JN Watch the video to see how APS reaches out to developing nations.
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

J Neurophysiol 94: 1971-1979, 2005. First published May 31, 2005; doi:10.1152/jn.00266.2005
0022-3077/05 $8.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
94/3/1971    most recent
00266.2005v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (6)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bhatt, T.
Right arrow Articles by Pai, Y.-C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bhatt, T.
Right arrow Articles by Pai, Y.-C.

Long-Term Retention of Gait Stability Improvements

T. Bhatt1,2 and Y.-C. Pai1,2

1Departments of Physical Therapy and 2Movement Sciences, University of Illinois, Chicago, Illinois

Submitted 14 March 2005; accepted in final form 19 May 2005


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 
Evidence of long-term modification of behavior—in particular, gait alterations in response to repeated exposure to slips—within the locomotor-balance control system is limited. The purpose of this study was to examine whether improvements in fall-resisting behavior as reflected by improvements in gait stability could be retained on a long-term basis. Eight healthy young subjects were exposed to a block of repeated slip trials during a single acquisition session consisting of five repeated slip exposures; the same subjects were then re-tested using the same protocol at a minimum of 12 mo later. Pre- and postslip gait stability for all slip trials was measured at touchdown (slipping limb) and liftoff (contralateral limb) based on the center of mass state (i.e., its instantaneous position and velocity) relative to the base of support (BOS) and the predicted thresholds for backward loss of balance. In the acquisition session, subjects were able to increase pre- and postslip stability, which significantly correlated with a decrease in the incidence of balance loss from 100% (1st slip) to 0% (5th slip). All subjects exhibited a similar balance loss on the first slip of the follow-up session. Nonetheless, subjects were able to retain the acquired preslip stability with feedforward control on the first slip but not the postslip stability related to the reactive response. Also, the subjects demonstrated a faster re-acquisition, with only one balance loss on the second slip of the follow-up session, as compared with seven balance losses on the acquisition session. Such rapid improvements were achieved by the significantly greater increase in post- compared with preslip stability; this increase was for the most part, a consequence of reductions in slip intensity (i.e., the peak BOS velocity). We concluded that a single acquisition session could only produce limited long-term retainable effects within the locomotor-balance control system. It appeared, however, that the CNS was still primed to more rapidly update its internal representation of gait stability during re-acquisition.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
METHODS
 RESULTS
 DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 
Falls are the leading cause of injury-related death or hospitalization in older adult populations and a significant contributor to a decrease in quality of life (Baker and Harvey 1985Go). Learning to increase the ability of the CNS to employ successful motor adaptation, i.e., strategies to deal with abrupt changing external constraints, could thus be fundamental in preventing falls and maintaining daily function. Although fundamentally different from an actual fall, unintended loss of balance is often a precursor to a fall. A successfully acquired and retained motor behavior that promotes an individual’s own neuromuscular protective mechanisms by improving stability and reducing occurrence of balance loss perturbation is therefore a prerequisite to the prevention of falls.

The adaptability of the posture and locomotion control system under various perturbation conditions is well established (Andres et al. 1991Go; Horak and Nashner 1986Go; Horak et al. 1997Go; Keshner et al. 1987Go; McIlroy and Maki 1995Go; Rand et al. 1998Go; Reynolds and Bronstein 2003Go). Only a few studies have established the beneficial effect of such adaptations with respect to any reduction in balance loss or fall incidence (Owings et al. 2001Go; Pai et al. 2003Go; Pavol and Pai 2002Go; Pavol et al. 2002Go). Recent theoretical and empirical evidence has, however, demonstrated that the adaptive control of one’s dynamic stability can indeed be successful in preventing backward loss of balance and slip-related falls if one can adjust his/her center of mass (COM) state (i.e., the COM position and its velocity) relative to the base of support (BOS) above the computationally predicted stability threshold level for backward loss of balance (Pai et al. 2003Go; Pavol and Pai 2002Go).

Because stability can be expressed as the relationship between the BOS and the COM, the CNS could achieve its central goal of adaptive stability control by adjusting the trunk and hence the COM motion (via anticipatory adjustments or reactive responses) to catch up with the perturbed BOS. Alternatively, it could increase stability by minimizing the potential BOS perturbation during gait (Lockhart et al. 2003Go; Marigold and Patla 2002Go; You et al. 2001Go; Bhatt et al. 2005a). However, little is known about the extent to which adapted fall-resisting behavior to external disturbances can be retained (Schwabe et al. 2004Go). Recently, Tjernstrom et al. (2002)Go demonstrated 1-mo retention of acquired changes in magnitude of body sway to calf vibrations during stance. It is possible that the retention effect in this study could have been the result of an extensive training period consisting of multiple sessions. In contrast, a few studies on applying fear conditioning in mice have shown that a single acquisition session is sufficient for long-term retention of the acquired stimulus-response behavior within the CNS (Kim and Fanselow 1992Go; Sacchetti et al. 2004Go). A similar analogy may hold true where a single session of acquisition to deal with a threatening environment may be sufficient to exhibit long-term retention of the acquired motor behavior. Thus variations in the time period required to acquire and convert these temporary sensori-motor associations into permanent motor programs are evident (Brashers-Krug et al. 1996Go; Kandel 2001Go). These differences may reflect the functional significance of the adaptation response, which itself is conditioned by the penalties imposed on an inappropriate response by the CNS and the increased potential of injury (Adkin et al. 2000Go; Carpenter et al. 2001Go).

