioned, the
intrinsic dynamics of the lead limb play a major role in de-
termining lead-foot trajectory over obstacles or obstructions.

The pattern of impulse modulations seen in Figure 5 is

consistent with the above observation concerning the lower
horizontal foot clearance beyond the surface edge by the
lead foot. As shown in Figures 3 and 4, horizontal step
clearance (both absolute and as a percentage of step length)
for the older participants when stepping on was signifi-
cantly lower. This feature of raised surface negotiation was
highlighted as potentially hazardous because of the older
person’s relatively small (approximately 6-cm) horizontal
clearance. The low anterior-posterior trail-foot propulsive
impulse combined with increased braking impulse associ-
ated with the older walkers did not therefore assist the lead
foot in traversing the step edge. Rather, the older individu-
als stepped onto the surface with a short lead-foot clearance
that was not assisted by impulses from the trail foot. An in-
crease in vertical impulse would assist in elevating the
whole body and consequently facilitate vertical foot clear-
ance. In this task, the significantly greater lead-foot clear-
ance of older individuals in stepping off, presented in Figure
2, may therefore have been engendered by their appreciably
greater vertical trail-foot impulses.

All the participants in the study were screened for mus-

culoskeletal or visual impairments by use of an in-house
questionnaire. Subjects reporting any balance or locomotor
problems were not included in the study and there was no
evidence during testing of participants having difficulties in
maintaining balance. The between-group differences re-
ported herein can therefore reasonably be considered as
primarily due to age-related declines in neuromuscular
function.

In the raised surface task investigated here there are a

number of risk-related consequences of nonoptimal trail-
foot placement by older individuals. First, when the trail
foot was positioned farther from the step, the lead-foot con-
tacted the raised surface later in the stride. As suggested
above, in the event of either obstacle contact or other distur-
bance to stability there would be less correction time or
“available response time” (14) in which to arrest the for-
ward motion of the whole body center of mass before it

M154

BEGG AND SPARROW

moves forward beyond the base of support provided by the
feet. Second, trail-foot position before the step is associated
with vertical clearance (between the heel and the surface
edge), and in this experiment in stepping on the signifi-
cantly lower lead-foot clearance by the older individuals
provided less margin for error in the vertical direction.
Third, the narrow margin between the edge of the step and
the lead-foot heel shown by the older individuals implies re-
duced capacity to lengthen the step to achieve stable foot
placement in the event of any imbalance. The observation
that the young subjects made heel contact on average 17 cm
from the step edge compared with a 6-cm horizontal clear-
ance for the older individuals indicates that lead-foot place-
ment was much less constrained in young people.

To investigate further the effects of lead-foot constraint

we calculated the mean within-subject step-to-heel SD be-
yond the step edge, in other words the variability associated
with the lead-foot horizontal clearance. The low SD of lead-
foot horizontal clearance in older individuals (2.9 compared
with 9.1 cm for the young) suggests that, in contrast to older
individuals who place the trail foot closer to the step, lead-
foot placement by young people is permitted, as it were, to
be considerably more variable because of the greater dis-
tance from the step edge. A foot placement strategy in nego-
tiating raised surfaces that allows greater variability in foot
placement, within safe bounds, would impose less con-
straint on lower limb targeting and presumably allow atten-
tion to be diverted to other features of the environment. In
road crossing, for example, attention could be diverted to
oncoming vehicles. The general implication of the findings
reported here is that falls and perhaps other accidents in
older individuals may, in part, be due to increased attention
demands associated with the primary task (negotiating ob-
stacles), leaving less time to monitor critical and in some
cases hazardous features of the external environment.

Acknowledgments

This research was supported by Grant ARC SGS19/95 from the Austra-

lian Research Council. The authors thank Kate Kloot and Daniel Halliday
for assistance with the data analysis and figures.

Address correspondence to Dr. Rezaul Begg, Biomechanics Unit, Cen-

tre for Rehabilitation, Exercise and Sport Science, Victoria University, PO
Box 14428, MCMC, Melbourne, Victoria 8001, Australia. E-mail: rezaul.
[e-mail]

References

1. Tinetti ME, Speechley M. Prevention of falls among the elderly.

New

Engl J Med.

1989;320:1055–1059.

2. Fildes B.

Injury Prevention Among the Elderly: Falls at Home and Pe-

destrian Accidents.

Melbourne: Dove, 1994.

3. Waller JA. Falls among the elderly—human and environmental fac-

tors.

Accid Anal Prev.

1978;10:21–33.

4. Prudham D, Evans JG. Factors associated with falls in the elderly: a

community study

. Age Ageing.

1981;10:141–146.

5. Lord S, McLean D, Stathers G. Physiological factors associated with

injurious falls in older people living in the community.

Gerontology.

1992;38:338–346.

6. Simoneau GG, Cavanagh PR, Ulbrecht JS, Leibowitz HW, Tyrell RA.

The influence of visual factors on falls-related kinematic variables during
stair descent by older women.

J Gerontol Med Sci.

1991;46:M118–M195.

7. Sparrow WA, Shinkfield AJ, Chow S, Begg RK. Gait characteristics

in stepping over obstacles.

Hum Mov Sci.

1996;15:605–622.

8. Begg RK, Sparrow WA, Lythgo ND. Time-domain analysis of foot-

ground reaction forces in negotiating obstacles.

Gait Posture.

1998;7:

99–109.

9. Patla AE, Rietdyk S. Visual control of limb trajectory over obstacles

during locomotion: effect of obstacle height and width.

Gait Posture.

1993;1:45–60.

10. Mcfayden BJ, Winter WA. An integrated biomechanical analysis of

normal stair ascent and descent.

J Biomech.

1988;21:733–744.

11. Sparrow WA, Shinkfield AJ, Summers JJ. Gait characteristics in indi-

viduals with mental retardation: unobstructed level walking, negotiating
obstacles and stair climbing.

Hum Mov Sci.

1998;17:167–187.

12. Chen HC, Ashton-Miller JA, Alexander NB, Schultz AB. Stepping

over obstacles: gait patterns of healthy young and old adults.

J Geron-

tol Med Sci.

1991;46:M196–M203.

13. Patla AE, Prentice SD, Robinson C, Newfold J. Visual control of loco-

motion: Strategies for changing direction and for going over obstacles.

J Exp Psychol.

1991;17:603–634.

14. Chen HC, Ashton-Miller JA, Alexander NB, Schultz AB. Effect of age

and available response time on ability to step over an obstacle.

J Ger-

ontol Med Sci.

1994;49:M227–M233.

Received April 6, 1999
Accepted August 6, 1999
Decision Editor: William B. Ershler, MD