Cognitive Brain Research 16 (2003) 192–198
www.elsevier.com / locate / cogbrainres
Research report
Triggering of protective stepping for the control of human balance:
age and contextual dependence
Mark W. Rogers a,b , *, Lois D. Hedman b , Marjorie E. Johnson b , Kathy M. Martinez b ,
Marie-Laure Mille b
a
b
Institute for Neuroscience, The Feinberg School of Medicine, Northwestern University, Chicago, IL, USA
Department of Physical Therapy and Human Movement Sciences, The Feinberg School of Medicine, Northwestern University,
645 North Michigan Ave., Suite 1100, Chicago, IL 60611, USA
Accepted 8 November 2002
Abstract
Human stepping is a commonly executed control strategy for maintaining standing balance in the natural environment. Aging changes
in the initiation triggering of both voluntary (longer latency) and perturbation-induced (shorter latency) stepping are associated with
falling, and are a complex function of altered sensorimotor, neuromuscular, and cognitive system factors. The aim of this study was to
determine the effect of contextual uncertainty about balance stability on the triggering of protective stepping in young and older
individuals. Subjects initiated forward stepping during simple reaction time and waist-pull perturbation conditions with and without
contextual uncertainty about balance stability. The results showed that, regardless of age, the initiation timing for triggering both
voluntary and induced stepping was delayed substantially (100–300 ms) by the presence of balance uncertainty, and that age-associated
timing differences were exacerbated with contextual uncertainty. The initiation timing of the first step liftoff for perturbation-induced
stepping did not reflect entirely an immediate necessity or last resort strategy to balance instability determined directly by specific sensory
input, but rather a decision to step. Moreover, the time to liftoff onset for perturbation-induced stepping was similar for the old and young
with contextual certainty, and occurred 130 ms earlier for the old than for the young when balance stability was uncertain. Overall, we
concluded that older individuals can retain a residual capacity to sustain stationary standing stability as a function of the prevailing task
conditions, and that the reduced timing threshold with age may involve a pre-selected strategy triggered earlier by non-specific
event-related sensory input rather than specific movement-related information.
 2002 Elsevier Science B.V. All rights reserved.
Theme: Motor systems and sensorimotor integration
Topic: Control of posture and movement
Keywords: Balance; Protective stepping; Aging; Intention and context
1. Introduction
Stepping is a commonly executed human movement
strategy for protection against falling in the everyday
environment [5]. As a control problem, the nervous system
must regulate the spatiotemporal relationship between the
position and motion of the body center of mass (COM)
and the changing base of support (BOS). The initiation
triggering of protective stepping is a complex function of
sensorimotor mechanisms, neuromuscular constraints and
*Corresponding author.
E-mail address: [email protected] (M.W. Rogers).
other physical factors, and cognitive influences including
situational context. Protective stepping may be triggered
voluntarily using predictive control when the characteristics of the movement produced instability are known in
advance, or steps may be triggered reactively whenever the
COM–BOS relationship is perturbed by external means.
Aging changes in psychomotor, sensorimotor and neuromuscular systems precipitate impaired control of postural
balance and falls [15,18,20]. For example, aging delays in
the initiation timing of voluntary movement have been
identified consistently for reaction time (RT) paradigms
[18] including RT stepping [2,4,12,16]. The longer latency
to initiate voluntary stepping is a marker for increased risk
0926-6410 / 02 / $ – see front matter  2002 Elsevier Science B.V. All rights reserved.
doi:10.1016/S0926-6410(02)00273-2
M.W. Rogers et al. / Cognitive Brain Research 16 (2003) 192–198
of falling [2]. Such RT delays are attributable to increased
central processing time and sensorimotor thresholds, and
slower neural conduction [18].