The purpose of this study was to determine whether gait-stability improvements acquired in a single session could be retained by subjects for ≥12 mo. Our hypothesis was that if the CNS were able to retain the acquired stability improvement, subjects would, when re-tested after 1 yr under the same environment, exhibit a preslip (at touchdown of the slipping limb) and postslip (at liftoff of the contralateral limb) stability significantly greater than the first slip exposure but not significantly different from the last exposure of the acquisition session. The retained improvement in pre- and postslip stability would, we expected, subsequently be reflected in a significantly reduced balance loss incidence on the first slip exposure of the follow-up session.


   METHODS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
RESULTS
 DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 
Subjects

Eight healthy young subjects (28.5 ± 6 yr, 3 males and 5 females) of the 14 subjects who initially participated in the acquisition session completed the study by participating in the follow-up session. Subjects were included in the study after being screened for exclusionary factors such as neurological, musculoskeletal, cardiopulmonary, other systemic disorders, and selected drug usage (e.g., sedatives, anti-anxiety, antihistamines). Of the eight, one subject was a newly joined member of the lab at the time of re-test. It was, however, determined definitively that the subject had not been exposed directly or indirectly to any aspects of the paradigm until the re-test, i.e., not allowed to watch the experiment being performed, analyze data, or participate in discussions about the slipping paradigm. The other seven subjects did not visit the lab nor did they have any contact with us until recruited for the re-test. Prior to participation, all subjects gave informed consent as approved by the University of Illinois at Chicago Institutional Review Board.

Experimental setup and protocol

Slips were induced using a sliding device consisting of a low-friction, nonmotorized moveable platform (29 x 40 cm, 3.85 kg) mounted on linear bearings to a support frame. The device was locked and embedded in a 7-m walkway and was hidden by the stationary decoy platforms surrounding it. The supporting frame of the sliding device was bolted to a force plate (OR6-5-1000, AMTI, Newton, MA) to measure ground reaction forces (GRF). Three additional force plates were placed in such a way that GRF from the steps before and after contact with the moveable platform could be recorded. Slips were induced by a computer-controlled release mechanism that unlocked the moveable platform after touchdown of the slipping foot, when the ratio of the horizontal to vertical GRF exceeded a preset threshold, comparable to a low coefficient of friction of 0.02. The trigger-release signal indicated that the preset threshold for the coefficient of friction was exceeded within a maximum of 20 ms from touchdown of the slipping foot. A computer program written in LabView (National Instruments, Austin, TX) was used for on-line monitoring of GRF and generation of lock-release signal. Once released, the movable platform slid freely on linear bearings and locked on reaching a maximum travel distance of 44 cm. The subjects wore their own athletic shoes and a full-body safety harness attached at the shoulders by a pair of shock absorbing dynamic ropes to a manually driven trolley on a ceiling-mounted I beam. The rope lengths were adjusted so that the knees could not touch the surface of the floor on suspension. The details of the setup are also given elsewhere (Bhatt et al. 2005bGo).

In the acquisition session, the subjects performed three blocks of 10 walking trials each at self-selected slow, regular, and fast speeds with the regular-speed block performed last. The other two speed blocks were performed first, in random order, with no slips induced. At the beginning of the experiment, the subjects were told that they would be walking for three blocks of trials (the number of trials at each speed was not specified) at each of their preferred "slow" and "fast" speeds and that they might be subjected to slips. They were told to try to recover their balance on any slip incidence and then to continue walking. The first slip was induced without prior warning or practice trials. The subjects were not aware of which trial, which speed, or where on the walkway the slip would occur. On the 11th trial at "regular" speed, a slip was induced. After exposure to the first unexpected slip, four consecutive repeated slip trials followed. The subjects were told to continue walking at the same speed as that of the previous trial and that they may or may not be exposed to slipping again. Each subject’s starting position was adjusted so that his or her right (slipping) foot would land entirely on the movable platform. All subjects were able to take at least three steps before stepping on the moveable platform. When called back for the follow-up session ≥12 mo later [16 ± 4 (SD) mo], the subjects followed precisely the same protocol as that of the acquisition session. Because slip-related falls are likely to occur during winter in a yearly cycle (Bentley and Haslam 1998Go; Gao and Abeysekera 2004Go), it was considered imperative to determine whether the acquired gait pattern would be stable enough to prevent slip-related balance loss over a 12-mo period.

Data collection and reduction

Full-body kinematics were recorded at 120 Hz using a six-camera motion-capture system (Motion Analysis, Santa Rosa, CA). Twenty-four light reflective markers were attached to bilateral upper and lower extremities and torso while one marker was attached to the movable platform. Marker coordinates were low-pass filtered at marker-specific optimal cutoff frequencies (range: 4.5–9 Hz) using a recursive second-order Butterworth Filter. Force plate and harness load cell data and trigger-release onset signals were collected at 600 Hz using a 64-channel, 16-bit A/D converter. The ground reaction force and motion data were time synchronized at the time of data collection.

Analyses were restricted to the anteroposterior direction and included the last unperturbed trial at regular speed and all the slip trials. When the contralateral limb landed posterior to the sliding heel (negative postslip step length), the trials were classified as loss of balance trials with protective stepping. Conversely, trials with the contralateral limb landing anterior to the sliding heel (positive postslip step length) were classified as "no loss of balance" trials in which protective stepping was not needed. Trials with false trigger, faulty landings on the moveable platform (3 in 40 trials), missing markers (1 of the 5th slip trials), and a breakdown of the release mechanism (1 person’s 4th and 5th trials), in the acquisition session were excluded from the analysis.

