- Email: [email protected]
Medio-lateral balance adjustments preceding reflexive limb withdrawal are modified by postural demands
Brain Research 914 (2001) 100–105 www.elsevier.com / locate / bres
Medio-lateral balance adjustments preceding reflexive limb withdrawal are modified by postural demands a, b c d Leah R. Bent *, Jim R. Potvin , John D. Brooke , William E. McIlroy a
School of Human Kinetics, Faculty of Education, University of British Columbia, 210 -6081 University Boulevard, Vancouver, British Columbia, Canada V6 T 1 Z1 b Department of Human Kinetics, University of Windsor, Windsor, Ontario, Canada c Department of Physical Therapy and Human Movement Sciences, Northwestern University Medical School, Chicago, Illinois, USA d Graduate Department of Rehabilitation Science, Department of Physical Therapy, University of Toronto, Toronto, Ontario, Canada Accepted 3 July 2001
Abstract We have recently observed medio-lateral balance adjustments (BA) preceding reflexive stepping elicited by noxious stimulation. While task specific modulation is evident for BA prior to voluntary leg movement, it is unclear whether rapid BA reactions (prior to ‘reflexive’ stepping) represent a generic response to evoked limb withdrawal or can be modified to suit task-conditions. This study was designed to establish whether the CNS is able to modify rapid onset latency BAs to match task conditions. Reflexive stepping was evoked by applying a noxious stimulus (50 ms stimulus train, 1 ms pulses, 300 Hz, 43perceptual threshold) to the plantar surface of the either the left or right foot. Task conditions were varied prior to stimulation by having subjects maintain one of three different static positions: (1) lean left (70% body weight (BW) on left), (2) neutral (50% BW both sides), (3) lean right (70% BW on right). BAs were denoted by centre-of-pressure (CoP) excursions towards the swing foot after the onset of noxious stimulation (average onset latency of 128 ms). There was a significant increase in frequency of occurrence and a significant increase in magnitude of CoP shift when the stimulation was applied to a loaded limb (leaning with 70% BW on the stimulated foot) as compared to an unloaded limb (30%BW). In addition, 78% of loaded trials featured steps taken with the unstimulated foot, which delayed removal of the stimulated foot. Collectively, the results indicate modifiability of the very rapid onset balance adjustments that precede the onset of limb withdrawal revealing complex control of balance exists even over very brief latencies. 2001 Elsevier Science B.V. All rights reserved. Theme: Motor systems and sensorimotor integration Topic: Control of posture and movement Keywords: Lower limb; Balance; Flexion withdrawal; Modulation; Anticipatory; Reflex
1. Introduction Voluntary movements are coupled with postural control mechanisms in order to reduce forthcoming destabilizing forces associated with the focal task [3]. These postural reactions elicited prior to voluntary movement, referred to as anticipatory postural adjustments (APAs), precede movements of both the lower [12,19,26,27], and upper limbs [3,4,22]. Such APAs are characteristically scaled in magnitude [6,9,15], duration [13,22] and direction [4,5] to *Corresponding author. Tel.: 11-604-822-4206; fax: 11-604-8226842. E-mail address: [email protected] (L.R. Bent).
appropriately oppose the postural disturbances created by the focal task. Recent work has demonstrated the novel observation of rapid onset balance adjustments (BA), in the medio-lateral direction, prior to reflexive withdrawal of the lower limb, evoked by noxious stimulation [20]. It is suggested that these BAs are similar to APAs in that they appear prior to the onset of primary limb movement and have the effect of stabilizing the individual prior to the instability created by unloading the withdrawing limb. What is not clear, is whether these rapid latency responses are generic unscaled reactions or if they are modifiable postural adjustments, similar to those APAs present preceding voluntary movement. The ability to scale rapid onset latency BAs prior to
0006-8993 / 01 / $ – see front matter 2001 Elsevier Science B.V. All rights reserved. PII: S0006-8993( 01 )02782-2
L.R. Bent et al. / Brain Research 914 (2001) 100 – 105
reflexive stepping has implications for the interpretation of the characteristics of compensatory stepping evoked by postural perturbations. In contrast to the tight coupling of the APA with voluntary stepping, the BA prior to compensatory stepping evoked by platform perturbations are inappropriately scaled or absent (McIlroy and Maki [18,19]). The present work is important in attempting to determine whether these absences arise due to limitations in control possibly related to the rapid onset latency. Previous studies have illustrated the ability to modify reflex responses to intense stimulation [7]. Such reflex modulation has been documented in humans, as reflex reversals during locomotion [8,30], phasic modulation over the movement cycle [2] as well as in modulation in quiet stance due to variations in limb loading [27]. However, it is not clear if the latter reflects changes associated with maintaining whole-body stability. The following experiment tests the hypothesis that these rapid onset latency BAs will in fact be modifiable as demonstrated by an increase in both magnitude and frequency of occurrence as the static load and associated postural demands on the stimulated limb increases.