Paradoxically, the triggering of perturbation induced
stepping is more frequent with aging [3,5,11] and often
occurs earlier such that the shorter initiation timing is also
a marker for risk of falling [3,4,11,17]. Furthermore,
induced stepping is triggered for smaller magnitudes of
balance perturbation for older versus younger individuals
[3,11]. Since it is unlikely that sensorimotor conduction is
faster or of lower threshold in aging, does this indicate that
stepping is not triggered directly by the specific sensory
input reflecting the evolving instability? Rather, do the
greater incidence and lower timing and magnitude thresholds for induced stepping with age, indicate that stepping
is pre-selected as a strategy based on an internal representation estimate of system dynamics? In this case, the
system knows from experience that pre-stepping feet-inplace compensatory responses (e.g., ankle torque strategy)
are inadequate [10,20] so it predetermines that a step will
be needed when an input trigger arrives even before the
step may actually be required.
We have recently provided indirect support for the
pre-selection alternative by determining the relationships
between sensorimotor and neuromuscular system functions, and the threshold boundary for triggering protective
stepping across a large array of displacement-velocity
combinations of waist-pull perturbations of stance in
young and older subjects [9]. Principally, the smallest
amplitude of perturbation that induced stepping at higher
velocities was lower in the old and was not correlated with
any of the somatosensory, visual, multi-sensory, or neuromuscular system variables evaluated. In view of psychophysiological studies [1,14] of limb proprioception identifying a threshold for detecting an ‘event’ that precedes
the threshold for detecting the specific properties of
movement (e.g., direction and magnitude), it is conceivable
that the old pre-selected to step such that stepping was
triggered when an input stimulus arrived even before the
step was required. Since diminished sensorimotor capacity
with age can compromise effective feet-in-place compensatory postural reactions [10,15,20], the system could
predetermine that a step will be needed based on an
estimate of altered system dynamics. Therefore, the ageassociated reduction in the threshold boundary for stepping
was not apparently related directly to sensorimotor or
neuromuscular deficits but appeared to be a pre-selected
strategy that was triggered by the perturbation ‘event’.
If the earlier triggering of perturbation induced stepping
with aging reflects a decision to step rather than a
necessity, then it should be possible to delay the onset
timing of stepping under certain experimental conditions.
This study determined the effect of contextual uncertainty
about balance stability on the initiation timing of protective
stepping in young and older adults. We hypothesized that if
subjects used a pre-selection strategy, then combining an
193
imperative voluntary RT instruction set with the potential
for perturbation would delay the triggering of induced
stepping in the presence of uncertainty about the impending state of stability that could accompany stepping. With
this paradigm, the premature triggering of a step would
potentially exacerbate instability through synchronization
of voluntarily generated stepping and perturbation induced
stepping. We also expected to observe aging differences in
the capacity to modify the timing of response triggering as
a function of contextual uncertainty.
2. Materials and methods
2.1. Subjects
Thirty-two healthy older subjects (26 women, six men;
mean age, 73 years) and 14 younger subjects (10 women,
four men; mean age, 31 years) participated in the study.
Older subjects were recruited through the Buehler Center
on Aging at Northwestern University Medical School and
through the Geriatric Evaluation Service at Northwestern
Memorial Hospital. Exclusion criteria for older subjects
included significant medical history of cardiovascular,
pulmonary, neurological, or musculoskeletal disease. Older
subjects were screened through an initial phone interview
and a subsequent physical examination by a physician
geriatrician. Prior to testing, all subjects provided written
informed consent to participate in the study.
2.2. Experimental protocol
Overall, three sets of trials involving four task conditions were administered. For all trials, subjects stood
with each foot on a separate force platform (Advanced
Mechanical Technology, Newton, MA) using a standardized foot position. An online visual display controlled
the initial postural weight-bearing conditions prior to each
trial [17]. To ensure safety, subjects wore a safety harness
that prevented injury but did not restrict normal movement.
A few practice trials were provided at the beginning of
each set.
The first set of six trials (Voluntary Certain) required
subjects to respond to a light cue reaction signal (RS),
preceded by an auditory warning signal (WS) by taking
two steps forward under simple RT conditions. The WS
was a high-pitched tone, while the RS was a light-emitting
diode mounted at eye level on a panel 2 m from the
subject. The time between WS and RS randomly varied
from 0.5 to 2.5 s. Subjects were instructed to lead with the
right leg and to step ‘as soon and as fast as possible’ in
response to the light-cue.