Analysis of gait stability

The COM position and its numerically derived derivative (velocity) were computed from the kinematic data using known gender-dependent segmental parameter information in a 13-segment representation of the body (de Leva 1996Go). The position of the COM in the anteroposterior direction was expressed relative to the rear of the BOS (XCOM/BOS) of the most recent foot to touchdown (i.e., the heel of the sliding foot for slip onset) and normalized to foot length. The COM velocity in the anteroposterior direction was expressed relative to the velocity of the BOS (COM/BOS) and normalized as a dimensionless fraction of (McMahon 1984Go), where g is the acceleration due to gravity and h is height of the subject.

Stability was assessed through comparison of the COM state with the previously published threshold values for backward balance loss under slip conditions (Pai and Iqbal 1999Go). Stability is defined as the shortest distance from this predicted boundary for backward balance loss to the instantaneous COM state (Bhatt et al. 2005bGo; Pai et al. 2003Go) as shown in Fig. 1. The model simulation predicts that a backward loss of balance must occur for COM states below the threshold (i.e., stability <0). Backward balance loss should not occur when the stability measure is above the predicted value for backward balance loss (i.e., stability > 0). Thus more positive values indicate greater stability against backward balance loss. Conversely, a COM state further below the threshold represents an increased likelihood of backward loss of balance under slipping conditions (Bhatt et al. 2005bGo; Pai et al. 2003Go).



View larger version (16K):
[in this window]
[in a new window]
  		FIG. 1. A: the instantaneous gait stability for an instantaneous center of mass (COM) state (diamond) is the shortest distance (double-headed arrow) between the boundary and the COM state. B: COM state trajectories for a slip and a nonslip trial for a representative subject from the acquisition session. The thick gray line represents the boundary for backward loss of balance. The time course of the COM state trajectory is indicated by small black arrows for the slip trial (solid line) and by gray arrows for the nonslip trial (dotted line). Anteroposterior COM position and velocity are normalized to foot length and , respectively, where g is acceleration due to gravity, and h is the body height. The model predicts that a person whose COM state is less stable at slip onset (star) is more likely to undergo a subsequent backward loss of balance (as indicated by the solid line below the threshold) and a touch down posterior to slipping foot.

 
Pre- and postslip response

The instants of step liftoff and touchdown were identified from the vertical ground reaction forces. These values were identified from foot kinematic data if the touchdown occurred outside of the force plates or if both feet were on the same force plate at an instance. Preslip stability was measured and noted at touchdown of the slipping limb prior to mechanical slip onset, i.e., 85 ± 30 ms before displacement onset of the moveable plate. This was comparable to onset of heel displacement on a "slippery surface" after its touchdown (Redfern et al. 2001Go). Postslip stability was recorded at liftoff of the contralateral limb, which occurred 100 ± 50 ms after slip onset. Thus the time between touchdown of slipping limb to liftoff of contralateral limb was on average 185 ms. Preslip step length was calculated as the difference between the heel markers of the slipping foot (right) and the contralateral foot (left) at the touchdown of the slipping foot. At liftoff of the contralateral foot, the BOS kinematics were obtained from its heel marker. There was no relative motion between the heel marker of the slipping foot and the movable plate from touchdown to that instant.

Statistics

The outcome for each subject on each trial was treated as a nominal variable and was assigned values of either 0 (balance loss) or 1 (no balance loss). Nonparametric statistics would therefore be best suitable to analyze the changes in incidence (%) of outcome across trials and between groups (Pavol et al. 2002Go). Thus the Cochran’s Q test, with post hoc {chi}2 tests were performed to test changes in incidence of balance loss with repeated slip trials. {chi}2 tests were also performed to test changes in incidence of balance loss on each trial between the acquisition and follow-up sessions. A mixed-factor repeated-measures ANOVA was performed with gait stability as the dependent variable and with the two sessions and the two time instants (pre- and postslip) as independent fixed factors and with slip trials (S1–S4) as the repeated independent factor. Only four subjects had good data on the fifth slip trial of the acquisition session. Thus only the first four trials were included in the ANOVA; results of the fifth slip trial between sessions and within sessions between trials were analyzed using t-test. Planned t-test were conducted on pre- and postslip stability between the first slip trials of the two sessions and between the first slip of the follow-up and last slip of the acquisition sessions. Significant main effects of the ANOVA were followed up with simple effects (1-way repeated-measures ANOVA for each event for each session) and planned t-test between consecutive trials. Significant interactions were resolved using event x session ANOVAs between consecutive trials with post hoc t-test. Two-way repeated-measures ANOVAs with session as the fixed and trial as the repeated factor were conducted on pre- and postslip XCOM/BOS and COM/BOS. Significant results were followed up with planned comparisons between sessions for each trial and between consecutive trials within each session. Paired t-test between consecutive trials within each session, and between sessions for each trial were also conducted on preslip step length, postslip BOS displacement, and velocity.

Separate Bonferroni corrections were applied to the post hoc t-test. Absolute P values between 0.05 and 0.001 for significant t-test comparisons were reported. A significance level of 0.05 was used for all the analyses. Analyses were performed using SPSS (Chicago, IL)


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 
Long-term retention of gait stability

All subjects exhibited a loss of balance on the first slip experience both during the initial acquisition and at the 12-mo follow-up. In the acquisition session with repeated slip exposure, subjects were able to significantly reduce the incidence of backward balance loss by the third slip trial (P < 0.05, Fig. 2). In the follow-up session, the reduction in incidence of balance loss, however, was much more rapid, i.e., from 100% in the first slip to only 13% in the second slip (P < 0.05, Fig. 2). Such differences in outcomes were directly related to gait stability. Results of the mixed-factor repeated-measures ANOVA indicated a significant main effect of session, [F(1,24) = 7.25, P = 0.01], and a significant main effect of trial [F(4,96) = 40.70, P < 0.001]. The relationship between pre- and postslip stability changed over trials [trial by event interaction: F(4,96) = 23.44, P < 0.001]. Significant main effects of the ANOVA (and interaction) are discussed in detail in the following sections.