2. Materials and methods Fourteen healthy subjects (six men, eight women) ranging in age from 19 to 39, of average weight 66.1611.5 kg, and average height 170.968.8 cm consented to participate in the study. All subjects indicated right foot as dominant (determined by limb preference for kicking a ball). Subjects were informed of the protocol but were not provided specific details of the purpose and the hypothesized outcome. The experiment was approved by the University of Guelph Human Ethics Committee. Subjects stood barefoot on a strain gauge force plate (Advanced Mechanical Technologies, Newton Mass, Model [OR6-5-1) with a standard foot position of 14 cm between the heels and an angle of 78 from the midline [16]. A noxious stimulus was applied during the study using two custom made aluminum plates (18328 cm) placed under and covering the whole sole of the right and left feet of the subject. Stimulation was delivered through the plates, which were wired to a Grass S88 stimulator. The bipolar stimulus consisted of a 50 ms train of 1 ms pulses, delivered at a frequency of 300 Hz [11,14,25]. The noxious stimulus was applied at a voltage approximating four times perceptual threshold (PT). The intensity of stimulation was determined from pilot data on a separate group of subjects ¨ to the to ensure subjects in the main study remained naıve stimulation until the time of data collection. The determination of PT was conducted for individual subjects, at each of the loading conditions, at the end of the experiment. It is noteworthy that there were no significant differences in PT across the loading conditions. Markings were placed on the force plate to ensure
101
consistent foot placement prior to each trial. To prevent a step taken to the side, which may have reduced the likelihood of a BA, two 60340 cm dummy plates (which subjects were told were stimulated) were positioned to the immediate right and left side of the subject. This was intended to discourage a lateral stepping strategy, which could have minimized reliance on BA responses. The study tested whether the initial centre of mass (CoM) location could affect modulation of the rapidly generated BAs. Six conditions were tested, using three levels of static lateral leaning (30, 50, 70% of body weight (BW)) on either leg. Subjects were instructed to shift their weight to one of these three locations. They monitored vertical loading via an oscilloscope to meet the appropriate target. The target line for weight shift on the oscilloscope was placed at positions where the subject would have approximately 70% of body weight on one leg (loaded) and concurrently 30% on the other leg (unloaded). This was the same for both right and left directions. The final condition was a left or right foot stimulus while standing in a neutral (50 / 50) position, with the weight equally distributed between the two feet. The order of the task conditions and side of stimulation was randomized. The noxious stimulus was delivered at varied times (ranging between 2 and 5 s) after the subject had reached the static location, to avoid anticipation. Data from the force plate were collected and digitally converted (National Instruments NB M10-16XE) at a sampling frequency of 200 Hz. The data were dual low pass filtered using a second order Butterworth filter with a 10 Hz cut off. Foot off and foot contact times were collected in addition to the forceplate data. The CoP parameters of interest were the presence or absence of medio-lateral (ML) BAs, onset latency, and peak magnitude. Previous studies have determined that a M–L CoP shift towards the swing limb in voluntary movement represents an increase in force to move the CoM onto the stance limb [17]. The criteria used in the present study to establish the occurrence of a BA was a shift of the CoP towards the swing foot, after stimulation onset. To denote the presence of a BA the CoP had to move towards the swing foot with a magnitude which surpassed the threshold band of 4 mm (calculation based on CoP sway during quiet stance) [21]. Such lateral excursions of the CoP are referred to in the present study as a BAs, since simple limb withdrawal or crossed extension would be characterized by movement of the CoP towards the stance limb. Onset latency was determined as the time from stimulus onset until the time when the CoP crossed the 4 mm threshold boundary. Magnitude was quantified as the peak ML shift of the CoP beyond the 4 mm threshold prior to unloading of the swing limb. The data were rejected if the latency exceeded the 95% confidence interval around the average onset latency for voluntary stimulation trials (1.53PT), determined as 183 ms to 217 ms in a previous study [20]. As a result, noxious
102
L.R. Bent et al. / Brain Research 914 (2001) 100 – 105
trials, in which the latency exceeded 183 ms, were excluded. Response frequency compared between task conditions was evaluated using Chi square analyses. One way repeated measures ANOVAs, blocked on subjects, with three replicates, were run to test the main effect of an increase in static loading on the stimulated limb for the dependent variable onset latency. Multiple comparisons were made using the Tukey’s honestly significant difference to test the experimental hypotheses. To address concerns for non normal distribution of data, the analysis of magnitude featured non-parametric Wilcoxon and Kruskall Wallis tests for two and three level comparisons respectively. ´ post-hoc test. Post-hoc analyses were made using Scheffes For all tests the significance level was set at P,0.05.
3. Results A total of 85 stepping responses to the stimulation were recorded during the study. Five of the fourteen subjects tested accounted for 77 of these 85 steps. These subjects are referred to as steppers in the present study. A separate group of five subjects accounted for 87 of the non-stepping responses (n5127). These five subjects are referred to as non-steppers. For the planned within-subject analyses, only the step data from the five steppers (n577) and the non-step responses from the non-steppers (n587) were used. Comparisons between step and no step data were made between these two groups. In addition, there were no significant differences found between the left and right legs for step frequency or for APA occurrence (P.0.05). Consequently data from left and right limbs were pooled for subsequent analysis. A third response, namely a toes up response (without foot removal from the ground), was observed in some subjects (n527). However, on account that the study focused on balance adjustments preceding step responses, and the toes up response could not be classified clearly as a step or no step response they were not analysed further. When considering individually the group of steppers, the frequency of occurrence of the CoP shift towards the swing foot (BA) was 34% (26 / 77). BA occurrence in the non steppers was 54% (46 / 87). As hypothesized, when comparing the frequency of BAs over conditions for step data, the occurrence was highest in the loaded condition as compared to the neutral or unloaded conditions; 61, 20, and 24% respectively (Fig. 1). This large increase in frequency of BAs in the loaded trials accounted for the significant differences between task conditions ( x 2 510.88, P50.004). This task specific influence on occurrence of BAs was also reflected in the task differences in initial direction of the CoP excursions measured in trials from non-steppers ( x 2 512.01, P50.002)(Fig. 1). Interesting to note is that although the BA frequency appears to be greater in non-steppers, it is the magnitude of
Fig. 1. Frequency of balance adjustment occurrence. Frequency of occurrence (%) of balance adjustments (BAs) for each loading condition in separated step and no-step data. As the postural demands on the limb increase, so does the occurrence of BAs. Differences between tasks were statistically significant (P,0.05).