The next set of three trials (Induced Certain) involved
forward stepping induced by a stepper-motor-driven waistpull system [13]. A flexible cable was attached at one end
to the puller and at the other end to a rigid connection
194
M.W. Rogers et al. / Cognitive Brain Research 16 (2003) 192–198
aligned with the umbilicus on a waist belt. Each forward
waist-pull included a displacement of 13.5 cm, a velocity
of 27 cm / s, and an acceleration of 540.0 cm / s 2 . As
demonstrated in previous studies [11,17], these magnitude
parameters always elicited stepping in young and older
subjects. In response to the waist-pull, subjects were
instructed to ‘react naturally to prevent themselves from
falling.’
The last set of 12 trials presented two variations of the
Voluntary and Induced conditions that introduced contextual uncertainly about balance stability during the
upcoming response. Subjects were instructed to respond to
the RT WS–RS sequence with a rapid voluntary step
regardless of whether or not a waist-pull was also introduced. Waist-pulls were absent in six trials (Voluntary
Uncertain) and were present in the other six trials (Induced
Uncertain). The order of the trials was pseudo-randomly
determined prior to the study, and was the same for all
subjects. In the trials that included a waist-pull, the pull
was introduced 1.5, 2.0, or 2.5 s prior to the RS. Seated
rest periods were included between blocks of trials.
2.3. Data analysis
Interactive graphical analysis programs were used to
compute the timing characteristics of anticipatory postural
adjustments (APAs) for lateral weight-transfer that accompany stepping, and first step unloading (UL) and liftoff
(LO) for each trial from the vertical ground reaction force
recordings (Fig. 1). The onset of the light RS or the
waist-pull was defined as time zero for determining the
onset timing measures. When balance stability was uncertain, onset timing was determined relative to the onset of
either the waist-pull stimulus (Induced Uncertain) or to the
light-cue stimulus (Voluntary Uncertain). APAs were detected by the presence of a bilaterally asymmetric step
limb loading / stance limb unloading force pattern with an
initial shift in the net medio-lateral center of pressure
towards the step side (see Fig. 1). These mechanical events
assist with unloading the initial stepping limb and precompensate for lateral postural instability at lift-off by
propelling the COM towards the upcoming single stance
side [5,8,17]. APA onset corresponded with the earliest
increase in vertical force beneath the first stepping limb.
UL onset was defined as the peak of vertical force beneath
the stepping foot. LO was identified as the time when the
vertical force beneath the stepping foot reached zero. The
duration of the APA comprised the time period between
the onset of the APA and the onset of UL, and the duration
of unloading was defined as the time period between the
onset of UL and onset of LO. The analysis program
automatically marked the events and data were confirmed
visually [16,17].
2.4. Statistical analysis
The effects of age, task and context on the mean
dependent timing variables were assessed using a threeway analysis of variance that included Group (Young–
Old) as the between-subjects factor with repeated measures
on the within-subjects factors of Task (Voluntary–Induced)
and Context (Certain–Uncertain). In cases of significance,
pairwise mean comparisons for meaningful subsets were
performed using the Scheffe´ method. For all tests, a
significance level was set at P#0.05.
3. Results
During waist-pull trials APAs were absent for two
younger subjects and four older subjects. Technical problems resulted in a total of three missing trials involving
one younger subject. Since the repeated measures ANOVA
requires that all dependent measures be present in all
subjects for inclusion, the absence of responses was treated
as missing data.
Fig. 1. Representative example of the timing events during a voluntary
stepping trial. The vertical ground reaction force beneath the first stepping
limb and the associated displacement of the net center of pressure (COP)
in the medio-lateral (M-L) direction recorded from two separate force
platforms are shown. The broken vertical line marks the onset of the
reaction stimulus (RS) light-cue, APA is the onset of the anticipatory
postural adjustment, UL is the onset of unloading beneath the stepping
limb, and LO marks the instant of limb liftoff. APA duration is interval of
time between APA and UL, and unloading duration is defined by the
interval between UL and LO. COP displacement is expressed as a
percentage of the width of the subject’s base of support (BOS).