View larger version (37K):
[in this window]
[in a new window]
  		FIG. 2. Decrease in number of balance losses (LOBs) from 1st to 4th slip trials for the acquisition (T1, thin-lined pattern bar) and the 1-yr follow-up (T2, thick-lined patterned bar) sessions. The corresponding registrations of no balance losses for each trial are indicated by thin and thick-lined open bars, respectively, for the acquisition and follow-up sessions.

 
Retention of gait stability observed in the first slip

During the follow-up session, subjects exhibited a significantly greater preslip stability on the regular as well as the first slip exposure compared with the first slip exposure of the acquisition session (P = 0.01 for both; Fig. 3, a and b). This significantly greater preslip stability was achieved by a significantly anterior XCOM/BOS on the follow-up session (P < 0.05; Fig. 4). There was no change in preslip gait stability (at touchdown of the slipping limb) between the last exposure of the acquisition session and the first exposure of the follow-up (P > 0.10). As shown in Fig. 5A, the COM state mean for trial 1 on the follow-up session was closer to the threshold than for the acquisition session as reflected in the greater preslip stability. There was no significant change in postslip stability (at liftoff of the contralateral limb) on the regular and first slip trial between the two sessions (P = 0.37, P = 0.10). Postslip stability on the first follow-up slip was significantly lower compared with the last slip of the acquisition session (P < 0.05; Fig. 3, b and e).



View larger version (17K):
[in this window]
[in a new window]
  		FIG. 3. Gait stability (means ± SD) at preslip touchdown of slipping limb (TD, {circ} and {square}) and postslip liftoff of contralateral limb (LO, {bullet} and {blacksquare}) from the acquisition ({circ} and {bullet}) and the follow-up ({square} and {blacksquare}) sessions, for regular walking and 1st through 4th slip trials (S1–S4). Significant differences in gait stability (for both pre- and postslip events) between the 2 sessions for each trial are indicated (*, P < 0.05). Note the differential increase from pre- to postslip stability for the follow-up session compared with the acquisition session on the 2nd slip trial (S2), indicated by the presence of interaction session by event interaction (S x E).

 


View larger version (23K):
[in this window]
[in a new window]
  		FIG. 4. Means ± SD of anteriorposterior COM position relative to the base of support (BOS) (XCOM/BOS) normalized to foot length at preslip touchdown of the slipping limb from the acquisition ({circ}) and follow-up sessions ({bullet}) for the regular walking and 1st through 4th slip trials (S1–S4). Significant between session changes, and trial-to-trial adaptive changes with respect to the preceding trial are indicated by #, and *, respectively (P < 0.05).

 


View larger version (30K):
[in this window]
[in a new window]
  		FIG. 5. The distribution of center of mass states relative to base of support (XCOM/BOS, COM/BOS) at preslip touchdown of slipping limb (A) and postslip liftoff of contralateral limb (b) for all trials classified into balance loss (LOB, open symbols) and no-balance loss (NLOB, filled symbols) for the acquisition (circles) and follow-up (squares) sessions. COM state means (± SD) for slip trials 1 (unfilled), 2 (half-filled), and 4 (filled) from both sessions are plotted as well.

 
There was no change in preslip stability from the regular walking to the first slip trial, but postslip stability between the two trials decreased significantly [significant event x trial interaction: F(1,28) = 124.49, P < 0.001] for both sessions [significant trial x session interaction: F(1,28) = 1.97, P > 0.10] with stability at liftoff (postslip) being greater than stability at touchdown (preslip) during the regular trial (P < 0.001), but postslip stability being significantly lower than preslip stability for the first slip trial (P < 0.001) for both sessions [session x event interaction: F(1,28) = 0.64, P > 0.10; Fig. 3, a and b].

Rapid re-acquisition with repeated slip exposure

As shown in Fig. 3, b–d, there was a significant increase in preslip stability for the acquisition session [F(4,5) = 7.69, P = 0.001] and the follow-up session [F(4,7) = 12.37, P < 0.001] with repeated slip exposure. A significant increase in preslip stability for the acquisition session was found from slip trials 1–2 (P = 0.02) and 2–3 (P = 0.04) compared with the majority of the re-acquisition completed in the second trial (P = 0.01) for the follow-up session, with no change thereafter (P > 0.10). The subjects had significantly greater preslip stability on the second slip trial of the follow-up session than they had during the acquisition session (P = 0.04).

Trial-to-trial increase in preslip stability for both sessions was achieved by an anterior shift in the XCOM/BOS [main effect trial: F(4,48) = 25.46, P < 0. 001, no significant session x trial interaction, P > 0.10] from the first to the second (P < 0.001) and second to the third slip trial (P = 0.02). The XCOM/BOS was, however, significantly anterior on the repeated slips for the follow-up session [main effect session: F(1,12) = 6.10, P = 0.03; Figs. 4 and 5A]. There was no change in COM/BOS (gait velocity) between trials and sessions (main effect: P > 0.10 for both).

The trial-to-trial increase in postslip stability with repeated slip exposure between the two sessions resembled the preslip change pattern with a significant increase in postslip stability for the acquisition session [F(4,5) = 15.18, P < 0.001] and the follow-up session [F(4,7) = 19.89, P < 0.001] with repeated slip exposure. In the acquisition session, these subjects required two repeated slips to continuously improve postonset stability (P < 0.01 for increase from trial 1 to 2 and 2 to 3). In the follow-up session, these same subjects only needed one repeated slip for the re-acquisition (P < 0.001 for increase from trial 1 to 2). Similar to the preslip stability, the subjects had a significantly greater postslip stability on the second slip trial of the follow-up session as compared with the acquisition session (P < 0.01, Fig. 3c).