the BAs that may reflect their effectiveness to shift CoM. The magnitude of M–L CoP shift towards the swing foot in step data was observed to increase significantly from the unloaded to neutral to the loaded condition; 1.1, 3.2, 5.4 cm ( x 2 513.15, P50.001)(Fig. 2A). In contrast, inspection of the magnitude of BAs exhibited by non-steppers revealed they were small and not different from the unloaded, to neutral to the loaded trials; 0.8, 1.6, 1.3 cm respectively (Fig. 2A). While most stepping responses to the noxious stimulation involved unloading of the stimulated foot, a unique and unexpected result was the removal of the unstimulated foot from the plates. All 5 stepping subjects exhibited trials in which steps were taken with the unstimulated foot, representing 27% (21 / 77) of the stepping responses. The majority of these steps with the unstimulated foot occurred when the stimulated foot was in the most loaded condition (78%, x 2 543.044, P,0.001). As a result of this occurrence, in order to address the primary control related to the rapid removal of the foot from the noxious stimulus, trials were isolated in which steps were taken only with the stimulated foot. Importantly this group of data, represented by trials only from the unloaded and neutral conditions, demonstrated an increase in magnitude between the two statically loaded positions. As the percent of BW increased on the stimulated limb from the unloaded to the neutral position, the magnitude of the CoP excursions towards the swing foot increased significantly from 0.36 cm to 2.1 cm respectively ( x 2 5 5.21, P50.02) (Fig. 2B). Modulations were also observed in onset latency of the BAs in accordance with the changing postural requirements. In the step data, increased static load on the stimulated leg had a significant effect on the onset latency (F2,8 59.33, P50.008). The step trial onset averages were 109, 128, and 125 ms for the unloaded, neutral, and loaded
L.R. Bent et al. / Brain Research 914 (2001) 100 – 105
Fig. 2. Magnitude of balance adjustments. Average magnitude of balance adjustment (BA), measured from baseline to the peak medio-lateral (ML) centre-of-pressure (CoP) shift, along with the associated standard errors. (A) Magnitude compared over the weight conditions (unloaded, neutral and loaded). Data is shown for stepping and non-stepping data. M–L CoP magnitude increases significantly for stepping trials only, as the weight on the stimulated limb increases (P,0.05). (B) Magnitude of BAs in responses where the stimulated foot was lifted. Only the neutral and unloaded trials are included because 78% of steps with the unstimulated foot occurred in the loaded condition. Unloaded versus neutral conditions featured a significant increase in BA magnitude, demonstrating a rapid postural response that is modulated according to the task (P,0.05).
conditions respectively. The significant difference found was accounted for by the rapid onset of the unloaded trials relative to the other two conditions. A significant increase was also found when focusing on trials featuring only stimulated foot removal in the unloaded and neutral conditions; 110 ms and 130 ms respectively (F1,4 544.44, P50.003). In opposition to this trend, the onset latency for non step data was not significantly different between any of the unloaded, neutral or loaded conditions P,0.05).
4. Discussion The results confirm previous observations of the presence of rapid onset latency BAs preceding noxious with-
103
drawal responses [20]. Importantly, the present study supports the hypothesis that it is possible to modify the characteristics of very rapid BAs that precede the withdrawal of the lower limb to noxious stimulation. This is supported by the observation that magnitude of the CoP excursion increases significantly as the static load on the stimulated limb is increased. Additional support also comes from the observation that the frequency of BA occurrence increases as the postural demands associated with the task increase from the unloaded to the loaded condition, as the loading on the stimulated limb increases. Also, the onset latency of the response increases significantly from the unloaded to the neutral condition. Finally one might also view the unique observation of frequent removal of the unstimulated foot in the loaded condition as support for the view that the CNS makes rapid latency BAs that are modulated for the purpose of preserving stability when executing rapid withdrawal responses. The rapid onset latency of 12864.6 ms (S.E.) observed in the neutral stance trials of the present experiment supports previous latencies of 12964.6 ms (S.E.) reported in the literature [20]. It is worth emphasizing that these response latencies reflect changes in ground reaction force which would be preceded by underlying EMG response onset latencies occurring 30–60 ms prior to measured force changes [23]. Interestingly, the observation in this study of onset latency modifiability with an increased static load on the stimulated limb, was not anticipated. There are several possible reasons for the presence of this modulation. One may be related to changes in ongoing sensory discharge and motoneuron activation. Previous studies have revealed a decrease in latency of a reflexive response with modest muscle contraction [1,28,29]. Another possibility for this timing change measured from ground reaction forces could be attributed to alterations in the electromechanical delay due to variations in tonic loading. The presence of BAs, in addition to the strategy observed of unstimulated foot removal, appears to support the view that these reactions exist to deal with the objective of reacting to the noxious stimulus while predictively minimizing whole-body instability. Clearly, the mere evidence of a rapid foot withdrawal with a preceding BA supports such an objective. This implementation of a predictive balance control strategy prior to the initiation of a reactive stepping reaction has been observed in studies exploring compensatory stepping reactions [18]. It is noteworthy that the present reactions occur at more rapid latencies and the amplitudes of the balance adjustments are much larger than seen in such compensatory reactions. In the present study, responses that include a BA, lead to a trade-off between the need to rapidly withdrawal the foot and the objective of maintaining stability. This seems to support the suggestion that the hierarchy of control is prioritized to address the maintenance of balance as an issue of primary importance even though this prolongs exposure to the noxious stimulus (longer time to foot-off).
104
L.R. Bent et al. / Brain Research 914 (2001) 100 – 105
The strategy of unloading the non-stimulated, when the stimulated foot is prior loaded, may be a specific example of this trade-off between balance and time to withdrawal the foot. This reaction may be an intentional strategy to help to move the COM towards the unstimulated foot under the influence of gravity or may be a result of a large medio-lateral balance adjustment. Despite the numerous possible responses; lifting the stimulated or unstimulated foot with or without a BA, it is apparent that the CNS may have several options for successfully achieving the often competing objectives of rapid limb withdrawal and maintaining whole-body stability. The present observation of rapid latency reactions (scaled to suit the task conditions) coincides with numerous studies revealing task-specific modulation of reflex activity [10,24,27]. The findings of the present study clearly highlight the ability to rapidly generate and modify postural adjustments to enable fast and appropriate responses to instabilities that may quickly arise in an environment that is consistently changing. The results have implications to the understanding of the coupling of BAs and rapid compensatory stepping reactions, specifically that the failure to appropriately scale BAs was not attributable to the rapid onset latency as once thought. There are also parallels between the presently observed balance adjustments and anticipatory postural adjustments (APAs) preceding voluntary movement. The one distinguishing characteristic is the rapid latency of the present reactions that precede withdrawal reactions, which are often considered to be spinally mediated. It is possible that BAs and APAs are regulated by common neural substrates, which might raise questions about the view of supraspinal control of APAs and their invariable coupling to voluntary movement.