3.1. APA characteristics
The results are summarized in Fig. 2. There were no
significant (P.0.050) between-group differences, group
interactions or main effect of Task condition for the APA
onset time.
A significant main effect of Context condition
(F(1,38)511.49, P,0.002) was observed indicating that,
M.W. Rogers et al. / Cognitive Brain Research 16 (2003) 192–198
195
Fig. 2. The influence of age and contextual uncertainty about balance stability on the initiation timing of voluntary and waist-pull induced stepping. (A)
Onset timing latency of the anticipatory postural adjustment (APA) for lateral weight transfer; (B) the duration of the APA; (C) the duration of the first step
unloading time; (D) the onset timing latency of the first step liftoff. Group mean values61 S.E.M. are shown for young (filled circles) and older (open
squares) subjects as a function of Certain versus Uncertain balance stability contexts.
across the groups and tasks, the APA onset was delayed
when the Context was Uncertain (Fig. 2A).
For APA duration (Fig. 2B), a significant Group3Task
interaction (F(1,38)59.38, P,0.004) was observed. Regardless of the Context, the APA was shorter (P50.027) in
duration for the Old than for the Young during the Induced
Task, but did not differ (P50.738) between the groups for
the Voluntary Task. A main effect of Context (F(1,38)5
22.42, P,0.001) and a Task3Context interaction
(F(1,38)54.25, P,0.046) showed that the effect of Context was mainly during the Induced stepping Task. The
APA duration was longer (P50.001) for Induced Uncertain
trials compared with Induced Certain trials while Voluntary
Uncertain and Certain durations were not statistically
different (P50.197). The mean APA durations were not
statistically different between Voluntary and Induced tasks
when compared within either the Certain Context (P5
0.419) or the Uncertain Context (P50.541).
3.2. Step characteristics
During the stepping phase, a main effect of the Task
(F(1,44)516.34, P,0.001) and of the Context (F(1,44)5
33.59, P,0.001) was observed for unloading duration
(Fig. 2C). A significant Group3Task interaction
(F(1,44)55.96, P,0.019) revealed that the Old had a
longer (P50.052) unloading duration than the Young for
Voluntary stepping while the groups were comparable
(P50.957) for Induced stepping. A significant Task3
Context interaction (F(1,44)529.92, P,0.001) further
indicated that, again, the effect of Context was primarily
during the Induced stepping task. The unloading duration
was similar (P50.929) between the tasks for Context
Certain but was significantly longer (P50.001) during
Induced stepping for Context Uncertain. The unloading
duration for Induced Uncertain trials was also significantly
(P50.001) longer than for Induced Certain and Voluntary
Certain trials.
The onset timing of the first step liftoff (Fig. 2D)
demonstrated significant main effects for Task (F(1,44)5
6.64, P,0.013) and Context condition (F(1,44)5104.19,
P,0.001). A Group3Task interaction (F(1,44)5123.29,
P,0.001) showed that, across the Context conditions,
Voluntary Task liftoff was marginally later (P50.098) for
the Old than for the Young while Induced Task liftoff was
not different (P50.411) between the groups. A Task3
196
M.W. Rogers et al. / Cognitive Brain Research 16 (2003) 192–198
Context interaction effect (F(1,44)519.21, P,0.001) also
revealed that the effect of the Context was greater during
the Induced Task than during the Voluntary Task. When the
Context was Uncertain, the time to liftoff onset was
delayed by (100 ms in the Voluntary Task (P50.001) and
by (260 ms during the Induced Task (P50.001). A
marginally significant (F(1,44)53.99, P,0.052) Group3
Task3Context interaction effect also indicated a trend for
the generally longer Voluntary liftoff onset latency for the
Old versus the Young to become even longer, and for the
Induced liftoff latency to become shorter for Context
Uncertain compared with the Certain condition (see Fig.
2D).