There was a significantly greater increase in postslip stability as compared with preslip stability from the first to the second slip [event by trial interaction: F(1,28) = 35.75, P < 0.001] and second to the third slip [event x trial interaction: F(1,28) = 11.53, P = 0.002, Fig. 3, c and d], during the acquisition session but not thereafter (significant session x event interaction, P > 0.10, Fig. 3, d and e). The re-acquisition in postslip stability for the follow-up session on the second slip trial after exposure to the first slip was so much greater than the original acquisition session [session x event interaction, F(1,28) = 4.39, P = 0.04], such that there were no further improvements thereafter. A similar increased improvement in postslip stability had developed by the third trial in the initial acquisition session [session x event interaction: F(1,28) = 4.85, P = 0.04, Fig. 3d).

The mechanisms for improving stability at liftoff (postslip) were similar for both sessions; these were achieved by anteriorly shifting the XCOM/BOS [main effect trial: F(4,48) = 24.60, P < 0. 001; session x trial interaction: P > 0.10] and increasing the COM/BOS [main effect trial: F(4,48) = 29.52, P < 0. 001]. Nevertheless, the XCOM/BOS was significantly more anterior in the follow-up session than it was in the acquisition session [main effect session: F(1,12) = 6.10, P = 0.03]. For both the sessions, on the first slip trial, the XCOM/BOS was significantly posterior compared with the regular walking trial due to the unexpected nature of the slip (P < 0.001), followed by a significant anterior shift in the second (P < 0.001) and the third slips (P = 0.002, Fig. 6A). The COM/BOS exhibited no significant main effect of session with repeated slips [F(1,12) = 1.40, P > 0.10], but it had a session x trial interaction (P = 0.07). The COM/BOS was traveling significantly slower on the first slip trial compared with regular walking due to the slip (P < 0.001), for both sessions but was followed by a significant increase in the second (P < 0.001) and the third slips (P = 0.01). The increase in COM/BOS from trial 2 to 3 was significantly greater in the follow-up session than it was in the acquisition session (session x trial interaction: P < 0.05, Fig. 6B). Figure 5B shows that by the second slip trial of the follow-up session, the mean of the COM state relative to the BOS, at postslip liftoff, lies above the computationally predicted stability threshold for backward balance loss. This shift, however, occurs by the third slip trial in the acquisition session. It must also be noted that although preslip adaptive change in COM state stability was characterized by an anterior shift toward the backward balance loss threshold, involving only a position change relative to the BOS (Fig. 5A), postslip adaptive changes were associated with an anterior and upward shift, i.e., with increase in relative velocity as well (Fig. 5B).



View larger version (18K):
[in this window]
[in a new window]
  		FIG. 6. Means ± SD of COM state with respect to the BOS (XCOM/BOS, COM/BOS) at postslip liftoff of the contralateral limb from the acquisition ({circ}) and follow-up sessions ({bullet}) for the regular walking and 1st to 4th slip trials (S1–S4). Significant between session differences, and adaptive changes with respect to preceding trials are indicated by #, and *, respectively (P < 0.05).

 
With repeated slip exposure, there was a significant reduction in postslip BOS displacement and velocity for the acquisition session with a significant reduction in displacement and velocity during the second slip (P = 0.044 and P = 0.037, respectively) and in velocity only during the third slip (P = 0.02, Fig. 7, A and B). In comparison, in the follow-up session, there was a significant reduction in BOS displacement and velocity only in the second slip (P = 0.014 and P = 0.006, respectively), with its displacement and velocity significantly lower than that in the acquisition session (P = 0.045 and P = 0.01, respectively; Fig. 7, A and B).



View larger version (12K):
[in this window]
[in a new window]
  		FIG. 7. Means ± SD of the state of the BOS (slipping foot heel displacement, XBOS, and its velocity, BOS) at liftoff of the contralateral limb from the acquisition ({circ}) and follow-up session ({bullet}) for the regular walking and 1st to 3rd slip trials (S1–S3). Significant between session differences, and adaptive changes with respect to preceding trials are indicated by #, and *, respectively (P < 0.05).

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 
Our results indicated that a single acquisition session could not produce a sustainable long-term effect that could reduce occurrences of backward balance loss in the initial slip after 12 mo. Improvement was observed in preslip gait stability during the first slip of follow-up session, but it was apparently insufficient as all subjects exhibited a loss of balance on the follow-up session similar to that seen on the first slip of the acquisition session. Nevertheless, a learning effect was most noticeable on the second slip in which occurrence of backward balance loss was reduced by 87%, a significant improvement compared with the same trial in the acquisition session.

Retention of gait stability observed in the first slip

Consistent with our hypothesis, subjects showed retention of preslip stability, which remained the same in the first slip of the follow-up session as the last slip of the acquisition session, and both these were significantly improved compared with the first slip of the acquisition session. This improved gait stability was achieved by the same mechanism as in the acquisition session, namely by anteriorly shifting the XCOM/BOS. Contrary to our expectation, however, the subjects were unable to retain improvement in the follow-up session for the postslip stability that they had acquired at the end of the acquisition session. It appeared that this increase in preslip stability from an improved feedforward mechanism (Bhatt et al. 2005aGo) was insufficient to significantly alter reactive postslip stability and hence the final outcome. It is worth noting that there was a definite trace of the training effect; we observed that the subjects tended to be more stable postslip on the first slip exposure in the follow-up session than they had been in the acquisition session, with a probability level of postslip stability close to 0.10. This gain in stability may be considered a partial retention. Yet, it appears that the combination of the preslip retention and partial postslip retention could not alter the final outcome of the first slip in the follow-up session.