Acknowledgements This work was supported by funding from the Natural Science and Engineering Council of Canada(WEM).
References [1] A.M. Aniss, S.C. Gandevia, D. Burke, Reflex responses in active muscles elicited by stimulation of low-threshold afferents from the human foot, J. Neurophysiol. 67 (1992) 1375–1384. [2] M. Belanger, A.E. Patla, Phase-dependent compensatory responses to perturbation applied during walking in humans, J. Motor Behav. 19 (1987) 434–453. [3] S. Bouisset, M. Zattara, Biomechanical study of the programming of anticipatory postural adjustments associated with voluntary movement, J. Biomech. 20 (1987) 735–742. [4] J.E. Brown, J.S. Frank, Influence of event anticipation on postural actions accompanying voluntary movement, Exp. Brain Res. 67 (1987) 645–650.
[5] P.J. Cordo, L.M. Nashner, Properties of postural adjustments with rapid arm movements, J. Neurophysiol. 47 (1982) 287–302. [6] M.C. Do, B. Bussel, Y. Breniere, Influence of plantar cutaneous afferents on early compensatory reactions to forward fall, Exp. Brain Res. 79 (1990) 319–324. [7] J. Duysens, G.E. Loeb, Modulation of ipsi-and contralateral reflex responses in unrestrained walking cats, J. Neurophysiol. 44 (1980) 1024–1037. [8] J.J. Eng, D.A. Winter, A.E. Patla, Strategies for recovery from a trip in early and late swing during human walking, Exp. Brain Res. 102 (1994) 339–349. [9] R. Forget, Y. Lamarre, Anticipatory postural adjustments in the absence of normal peripheral feedback, Brain Res. 508 (1990) 176–179. [10] H. Forssberg, Stumbling corrective reaction: a phase-dependent compensatory reaction during locomotion, J. Neurophysiol. 42 (1979) 936–953. [11] K.E. Hagbarth, B.L. Finer, The plasticity of human withdrawal reflexes to noxious skin stimuli in lower limbs, Prog. Brain Res. 1 (1963) 65–78. [12] R. Herman, T. Cook, B. Cozzens, W. Freedman, Control of postural reactions in man: the initiation of gait, in: R.B. Stein, K.G. Pearson, R.S. Smith, J.B. Redford (Eds.), Control of Posture and Gait, Plenum, New York, 1973, pp. 363–388. [13] T. Kasai, K. Kawai, Quantitative EMG analysis of anticipatory postural adjustments of voluntary contraction of leg muscles in standing man, Electroenceph. Clin. Neurophys. 93 (1994) 184–187. [14] E. Kugelberg, K. Eklund, L. Grimby, An electromyographic study of the nociceptive reflexes of the lower limb. Mechanisms of the plantar response, Brain 83 (1960) 394–410. [15] W.A. Lee, C.F. Michaels, Y.C. Pai, The organization of torque and EMG activity during bilateral handle pulls by standing humans, Exp. Brain Res. 82 (1990) 304–314. [16] B.E. Maki, W.E. McIlroy, The role of limb movements in maintaining upright stance: The ‘change-in-support’ strategy, Phys. Therapy. 77 (1997) 488–507. [17] J. Massion, Movement, posture and equilibrium: Interaction and coordination, Prog. Neurobiol. 38 (1992) 35–56. [18] W.E. McIlroy, B.E. Maki, Do anticipatory postural adjustments precede compensatory stepping reactions evoked by perturbation?, Neurosci. Lett. 164 (1993) 199–202. [19] W.E. McIlroy, B.E. Maki, Adaptive changes to compensatory stepping responses, Gait Posture 3 (1995) 43–50. [20] W.E. McIlroy, L.R. Bent, J.R. Potvin, J.D. Brooke, B.E. Maki, Preparatory balance adjustments precede withdrawal response to noxious stimulation in standing humans, Neurosci. Lett. 267 (1999) 197–200. [21] W.E. McIlroy, B.E. Maki, The control of lateral stability during rapid stepping reactions evoked by antero-posterior perturbation: does anticipatory control play a role?, Gait Posture 9 (1999) 190– 198. [22] A. Nardone, M. Schieppati, Postural adjustments associated with voluntary contraction of leg muscles in standing man, Exp. Brain Res. 69 (1988) 469–480. [23] R.W. Norman, P.V. Komi, Electromechanical delay in skeletal muscle under normal movement conditions, Acta Physiol. Scand. 106 (1979) 241–248. [24] K.G. Pearson, J.E. Misiaszek, K. Fouad, Enhancement and resetting of locomotor activity by muscle afferents, Ann. NY Acad. Sci. 860 (1998) 203–215. [25] A. Roby-Brami, B. Bussel, Long latency spinal reflex in man after flexor reflex afferent stimulation, Brain 110 (1987) 707–725. [26] M.W. Rogers, Y.C. Pai, Dynamic transitions in stance support accompanying leg flexion movements in man, Exp. Brain Res. 81 (1990) 398–402. [27] A. Rossi, B. Decchi, Flexibility of lower limb reflex responses to painful cutaneous stimulation in standing humans: evidence of load-dependent modulation, J. Physiol. 481 (1994) 521–532.
L.R. Bent et al. / Brain Research 914 (2001) 100 – 105 [28] A. Rossi, A. Zalaffi, B. Decchi, Interaction of nociceptive and non-nociceptive cutaneous afferents from the foot sole in common reflex pathways to tibialis anterior motoneurones in humans, Brain Res. 714 (1996) 76–86.
105
[29] B.T. Shahani, R.R. Young Human flexor reflexes, J. Neurol. Neurosurg. Psych. 34 (1971) 616–627. [30] J.F. Yang, R.B. Stein, Phase-dependent reflex reversal in human leg muscles during walking, J. Neurophysiol. 63 (1990) 1109–1117.