4. Discussion
The results of this study supported the hypothesis that
the initiation timing for triggering perturbation-induced
stepping does not reflect entirely an immediate necessity or
last resort response to balance instability determined
directly by specific sensory input, but rather a decision to
step. Both young and older subjects delayed their step
initiation timing in response to the waist-pull perturbation
applied while awaiting an imperative RT light-cue to step,
compared with when the same perturbation was presented
alone and they reacted naturally. The results further
indicated that step triggering depends on factors additional
to the instantaneous ability to maintain a corrective feet-inplace standing position. Aging differences in step liftoff
initiation timing also varied as a function of the form of
stepping.
To identify the locus of the changes in step triggering,
we determined which phases of the step timing sequence
were altered by the introduction of contextual uncertainty
about balance stability. Across the age groups, delaying the
Voluntary and Induced onset timing of the APA and
prolonging the duration of the APA and unloading phases
for Induced stepping accounted for the delayed liftoff
timing when uncertainty was introduced. These results
indicated that both the anticipatory postural and stepping
phases were modified as a function of Context Uncertainty,
and reflect, at least in part the influence of prior-planning
on perturbation-induced stepping as well as for voluntary
stepping.
As indicated by the Task3Context interaction effect for
unloading duration, the time of first step unloading for
Induced stepping was prolonged by contextual uncertainty
to a greater extent than for Voluntary stepping. This
augmented delay manifested in the later stages of the
postural–step sequence suggested that the Induced steps
were not totally preplanned and that their release was
subject to being modulated ‘on-line.’
For Voluntary stepping, the delay in responding to the
light-cue with contextual uncertainty when the possibility
of perturbation was introduced, might have involved a
division of attention in order to avoid initiating a voluntary
step at the same time as an induced step was triggered.
Such synchronization of stepping responses would exacerbate instability by increasing the forward momentum
of the body through the addition of passive and active
motion.
When compared with the Young, the Old tended to
trigger Induced liftoff timing at a similar latency for the
Context Certain condition, and earlier during the Context
Uncertain condition. This trend was associated with a
significantly shorter APA duration for the Old during
Context Uncertain trials (see Fig. 2B). These results
confirmed previous observations that older individuals can
trigger Induced stepping reactions as rapidly or more
rapidly than younger individuals [4,17]. However, by
reducing the overall time that the APA was sustained, older
subjects may have compromised their lateral stability in
order to achieve a more rapid adjustment in their BOS.
Compared with the Young, older individuals also demonstrated a longer unloading duration for Voluntary stepping regardless of Context. Consistent with a previous
report [16], the onset timing of the APA was not affected
by age compared with the later phase of Voluntary
stepping preceding lift-off (unloading duration). Thus, once
the decision to step has been made, older individuals can
trigger anticipatory postural events as rapidly as younger
subjects but, once activated, require more time to execute
the evolving locomotion phase. Moreover, the introduction
of uncertainty appeared to further increase the Voluntary
liftoff timing differences between the Old and Young.
Overall, the present results provide additional insight
into the paradox of triggering protective stepping in aging.
The general observations that older individuals have a
longer latency to trigger voluntary stepping [2,4,12,16] and
that perturbation-induced stepping is triggered as fast or
earlier in the old versus young [3,4,17], even when the
direction of perturbation is not known in advance [4], were
unaltered by the situational context of balance uncertainty.
However, while liftoff was triggered 50 ms earlier for the
old during the Induced Certain condition, their initiation
timing during Induced Uncertain trials occurred 130 ms
earlier than for the young (see Fig. 2D). Thus, the trend for
older individuals to trigger stepping earlier was exacerbated by the contextual uncertainty about balance stability.
Compared with the contextually more certain perturbation
condition, it is possible that the old triggered stepping
much earlier than the young because of the even greater
potential cost to stability in conjunction with reduced
balance confidence, anxiety about falling, or diminished
attention [6,7,19].