We postulate that the CNS was attempting to recall and execute the stable COM state representation acquired during the acquisition session, but only partially succeeded due to uncertainty when encountering the first slip and possibly the deterioration in motor memory over the 12-mo period. When the perturbation could not be fully compensated by preslip feedforward adjustment, the CNS must then have relied on feedback mechanisms after slip onset to improve stability. These findings suggest that while the retention of feedforward control affecting gait stability prior to slip onset was evident, the training effect on the reactive response, which could heavily rely on feedback control when perturbation was unexpected, might be more difficult to retain over a prolonged period. To the best of our knowledge, these findings illustrated for the first time the significance of the reactive response and the possible difficulties of retaining the training effect on feedback control. It is also possible that a single acquisition session that included five repetitions was not sufficiently intense to elicit sustainable changes, although it was apparently effective and successful in reducing the incidence of backward balance loss from 100 to 0% by the fifth trial.

Rapid re-acquisition with repeated slip exposure

Our results indicate that with training, the subjects were able to rapidly reduce balance loss incidence after the first slip. It was noteworthy that although there was a significant increase in stability from the first to the second slip during acquisition session, it was insufficient resulting in a high balance loss incidence (75%) on that trial. The significant reduction (by 87%) in balance loss on the second slip trial of the follow-up session could be explained by the significant and sufficient stability improvements during this session.

This achievement was made by improving both pre- and postslip stability, whereby the postslip improvement was significantly greater than the preslip improvements and likely resulted from improvements made in both feedforward and feedback control. Our findings support previous findings indicating that increasing one’s stability during the period of transition from double to single stance after slip onset is critical for prevention of balance loss (You et al. 2001Go; Bhatt, et al. 2005aGo). The control of balance can be characterized by the control of the COM state with respect to the BOS (i.e., XCOM/BOS and COM/BOS). During this period, the CNS controls stability by altering the state of the COM through the control of individual segment positions and their velocities, especially that of the upper body that carries the majority of body mass. The CNS may alternatively choose to closely regulate and adjust the BOS state, which is dictated by the lower limb kinematics. Previous findings have indicated that braking impulse exerted under the slipping limb is the single best predictor of BOS velocity (Bhatt et al. 2005aGo). The trial-to-trial adaptive changes in gait stability with repeated slip exposure for both sessions were achieved by both an anterior shift in XCOM/BOS and an increase in forward COM/BOS. However, the significantly improved stability on trial 2 of the follow-up session was achieved due to a significantly greater increase in COM/BOS achieved by a significantly lower BOS velocity. Such changes in BOS velocity after slip onset could in part be explained by feedforward adjustments in preslip gait pattern (e.g., foot angle, step length, heel velocity) (Lockhart et al. 2003Go; Bhatt et al. 2005aGo). that influence postslip kinetics (braking impulse) and the BOS perturbation intensity (Bhatt et al. 2005aGo).

Feedforward control would require prior experience and learning of the environmental constraints as well as the physiological properties and limitations of the system to be controlled. Interestingly, the experimental results can be fully accounted for if we assume that probability of balance loss and stability limits are predictable and that the feedforward stability control that the CNS employs must require an internal representation of the stability limits (Pai et al. 2003Go). An improved internal representation derived from repeated slip exposure could then improve preslip feedforward control, which would have significant impact on postslip reactive response. Successful feedforward control would likely reduce or even eliminate the need for feedback correction in the reactive response. Earlier evidence from perturbations studies also support this notion that the CNS builds or updates internal representation for the feedforward control of stability based on experience and anticipation. For example, when the perturbation type during stance changes from toes-up rotation to backward translation, subjects still show some postural strategies adapted to the previous perturbation (Horak and Nashner 1986Go). Similarly, when subjects anticipate a smaller perturbation but experience a larger one, they still show smaller magnitude responses (Horak et al. 1989Go; Timmann and Horak 1997Go). These responses could be preselected before the perturbation began due to formation of a "postural set" (Horak et al. 1989Go; Timmann and Horak 1997Go).

Insufficient retention in postslip stability on the first slip in the follow-up session may also indicate that the CNS may have altered the acquired internal representation of the stable gait pattern as the result of interference from performance of their preferred regular gait (Del Rey 1989Go), reinforced over the extensive 12-mo period during activities of daily living. Preferred gait patterns are often associated with minimum energy expenditure (Cavagna et al. 1976Go; Griffin et al. 2003Go), which may not be a primary concern when one is to avoid backward loss of balance. Therefore it is possible that the gait patterns acquired for resisting backward balance loss are not reinforced sufficiently during regular gait of daily living and continue to deteriorate over time. Our findings revealed, however, that once a stable postural response had been acquired, the refreshment of motor memory after a 12-mo interval would not take nearly as long as the original acquisition (Neumann and Ammons 1957Go).