Medio-lateral balance adjustments preceding reflexive limb withdrawal are modified by postural demands a, b c d Leah R. Bent *, Jim R. Potvin , John D. Brooke , William E. McIlroy a
School of Human Kinetics, Faculty of Education, University of British Columbia, 210 -6081 University Boulevard, Vancouver, British Columbia, Canada V6 T 1 Z1 b Department of Human Kinetics, University of Windsor, Windsor, Ontario, Canada c Department of Physical Therapy and Human Movement Sciences, Northwestern University Medical School, Chicago, Illinois, USA d Graduate Department of Rehabilitation Science, Department of Physical Therapy, University of Toronto, Toronto, Ontario, Canada Accepted 3 July 2001
Abstract We have recently observed medio-lateral balance adjustments (BA) preceding reflexive stepping elicited by noxious stimulation. While task specific modulation is evident for BA prior to voluntary leg movement, it is unclear whether rapid BA reactions (prior to ‘reflexive’ stepping) represent a generic response to evoked limb withdrawal or can be modified to suit task-conditions. This study was designed to establish whether the CNS is able to modify rapid onset latency BAs to match task conditions. Reflexive stepping was evoked by applying a noxious stimulus (50 ms stimulus train, 1 ms pulses, 300 Hz, 43perceptual threshold) to the plantar surface of the either the left or right foot. Task conditions were varied prior to stimulation by having subjects maintain one of three different static positions: (1) lean left (70% body weight (BW) on left), (2) neutral (50% BW both sides), (3) lean right (70% BW on right). BAs were denoted by centre-of-pressure (CoP) excursions towards the swing foot after the onset of noxious stimulation (average onset latency of 128 ms). There was a significant increase in frequency of occurrence and a significant increase in magnitude of CoP shift when the stimulation was applied to a loaded limb (leaning with 70% BW on the stimulated foot) as compared to an unloaded limb (30%BW). In addition, 78% of loaded trials featured steps taken with the unstimulated foot, which delayed removal of the stimulated foot. Collectively, the results indicate modifiability of the very rapid onset balance adjustments that precede the onset of limb withdrawal revealing complex control of balance exists even over very brief latencies. 2001 Elsevier Science B.V. All rights reserved. Theme: Motor systems and sensorimotor integration Topic: Control of posture and movement Keywords: Lower limb; Balance; Flexion withdrawal; Modulation; Anticipatory; Reflex
1. Introduction Voluntary movements are coupled with postural control mechanisms in order to reduce forthcoming destabilizing forces associated with the focal task [3]. These postural reactions elicited prior to voluntary movement, referred to as anticipatory postural adjustments (APAs), precede movements of both the lower [12,19,26,27], and upper limbs [3,4,22]. Such APAs are characteristically scaled in magnitude [6,9,15], duration [13,22] and direction [4,5] to *Corresponding author. Tel.: 11-604-822-4206; fax: 11-604-8226842. E-mail address: [email protected] (L.R. Bent).
appropriately oppose the postural disturbances created by the focal task. Recent work has demonstrated the novel observation of rapid onset balance adjustments (BA), in the medio-lateral direction, prior to reflexive withdrawal of the lower limb, evoked by noxious stimulation [20]. It is suggested that these BAs are similar to APAs in that they appear prior to the onset of primary limb movement and have the effect of stabilizing the individual prior to the instability created by unloading the withdrawing limb. What is not clear, is whether these rapid latency responses are generic unscaled reactions or if they are modifiable postural adjustments, similar to those APAs present preceding voluntary movement. The ability to scale rapid onset latency BAs prior to
0006-8993 / 01 / $ – see front matter 2001 Elsevier Science B.V. All rights reserved. PII: S0006-8993( 01 )02782-2
L.R. Bent et al. / Brain Research 914 (2001) 100 – 105
reflexive stepping has implications for the interpretation of the characteristics of compensatory stepping evoked by postural perturbations. In contrast to the tight coupling of the APA with voluntary stepping, the BA prior to compensatory stepping evoked by platform perturbations are inappropriately scaled or absent (McIlroy and Maki [18,19]). The present work is important in attempting to determine whether these absences arise due to limitations in control possibly related to the rapid onset latency. Previous studies have illustrated the ability to modify reflex responses to intense stimulation [7]. Such reflex modulation has been documented in humans, as reflex reversals during locomotion [8,30], phasic modulation over the movement cycle [2] as well as in modulation in quiet stance due to variations in limb loading [27]. However, it is not clear if the latter reflects changes associated with maintaining whole-body stability. The following experiment tests the hypothesis that these rapid onset latency BAs will in fact be modifiable as demonstrated by an increase in both magnitude and frequency of occurrence as the static load and associated postural demands on the stimulated limb increases.