At least for the conditions investigated here, protective
stepping was apparently not triggered directly by specific
sensory input reflecting the state of balance stability but
appeared to involve a pre-selection process that was
initiated before it may have actually been needed. The
finding that the latency for triggering perturbation-induced
M.W. Rogers et al. / Cognitive Brain Research 16 (2003) 192–198
stepping is delayed substantially by contextual uncertainty,
and that a prior investigation [9] determined that altered
sensorimotor and neuromuscular system functions in aging
do not associate with a lower spatiotemporal threshold
boundary for stepping, is consistent with this perspective.
With balance uncertainty, however, subjects might have
prolonged their initiation timing to a threshold level that
more directly approached their mechanical limits of stability to recover balance with steps as estimated via on-line
sensory information. Whatever the origin of this shift in
threshold timing, it was clearly demonstrated that, at least
for current experimental conditions, younger and older
individuals possess a previously unreported and potent
contextual dependent capacity to delay substantially (by
200–300 ms) the triggering of Induced stepping. In this
regard, a previous report [8] showed that the incidence of
stepping in younger individuals for ‘react naturally’ trials
could be reduced by instructing subjects to ‘try not to
step.’ However, the extent and consistency of the delays
identified in the present study were striking since our past
studies [11,17] found that subjects always stepped for the
same perturbation magnitude applied here even when they
were instructed to avoid stepping [9]. Further, using
biomechanical analyses, we have also previously observed
[11] that the same subjects often triggered stepping well in
advance of the limit of their mechanical margin of
stability. This was particularly true for the old. Therefore, a
relatively conservative margin of safety was apparently
adopted whereby subjects initiated stepping well before
reaching their limits of stability beyond which point steps
would be required to prevent falling [11].
Psychophysiological studies [1,14] of limb proprioception in young adults have revealed that there is a threshold
for detecting an event that is lower than the threshold for
detecting the specific properties of movement. Such sensory information about the occurrence of an event does not
contain information about movement size and direction
[1]. Since diminished proprioception associated with instability including inadequate triggering of feet-in-place
postural reactions are common accompaniments of aging
[10,15,20], the system could predetermine that a step will
be needed based on an internal representation estimate of
system dynamics. In this case, stepping will be initiated
when an input trigger arrives even before the step may
actually be required. We postulate that if older individuals
step primarily to perturbation ‘event’ detection rather than
specific movement information detection, then they will
make errors in their initiation timing. Thus, stepping may
be triggered earlier by non-specific ‘event’-related sensory
input rather than specific movement-related information.
In summary, the present study demonstrated that protective stepping was not likely triggered directly by specific
sensory input but appeared, at least in part, to be a
pre-selected strategy that was triggered by the perturbation
‘event’. Regardless of age, the triggering of step initiation
in response to waist-pull perturbation of sufficient mag-
197
nitude to force stepping in a predictable direction involves
a decision component additional to the need to step due to
ineffective feet-in-place strategies. The present results
emphasize that the situational context of testing conditions
must be taken into consideration when interpreting the
postural balance control capabilities of older individuals
and other clinical populations. The results indicate further
that older individuals can retain a residual capacity to
sustain stationary standing stability as a function of the
prevailing task conditions, and may pre-select a stepping
strategy based on an on internal representation estimate
linking intrinsic (bodily) and extrinsic (environmental)
dynamics.
Acknowledgements
This work was supported by the National Institutes of
Health grants K01 AG00581 and R01 AG16780 to MWR.
The contributions of M. Payton, L. Vervaeke-Robinson, A.
Ross, A. Smith and R. Weir are gratefully acknowledged.
We thank Dr. Richard Fitzpatrick for helpful discussion.
Portions of this work were presented at the Combined
Sections Meeting of the American Physical Therapy
Association, New Orleans, LA, 2000.
References
[1] L.A. Hall, D.I. McCloskey, Detections of movements imposed on
finger, elbow and shoulder joints, J. Physiol. 335 (1983) 519–533.
[2] S.R. Lord, R.C. Fitzpatrick, Choice stepping reaction time: a
composite measure of falls risk in older people, J. Gerontol. Biol.
Med. Sci. 56 (2001) M627–632.