It should be noted that the neurophysiological mechanisms underlying long-term retention for posture and gait adaptation are far from fully understood; moreover, current understanding of related phenomena at the behavior level is limited. Long-term retention has been well observed in the context of practice-related performance changes in skilled voluntary movements (Newell 1991Go). These studies have seldom examined retention at or 12 mo later with a single acquisition session. Systematic study of the effects of a single acquisition session at predetermined time intervals would be very costly to conduct. The present study charted the temporal boundary of such investigation, where a single acquisition session resulting in only partial longer-term retention in motor behavior was insufficient to yield significant results. The absence of long-term retention in reactive (postslip) stability on the first slip may be due to the low intensity of the acquisition session. This indicates that long-term structural changes could not be brought about after a mere five repetitions. Our acquisition session to improve gait stability was highly efficient and successful in temporarily reducing incidence of backward balance loss. This success is attributable, at least in part, to the potentially threatening nature of the perturbation employed. In fact, severe injury or even death is the most serious consequence of loss of balance and resulting falls among older adults (Cummings et al. 1990Go; Stalenhoef et al. 1999Go). Slip during gait is one of the leading contributors for these causes (Luukinen et al. 2000Go; Michelson et al. 1995Go).

In summary, the present study revealed that gait stability improvements acquired in a single acquisition session can be retained in varying degrees over a period of 12 mo, though no degree of retention was sufficient to alter the outcome of balance control when slip occurred for the first time. These improvements, however, could be more quickly refreshed than they were acquired. Based on the findings of the present study, our conclusion is that future investigations should focus on the effect of retention shorter than the 12-mo interval and possibly also on the effectiveness of stimulus intensity. Studies that take into account stimulus intensity should include the determination of the minimal number of repetitions and acquisition sessions required to produce a retainable effect in behavior.


   GRANTS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 GRANTS
ACKNOWLEDGMENTS
 REFERENCES
 
This work was supported by National Institute on Aging Grant 2R01 AG-16727.


   ACKNOWLEDGMENTS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
REFERENCES
 
The authors thank J. Wening and D. Espy for assisting in data collection and processing, Dr. E. Wang for statistical consultation, and Dr. James Girsch and K. Breving for editing the text.


   FOOTNOTES
 
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Address for reprint requests and other correspondence: Y.-C. Pai, Dept. of Physical Therapy, University of Illinois at Chicago, 1919 West Taylor St., Room 426 (M/C 898), Chicago, Illinois 60612 (E-mail: cpai{at}uic.edu)


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 
Adkin AL, Frank JS, Carpenter MG, and Peysar GW. Postural control is scaled to level of postural threat. Gait Posture 12: 87–93, 2000.[CrossRef][ISI][Medline]

Andres RO, Eng T, and O’Connor D. Sagittal plane kinematics of walking onto a slippery surface and gait adaptations from continuous walking on slippery surfaces. Slipping, Tripping, and Falling Accident Conference, London, 1991.

Baker SP and Harvey AH. Fall injuries in the elderly. Clin Geriat Med 1: 501–512, 1985.[Medline]

Bentley TA and Haslam RA. Slip, trip and fall accidents occurring during the delivery of mail. Ergonomics 41: 1859–1872, 1998.[Medline]

Bhatt T, Wening JD, and Pai Y-C. Adaptive control of gait stability in reducing slip-related backward loss of balance. Exp Brain Res In press 2005a.

Bhatt T, Wening JD, and Pai Y-C. Influence of gait speed on stability: recover from anterior slips and compensatory stepping. Gait Posture 21: 146–156, 2005b.[CrossRef][ISI][Medline]

Brashers-Krug T, Shadmehr R, and Bizzi E. Consolidation in human motor memory. Nature 382: 252–255, 1996.[CrossRef][Medline]

Carpenter MG, Frank JS, Silcher CP, and Peysar GW. The influence of postural threat on the control of upright stance. Exp Brain Res 138: 210–218, 2001.[CrossRef][ISI][Medline]

Cavagna GA, Thys H, and Zamboni A. The sources of external work in level walking and running. J Physiol 262: 639–657, 1976.[Abstract/Free Full Text]

Cummings SR, Rubin SM, and Black D. The future of hip fractures in the United States. Numbers, costs, and potential effects of postmenopausal estrogen. Clin Orthopaed Related Res 252: 163–166, 1990.

de Leva P. Adjustments to Zatsiorsky-Seluyanov’s segment inertia parameters. J Biomechan 29: 1223–1230, 1996.[CrossRef][ISI][Medline]

Del Rey P. Training and contextual interference effects on memory and transfer. Res Q Exercise Sport 60: 342–347, 1989.

Gao C and Abeysekera J. A systems perspective of slip and fall accidents on icy and snowy surfaces. Ergonomics 47: 573–598, 2004.[Medline]

Griffin TM, Roberts TJ, and Kram R. Metabolic cost of generating muscular force in human walking: insights from load-carrying and speed experiments. J Appl Physiol 95: 172–183, 2003.[Abstract/Free Full Text]

Horak FB, Diener HC, and Nashner LM. Influence of central set on human postural responses. J Neurophysiol 62: 841–853, 1989.[Abstract/Free Full Text]

Horak FB, Henry SM, and Shumway-Cook A. Postural perturbations: new insights for treatment of balance disorders. Phys Ther 77: 517–533, 1997.[Abstract/Free Full Text]

Horak FB and Nashner LM. Central programming of postural movements: adaptation to altered support-surface configurations. J Neurophysiol 55: 1369–1381, 1986.[Abstract/Free Full Text]

Kandel ER. The molecular biology of memory storage: a dialogue between genes and synapses. Science 294: 1030–1038, 2001.[Abstract/Free Full Text]

Keshner EA, Allum JH, and Pfaltz CR. Postural coactivation and adaptation in the sway stabilizing responses of normals and patients with bilateral vestibular deficit. Exp Brain Res 69: 77–92, 1987.[CrossRef][ISI][Medline]

Kim JJ and Fanselow MS. Modality-specific retrograde amnesia of fear. Science 256: 675–677, 1992.[Abstract/Free Full Text]

Lockhart TE, Woldstad JC, and Smith JL. Effects of age-related gait changes on the biomechanics of slips and falls. Ergonomics 46: 1136–1160, 2003.[Medline]

Luukinen H, Herala M, Koski K, Honkanen R, Laippala P, and Kivela SL. Fracture risk associated with a fall according to type of fall among the elderly. Osteoporosis Int 11: 631–634, 2000.[CrossRef][ISI][Medline]

Marigold DS and Patla AE. Strategies for dynamic stability during locomotion on a slippery surface: effects of prior experience and knowledge. J Neurophysiol 88: 339–353, 2002.[Abstract/Free Full Text]

McIlroy WE and Maki BE. Adaptive changes to compensatory stepping responses. Gait Posture 3: 43–50, 1995.