2. Materials and methods Fourteen healthy subjects (six men, eight women) ranging in age from 19 to 39, of average weight 66.1611.5 kg, and average height 170.968.8 cm consented to participate in the study. All subjects indicated right foot as dominant (determined by limb preference for kicking a ball). Subjects were informed of the protocol but were not provided specific details of the purpose and the hypothesized outcome. The experiment was approved by the University of Guelph Human Ethics Committee. Subjects stood barefoot on a strain gauge force plate (Advanced Mechanical Technologies, Newton Mass, Model [OR6-5-1) with a standard foot position of 14 cm between the heels and an angle of 78 from the midline [16]. A noxious stimulus was applied during the study using two custom made aluminum plates (18328 cm) placed under and covering the whole sole of the right and left feet of the subject. Stimulation was delivered through the plates, which were wired to a Grass S88 stimulator. The bipolar stimulus consisted of a 50 ms train of 1 ms pulses, delivered at a frequency of 300 Hz [11,14,25]. The noxious stimulus was applied at a voltage approximating four times perceptual threshold (PT). The intensity of stimulation was determined from pilot data on a separate group of subjects ¨ to the to ensure subjects in the main study remained naıve stimulation until the time of data collection. The determination of PT was conducted for individual subjects, at each of the loading conditions, at the end of the experiment. It is noteworthy that there were no significant differences in PT across the loading conditions. Markings were placed on the force plate to ensure
101
consistent foot placement prior to each trial. To prevent a step taken to the side, which may have reduced the likelihood of a BA, two 60340 cm dummy plates (which subjects were told were stimulated) were positioned to the immediate right and left side of the subject. This was intended to discourage a lateral stepping strategy, which could have minimized reliance on BA responses. The study tested whether the initial centre of mass (CoM) location could affect modulation of the rapidly generated BAs. Six conditions were tested, using three levels of static lateral leaning (30, 50, 70% of body weight (BW)) on either leg. Subjects were instructed to shift their weight to one of these three locations. They monitored vertical loading via an oscilloscope to meet the appropriate target. The target line for weight shift on the oscilloscope was placed at positions where the subject would have approximately 70% of body weight on one leg (loaded) and concurrently 30% on the other leg (unloaded). This was the same for both right and left directions. The final condition was a left or right foot stimulus while standing in a neutral (50 / 50) position, with the weight equally distributed between the two feet. The order of the task conditions and side of stimulation was randomized. The noxious stimulus was delivered at varied times (ranging between 2 and 5 s) after the subject had reached the static location, to avoid anticipation. Data from the force plate were collected and digitally converted (National Instruments NB M10-16XE) at a sampling frequency of 200 Hz. The data were dual low pass filtered using a second order Butterworth filter with a 10 Hz cut off. Foot off and foot contact times were collected in addition to the forceplate data. The CoP parameters of interest were the presence or absence of medio-lateral (ML) BAs, onset latency, and peak magnitude. Previous studies have determined that a M–L CoP shift towards the swing limb in voluntary movement represents an increase in force to move the CoM onto the stance limb [17]. The criteria used in the present study to establish the occurrence of a BA was a shift of the CoP towards the swing foot, after stimulation onset. To denote the presence of a BA the CoP had to move towards the swing foot with a magnitude which surpassed the threshold band of 4 mm (calculation based on CoP sway during quiet stance) [21]. Such lateral excursions of the CoP are referred to in the present study as a BAs, since simple limb withdrawal or crossed extension would be characterized by movement of the CoP towards the stance limb. Onset latency was determined as the time from stimulus onset until the time when the CoP crossed the 4 mm threshold boundary. Magnitude was quantified as the peak ML shift of the CoP beyond the 4 mm threshold prior to unloading of the swing limb. The data were rejected if the latency exceeded the 95% confidence interval around the average onset latency for voluntary stimulation trials (1.53PT), determined as 183 ms to 217 ms in a previous study [20]. As a result, noxious
102
L.R. Bent et al. / Brain Research 914 (2001) 100 – 105
trials, in which the latency exceeded 183 ms, were excluded. Response frequency compared between task conditions was evaluated using Chi square analyses. One way repeated measures ANOVAs, blocked on subjects, with three replicates, were run to test the main effect of an increase in static loading on the stimulated limb for the dependent variable onset latency. Multiple comparisons were made using the Tukey’s honestly significant difference to test the experimental hypotheses. To address concerns for non normal distribution of data, the analysis of magnitude featured non-parametric Wilcoxon and Kruskall Wallis tests for two and three level comparisons respectively. ´ post-hoc test. Post-hoc analyses were made using Scheffes For all tests the significance level was set at P,0.05.
3. Results A total of 85 stepping responses to the stimulation were recorded during the study. Five of the fourteen subjects tested accounted for 77 of these 85 steps. These subjects are referred to as steppers in the present study. A separate group of five subjects accounted for 87 of the non-stepping responses (n5127). These five subjects are referred to as non-steppers. For the planned within-subject analyses, only the step data from the five steppers (n577) and the non-step responses from the non-steppers (n587) were used. Comparisons between step and no step data were made between these two groups. In addition, there were no significant differences found between the left and right legs for step frequency or for APA occurrence (P.0.05). Consequently data from left and right limbs were pooled for subsequent analysis. A third response, namely a toes up response (without foot removal from the ground), was observed in some subjects (n527). However, on account that the study focused on balance adjustments preceding step responses, and the toes up response could not be classified clearly as a step or no step response they were not analysed further. When considering individually the group of steppers, the frequency of occurrence of the CoP shift towards the swing foot (BA) was 34% (26 / 77). BA occurrence in the non steppers was 54% (46 / 87). As hypothesized, when comparing the frequency of BAs over conditions for step data, the occurrence was highest in the loaded condition as compared to the neutral or unloaded conditions; 61, 20, and 24% respectively (Fig. 1). This large increase in frequency of BAs in the loaded trials accounted for the significant differences between task conditions ( x 2 510.88, P50.004). This task specific influence on occurrence of BAs was also reflected in the task differences in initial direction of the CoP excursions measured in trials from non-steppers ( x 2 512.01, P50.002)(Fig. 1). Interesting to note is that although the BA frequency appears to be greater in non-steppers, it is the magnitude of
Fig. 1. Frequency of balance adjustment occurrence. Frequency of occurrence (%) of balance adjustments (BAs) for each loading condition in separated step and no-step data. As the postural demands on the limb increase, so does the occurrence of BAs. Differences between tasks were statistically significant (P,0.05).