[3] C.W. Luchies, N.B. Alexander, A.B. Schultz, J. Ashton-Miller,
Stepping responses of young and old adults to postural disturbances:
kinematics, J. Am. Geriatr. Soc. 42 (1994) 506–512.
[4] C.W. Luchies, D. Wallace, R. Pazdur, S. Young, A.J. DeYoung,
Effects of age on balance assessment using voluntary and involuntary step tasks, J. Gerontol. Biol. Med. Sci. 3 (1999) M140–144.
[5] B.E. Maki, W.E. McIlroy, The role of limb movements in maintaining upright stance: the ‘change-in-support’ strategy, Phys. Ther. 77
(1997) 488–507.
[6] B.E. Maki, P.J. Holliday, A.K. Topper, Fear of falling and postural
performance in the elderly, J. Gerontol. 46 (1991) M123–131.
[7] E.A. Maylor, A.M. Wing, Age differences in postural stability are
increased by additional cognitive demands, J. Gerontol. 51B (1996)
P143–154.
[8] W.E. McIlroy, B.E. Maki, Task constraints on foot movement and
the incidence of compensatory stepping following perturbation of
upright stance, Brain Res. 616 (1993) 30–38.
[9] M.-L. Mille, M.W. Rogers, S.R. Lord, R.C. Fitzpatrick, K. Martinez,
L.D. Hedman, M.E. Johnson, Sensorimotor threshold for protective
stepping in young and older adults, Soc. Neurosci. Abstr. 27 (2001)
834.8.
[10] S. Okada, H. Kazufumi, Y. Takada, H. Kinoshita, Age-related
differences in postural control in humans in response to a sudden
deceleration generated by postural disturbance, Eur. J. Appl. Physiol. 85 (2001) 10–18.
[11] Y.-C. Pai, M.W. Rogers, J. Patton, T.D. Cain, T.A. Hanke, Static
versus dynamic predictions of protective stepping following waist-
198
[12]
[13]
[14]
[15]
[16]
M.W. Rogers et al. / Cognitive Brain Research 16 (2003) 192–198
pull perturbations in young and older adults, J. Biomech. 31 (1998)
1111–1118.
A.E. Patla, J.S. Frank, D.A. Winter, S. Rietdyk, S. Prentice, S.
Prasad, Age-related changes in balance control system: initiation of
stepping, Clin. Biomech. 8 (1993) 179–184.
P.E. Pidcoe, M.W. Rogers, A closed-loop stepper motor waist-pull
system for inducing protective stepping in humans, J. Biomech. 31
(1998) 377–381.
K.M. Refshauge, R. Chan, J.L. Taylor, D.I. McCloskey, Detection
of movements imposed on human hip, knee, ankle and toe joints, J.
Physiol. 488 (1995) 231–241.
M.W. Rogers, D.L. Wardman, S.R. Lord, R.C. Fitzpatrick, Passive
tactile sensory input improves stability during standing, Exp. Brain
Res. 136 (2001) 514–522.
M.W. Rogers, C.G. Kukulka, D. Brunt, T.D. Cain, T.A. Hanke, The
[17]
[18]
[19]
[20]
influence of stimulus cue on the initiation of stepping in young and
older adults, Arch. Phys. Med. Rehabil. 82 (2001) 619–624.
M.W. Rogers, L.D. Hedman, M.E. Johnson, T.D. Cain, T.A. Hanke,
Lateral stability during forward-induced stepping for dynamic
balance recovery in young and older subjects, J. Gerontol. Biol.
Med. Sci. 56 (2001) M589–594.
G.E. Stelmach, C.J. Worringham, Sensorimotor deficits related to
postural stability. Implications for falling in the elderly, Clin.
Geriatr. Med. 1 (1985) 679–694.
N. Teasdale, M. Simoneau, Attentional demands for postural
control: the effects of aging and sensory reintegration, Gait Posture
14 (2001) 203–210.
M.H. Woollacott, Systems contributing to balance disorders in older
adults, J. Gerontol. Biol. Med. Sci. 55 (2000) M424–428.