McMahon TA. Muscles, Reflexes, and Locomotion. Princeton, NJ: Princeton University Press, 1984.

Michelson JD, Myers A, Jinnah R, Cox Q, and Van Natta M. Epidemiology of hip fractures among the elderly. Risk factors for fracture type. Clin Orthopaed Related Res 311: 129–135, 1995.

Neumann E and Ammons RB. Acquisition and long-term retention of a simple serial perceptual-motor skill. J Exp Psychol 53: 159–161, 1957.[CrossRef][ISI][Medline]

Newell KM. Motor skill acquisition. Annu Rev Psychol 42: 213–237, 1991.[CrossRef][ISI][Medline]

Owings TM, Pavol MJ, and Grabiner MD. Mechanisms of failed recovery following postural perturbations on a motorized treadmill mimic those associated with an actual forward trip. Clin Biomech 16: 813–819, 2001.[CrossRef][Medline]

Pai Y-C and Iqbal K. Simulated movement termination for balance recovery: Can movement strategies be sought to maintain stability even in the presence of slipping or forced sliding? J Biomech 32: 779–786, 1999.[CrossRef][ISI][Medline]

Pai Y-C, Wening JD, Runtz EF, Iqbal K, and Pavol MJ. Role of feedforward control of movement stability in reducing slip-related balance loss and falls among older adults. J Neurophysiol 90: 755–762, 2003.[Abstract/Free Full Text]

Pavol MJ and Pai YC. Feedforward adaptations are used to compensate for a potential loss of balance. Exp Brain Res 145: 528–538, 2002.[CrossRef][ISI][Medline]

Pavol MJ, Runtz EF, Edwards BJ, and Pai Y-C. Age influences the outcome of a slipping perturbation during initial but not repeated exposures. J Gerontol A Biol Sci Med Sci 57: M496–503, 2002.[Abstract/Free Full Text]

Rand MK, Wunderlich DA, Martin PE, Stelmach GE, and Bloedel JR. Adaptive changes in responses to repeated locomotor perturbations in cerebellar patients. Exp Brain Res 122: 31–43, 1998.[CrossRef][ISI][Medline]

Redfern MS, Cham R, Gielo-Perczak K, Grönqvist R, Hirvonen M, Lanshammar H, Marpet M, Pai Y-C, and Powers C. Biomechanics of slips. Ergonomics 44: 1138–1166, 2001.[Medline]

Reynolds RF and Bronstein AM. The Moving Platform aftereffect: limited generalization of a locomotor adaptation. J Neurophysiol 91: 92–100, 2003.

Sacchetti B, Scelfo B, Tempia F, and Strata P. Long-term synaptic changes induced in the cerebellar cortex by fear conditioning. Neuron 42: 973–982, 2004.[CrossRef][ISI][Medline]

Schwabe A, Drepper J, Maschke M, Diener HC, and Timmann D. The role of the human cerebellum in short- and long-term habituation of postural responses. Gait Posture 19: 16–23, 2004.[CrossRef][ISI][Medline]

Stalenhoef PA, Diederiks JP, de Witte LP, Schiricke KH, and Crebolder HF. Impact of gait problems and falls on functioning in independent living persons of 55 years and over: a community survey. Patient Ed Counsel 36: 23–31, 1999.

Timmann D, and Horak FB. Prediction and set-dependent scaling of early postural responses in cerebellar patients. Brain 120: 327–337, 1997.[Abstract/Free Full Text]

Tjernstrom F, Fransson PA, Hafstrom A, and Magnusson M. Adaptation of postural control to perturbations–a process that initiates long-term motor memory. Gait Posture 15: 75–82, 2002.[CrossRef][ISI][Medline]

You J, Chou Y, Lin C, and Su F. Effect of slip on movement of body center of mass relative to base of support. Clin Biomech 16: 167–173, 2001.




This article has been cited by other articles:


Home page
 J. Neurophysiol.Home page
T. Bhatt and Y.-C. Pai
Can Observational Training Substitute Motor Training in Preventing Backward Balance Loss After an Unexpected Slip During Walking?
J Neurophysiol, February 1, 2008; 99(2): 843 - 852.
[Abstract] [Full Text] [PDF]

Home page
 ptjournalHome page
Y.-C. Pai and T. S Bhatt
Repeated-Slip Training: An Emerging Paradigm for Prevention of Slip-Related Falls Among Older Adults
Physical Therapy, November 1, 2007; 87(11): 1478 - 1491.
[Abstract] [Full Text] [PDF]

Home page
 J. Neurophysiol.Home page
T. Bhatt, E. Wang, and Y.-C. Pai
Retention of Adaptive Control Over Varying Intervals: Prevention of Slip- Induced Backward Balance Loss During Gait
J Neurophysiol, May 1, 2006; 95(5): 2913 - 2922.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
94/3/1971    most recent
00266.2005v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (6)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bhatt, T.
Right arrow Articles by Pai, Y.-C.
Right arrow Search for Related Content
PubMed