the BAs that may reflect their effectiveness to shift CoM. The magnitude of M–L CoP shift towards the swing foot in step data was observed to increase significantly from the unloaded to neutral to the loaded condition; 1.1, 3.2, 5.4 cm ( x 2 513.15, P50.001)(Fig. 2A). In contrast, inspection of the magnitude of BAs exhibited by non-steppers revealed they were small and not different from the unloaded, to neutral to the loaded trials; 0.8, 1.6, 1.3 cm respectively (Fig. 2A). While most stepping responses to the noxious stimulation involved unloading of the stimulated foot, a unique and unexpected result was the removal of the unstimulated foot from the plates. All 5 stepping subjects exhibited trials in which steps were taken with the unstimulated foot, representing 27% (21 / 77) of the stepping responses. The majority of these steps with the unstimulated foot occurred when the stimulated foot was in the most loaded condition (78%, x 2 543.044, P,0.001). As a result of this occurrence, in order to address the primary control related to the rapid removal of the foot from the noxious stimulus, trials were isolated in which steps were taken only with the stimulated foot. Importantly this group of data, represented by trials only from the unloaded and neutral conditions, demonstrated an increase in magnitude between the two statically loaded positions. As the percent of BW increased on the stimulated limb from the unloaded to the neutral position, the magnitude of the CoP excursions towards the swing foot increased significantly from 0.36 cm to 2.1 cm respectively ( x 2 5 5.21, P50.02) (Fig. 2B). Modulations were also observed in onset latency of the BAs in accordance with the changing postural requirements. In the step data, increased static load on the stimulated leg had a significant effect on the onset latency (F2,8 59.33, P50.008). The step trial onset averages were 109, 128, and 125 ms for the unloaded, neutral, and loaded
L.R. Bent et al. / Brain Research 914 (2001) 100 – 105
Fig. 2. Magnitude of balance adjustments. Average magnitude of balance adjustment (BA), measured from baseline to the peak medio-lateral (ML) centre-of-pressure (CoP) shift, along with the associated standard errors. (A) Magnitude compared over the weight conditions (unloaded, neutral and loaded). Data is shown for stepping and non-stepping data. M–L CoP magnitude increases significantly for stepping trials only, as the weight on the stimulated limb increases (P,0.05). (B) Magnitude of BAs in responses where the stimulated foot was lifted. Only the neutral and unloaded trials are included because 78% of steps with the unstimulated foot occurred in the loaded condition. Unloaded versus neutral conditions featured a significant increase in BA magnitude, demonstrating a rapid postural response that is modulated according to the task (P,0.05).
conditions respectively. The significant difference found was accounted for by the rapid onset of the unloaded trials relative to the other two conditions. A significant increase was also found when focusing on trials featuring only stimulated foot removal in the unloaded and neutral conditions; 110 ms and 130 ms respectively (F1,4 544.44, P50.003). In opposition to this trend, the onset latency for non step data was not significantly different between any of the unloaded, neutral or loaded conditions P,0.05).
4. Discussion The results confirm previous observations of the presence of rapid onset latency BAs preceding noxious with-
103
drawal responses [20]. Importantly, the present study supports the hypothesis that it is possible to modify the characteristics of very rapid BAs that precede the withdrawal of the lower limb to noxious stimulation. This is supported by the observation that magnitude of the CoP excursion increases significantly as the static load on the stimulated limb is increased. Additional support also comes from the observation that the frequency of BA occurrence increases as the postural demands associated with the task increase from the unloaded to the loaded condition, as the loading on the stimulated limb increases. Also, the onset latency of the response increases significantly from the unloaded to the neutral condition. Finally one might also view the unique observation of frequent removal of the unstimulated foot in the loaded condition as support for the view that the CNS makes rapid latency BAs that are modulated for the purpose of preserving stability when executing rapid withdrawal responses. The rapid onset latency of 12864.6 ms (S.E.) observed in the neutral stance trials of the present experiment supports previous latencies of 12964.6 ms (S.E.) reported in the literature [20]. It is worth emphasizing that these response latencies reflect changes in ground reaction force which would be preceded by underlying EMG response onset latencies occurring 30–60 ms prior to measured force changes [23]. Interestingly, the observation in this study of onset latency modifiability with an increased static load on the stimulated limb, was not anticipated. There are several possible reasons for the presence of this modulation. One may be related to changes in ongoing sensory discharge and motoneuron activation. Previous studies have revealed a decrease in latency of a reflexive response with modest muscle contraction [1,28,29]. Another possibility for this timing change measured from ground reaction forces could be attributed to alterations in the electromechanical delay due to variations in tonic loading. The presence of BAs, in addition to the strategy observed of unstimulated foot removal, appears to support the view that these reactions exist to deal with the objective of reacting to the noxious stimulus while predictively minimizing whole-body instability. Clearly, the mere evidence of a rapid foot withdrawal with a preceding BA supports such an objective. This implementation of a predictive balance control strategy prior to the initiation of a reactive stepping reaction has been observed in studies exploring compensatory stepping reactions [18]. It is noteworthy that the present reactions occur at more rapid latencies and the amplitudes of the balance adjustments are much larger than seen in such compensatory reactions. In the present study, responses that include a BA, lead to a trade-off between the need to rapidly withdrawal the foot and the objective of maintaining stability. This seems to support the suggestion that the hierarchy of control is prioritized to address the maintenance of balance as an issue of primary importance even though this prolongs exposure to the noxious stimulus (longer time to foot-off).
104
L.R. Bent et al. / Brain Research 914 (2001) 100 – 105
The strategy of unloading the non-stimulated, when the stimulated foot is prior loaded, may be a specific example of this trade-off between balance and time to withdrawal the foot. This reaction may be an intentional strategy to help to move the COM towards the unstimulated foot under the influence of gravity or may be a result of a large medio-lateral balance adjustment. Despite the numerous possible responses; lifting the stimulated or unstimulated foot with or without a BA, it is apparent that the CNS may have several options for successfully achieving the often competing objectives of rapid limb withdrawal and maintaining whole-body stability. The present observation of rapid latency reactions (scaled to suit the task conditions) coincides with numerous studies revealing task-specific modulation of reflex activity [10,24,27]. The findings of the present study clearly highlight the ability to rapidly generate and modify postural adjustments to enable fast and appropriate responses to instabilities that may quickly arise in an environment that is consistently changing. The results have implications to the understanding of the coupling of BAs and rapid compensatory stepping reactions, specifically that the failure to appropriately scale BAs was not attributable to the rapid onset latency as once thought. There are also parallels between the presently observed balance adjustments and anticipatory postural adjustments (APAs) preceding voluntary movement. The one distinguishing characteristic is the rapid latency of the present reactions that precede withdrawal reactions, which are often considered to be spinally mediated. It is possible that BAs and APAs are regulated by common neural substrates, which might raise questions about the view of supraspinal control of APAs and their invariable coupling to voluntary movement.
Acknowledgements This work was supported by funding from the Natural Science and Engineering Council of Canada(WEM).
References [1] A.M. Aniss, S.C. Gandevia, D. Burke, Reflex responses in active muscles elicited by stimulation of low-threshold afferents from the human foot, J. Neurophysiol. 67 (1992) 1375–1384. [2] M. Belanger, A.E. Patla, Phase-dependent compensatory responses to perturbation applied during walking in humans, J. Motor Behav. 19 (1987) 434–453. [3] S. Bouisset, M. Zattara, Biomechanical study of the programming of anticipatory postural adjustments associated with voluntary movement, J. Biomech. 20 (1987) 735–742. [4] J.E. Brown, J.S. Frank, Influence of event anticipation on postural actions accompanying voluntary movement, Exp. Brain Res. 67 (1987) 645–650.
[5] P.J. Cordo, L.M. Nashner, Properties of postural adjustments with rapid arm movements, J. Neurophysiol. 47 (1982) 287–302. [6] M.C. Do, B. Bussel, Y. Breniere, Influence of plantar cutaneous afferents on early compensatory reactions to forward fall, Exp. Brain Res. 79 (1990) 319–324. [7] J. Duysens, G.E. Loeb, Modulation of ipsi-and contralateral reflex responses in unrestrained walking cats, J. Neurophysiol. 44 (1980) 1024–1037. [8] J.J. Eng, D.A. Winter, A.E. Patla, Strategies for recovery from a trip in early and late swing during human walking, Exp. Brain Res. 102 (1994) 339–349. [9] R. Forget, Y. Lamarre, Anticipatory postural adjustments in the absence of normal peripheral feedback, Brain Res. 508 (1990) 176–179. [10] H. Forssberg, Stumbling corrective reaction: a phase-dependent compensatory reaction during locomotion, J. Neurophysiol. 42 (1979) 936–953. [11] K.E. Hagbarth, B.L. Finer, The plasticity of human withdrawal reflexes to noxious skin stimuli in lower limbs, Prog. Brain Res. 1 (1963) 65–78. [12] R. Herman, T. Cook, B. Cozzens, W. Freedman, Control of postural reactions in man: the initiation of gait, in: R.B. Stein, K.G. Pearson, R.S. Smith, J.B. Redford (Eds.), Control of Posture and Gait, Plenum, New York, 1973, pp. 363–388. [13] T. Kasai, K. Kawai, Quantitative EMG analysis of anticipatory postural adjustments of voluntary contraction of leg muscles in standing man, Electroenceph. Clin. Neurophys. 93 (1994) 184–187. [14] E. Kugelberg, K. Eklund, L. Grimby, An electromyographic study of the nociceptive reflexes of the lower limb. Mechanisms of the plantar response, Brain 83 (1960) 394–410. [15] W.A. Lee, C.F. Michaels, Y.C. Pai, The organization of torque and EMG activity during bilateral handle pulls by standing humans, Exp. Brain Res. 82 (1990) 304–314. [16] B.E. Maki, W.E. McIlroy, The role of limb movements in maintaining upright stance: The ‘change-in-support’ strategy, Phys. Therapy. 77 (1997) 488–507. [17] J. Massion, Movement, posture and equilibrium: Interaction and coordination, Prog. Neurobiol. 38 (1992) 35–56. [18] W.E. McIlroy, B.E. Maki, Do anticipatory postural adjustments precede compensatory stepping reactions evoked by perturbation?, Neurosci. Lett. 164 (1993) 199–202. [19] W.E. McIlroy, B.E. Maki, Adaptive changes to compensatory stepping responses, Gait Posture 3 (1995) 43–50. [20] W.E. McIlroy, L.R. Bent, J.R. Potvin, J.D. Brooke, B.E. Maki, Preparatory balance adjustments precede withdrawal response to noxious stimulation in standing humans, Neurosci. Lett. 267 (1999) 197–200. [21] W.E. McIlroy, B.E. Maki, The control of lateral stability during rapid stepping reactions evoked by antero-posterior perturbation: does anticipatory control play a role?, Gait Posture 9 (1999) 190– 198. [22] A. Nardone, M. Schieppati, Postural adjustments associated with voluntary contraction of leg muscles in standing man, Exp. Brain Res. 69 (1988) 469–480. [23] R.W. Norman, P.V. Komi, Electromechanical delay in skeletal muscle under normal movement conditions, Acta Physiol. Scand. 106 (1979) 241–248. [24] K.G. Pearson, J.E. Misiaszek, K. Fouad, Enhancement and resetting of locomotor activity by muscle afferents, Ann. NY Acad. Sci. 860 (1998) 203–215. [25] A. Roby-Brami, B. Bussel, Long latency spinal reflex in man after flexor reflex afferent stimulation, Brain 110 (1987) 707–725. [26] M.W. Rogers, Y.C. Pai, Dynamic transitions in stance support accompanying leg flexion movements in man, Exp. Brain Res. 81 (1990) 398–402. [27] A. Rossi, B. Decchi, Flexibility of lower limb reflex responses to painful cutaneous stimulation in standing humans: evidence of load-dependent modulation, J. Physiol. 481 (1994) 521–532.
L.R. Bent et al. / Brain Research 914 (2001) 100 – 105 [28] A. Rossi, A. Zalaffi, B. Decchi, Interaction of nociceptive and non-nociceptive cutaneous afferents from the foot sole in common reflex pathways to tibialis anterior motoneurones in humans, Brain Res. 714 (1996) 76–86.
105
[29] B.T. Shahani, R.R. Young Human flexor reflexes, J. Neurol. Neurosurg. Psych. 34 (1971) 616–627. [30] J.F. Yang, R.B. Stein, Phase-dependent reflex reversal in human leg muscles during walking, J. Neurophysiol. 63 (1990) 1109–1117.