- Email: [email protected]
Exploratory behavior during stance persists with visual feedback
Neuroscience 195 (2011) 54 –59
EXPLORATORY BEHAVIOR DURING STANCE PERSISTS WITH VISUAL FEEDBACK C. D. MURNAGHAN, B. C. HORSLEN, J. T. INGLIS AND M. G. CARPENTER*
or corrects for, deviations of the COM from a desired position or point of equilibrium (Horak and MacPherson, 1996). Therefore, it has been assumed that errors or delays in this control system lead to continuous movement or sway of the body during stance (Maurer and Peterka, 2005), and from this perspective, increases in the amplitude of postural sway are often used as evidence of poor postural control (van Emmerik and van Wegen, 2002). However, recent evidence has begun to challenge the view that postural sway is a consequence of an unstable balance system or errors in feedback control. By using a novel method to minimize or “lock” COM displacements in the anterior–posterior plane during stance without subject awareness, Carpenter et al. (2010) demonstrated that COP amplitude increased compared to when the COM swayed freely. Since locking stabilized movements of the COM, the increase in COP amplitude could not be attributed to increases in movement of the COM. Therefore, it was interpreted that COP displacements during locking may be actively driven by the central nervous system (CNS), potentially as an exploratory means to acquire sensory information. While the conclusions of Carpenter et al. (2010) were consistent with other evidence of exploratory behavior in nonpostural tasks (Burgess, 1989; Martinez-Conde et al., 2006; Sasaki et al., 1995), at least two potential counterarguments had to be acknowledged. One argument was that the apparatus used to stabilize body sway (Fig. 1) may have allowed sensory illusions to emerge from a perceived mismatch between the actual displacement of the body and changes registered by either cutaneous receptors (in the feet and/or dorsal aspects of the body that were in contact with the apparatus) and/or musculotendinous force feedback receptors at the ankle. While it is not possible to accurately account for all of the forces or other sources of feedback that could potentially contribute to sensory illusions, it is possible to reduce the likelihood that sensory illusions will develop. One proven method of reducing sensory illusions is to provide participants with accurate visual input of their actual limb/body position during a task (Lackner and Taublieb, 1984; Izumizaki et al., 2010). Therefore, the first goal of the experiment was to determine if the previously reported increases in COP displacement that occurred following an external stabilization of sway (Carpenter et al., 2010) would still persist when sensory illusions were minimized using real-time visual feedback of postural sway. If there was an influence of sensory illusions, we hypothesized that the effects of locking would be dependent on visual feedback of postural sway, with COP displacements increasing with stabilization under normal
School of Kinesiology, University of British Columbia, Vancouver, BC, Canada, V6T 1Z3
Abstract—Recent evidence showing center of pressure (COP) displacements increase following an external stabilization of the center of mass (COM) supports the theory that postural sway may be exploratory and serve as a means of acquiring sensory information. The aim of the current study was to further test this theory and rule out potential confounding effects of sensory illusions or motor drift on prior observations. Participants stood as still as possible in an apparatus which allowed movements of the COM to be stabilized (“locked”) without subject awareness, and they were provided real-time visual feedback of their COM or COP throughout the trial. If there was an influence of sensory illusions or motor drift, we hypothesized that the change in COP displacement with locking would be reduced when participants were provided visual confirmation of COM stabilization (COM feedback), or when they were aware of the position of the COP throughout the trial (COP feedback). Confirming our previous results, increases in COP displacement were observed when movements of the COM were stabilized. In addition, our results showed that increases in COP displacement could not be explained by the presence of sensory illusions or motor drift, since increases in COP were observed despite being provided convincing evidence that the COM had been stabilized, and when participants were aware of their COP position throughout the trial. These results provide further support for an exploratory role of postural sway. The theoretical basis of current clinical practices designed to deal with balance control deficits due to age or disease is typically based on the opinion that increases in sway are a consequence of a failing balance control system. Our results suggest that this may not be the case, and if sway is in fact exploratory, a serious re-evaluation of current clinical practices may be warranted. © 2011 IBRO. Published by Elsevier Ltd. All rights reserved. Key words: posture, stability, postural sway, neural control, sensorimotor.
Postural sway is experienced by all humans when standing quietly, yet the exact nature of postural sway is still unknown. The continuous and random nature of postural sway is commonly thought to result from the interplay between movements of the body or center of mass (COM), and the resultant ground-reaction force acting beneath the feet (center of pressure [COP]), where the COP controls, *Corresponding author. Tel: ⫹1-604-822-8614. E-mail address: [email protected] (M. G. Carpenter). Abbreviations: AP, anterior–posterior; COM, center of mass; COP, center of pressure; RMS, root mean square.
0306-4522/11 $ - see front matter © 2011 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2011.08.020
54
C. D. Murnaghan et al. / Neuroscience 195 (2011) 54 –59
55
Fig. 1. Sagittal plane (A) and overhead (B) views of the apparatus used to minimize or “lock” AP movements of the COM without the subject’s awareness. Participants performed three trials corresponding to each of the feedback conditions. In the NOFB condition, participants stood as still as possible without any visual feedback. In the COPFB and COMFB conditions, participants were provided real-time visual feedback of either the COP or COM throughout the trial on the monitor illustrated in the figure.
conditions, yet decreasing with stabilization when sway feedback was available. A second possible counter-argument to the conclusions of Carpenter et al. (2010) was that the observed increases in COP amplitude with locking were the result of a drift in motor output, as opposed to exploratory behavior. Motor drift is a phenomenon observed in de-afferented individuals during tasks requiring either constant position or constant force output. Without visual feedback, these individuals, over time, have motor output that tends to drift either in a specific or random direction, depending on the goal of the task (Rothwell et al., 1982; Sanes et al., 1984). Assuming that the locking apparatus used by Carpenter et al. (2010) could minimize postural sway to such an extent that proprioceptive inputs were unable to exceed receptor thresholds, it could be argued that subjects may be susceptible to motor drift, in the same way as de-afferented patients. Therefore, the second goal of the study was to determine if the previously reported increases in COP displacement that occurred following an external stabiliza-
tion of sway (Carpenter et al., 2010) still persist with the availability of real-time visual feedback of the COP position. The COP position reflects the net neuromuscular output at the ankle joint (Winter et al., 1998), and when participants are provided with such an indication of motor output, the potential for motor drift is expected to be minimized. If there was an influence of motor drift, we hypothesized that the effects of stabilization would be dependent on visual feedback of COP position, with COP displacements increasing with stabilization under normal conditions, yet decreasing with stabilization when COP feedback was available.
EXPERIMENTAL PROCEDURES Participants Sixteen healthy young adults (eight females; mean⫾SD for age⫽22.9⫾3.9 years; height⫽174.3⫾9.7 cm; weight⫽67.0⫾9.8 kg) volunteered in the study. Each participant provided informed written consent, and the experimental protocol was approved by the
56
C. D. Murnaghan et al. / Neuroscience 195 (2011) 54 –59
Behavioural Research Ethics Board at the University of British Columbia. All subjects were completely naïve to the goals of the experiment and the intended effect of the apparatus on postural sway.
Apparatus Throughout the entire experiment, subjects were fixed firmly to a rigid board that extended from the head to calf, using adjustable straps tightened firmly around the head, shoulders, chest, waist, hips/upper thighs, and upper shank, to prevent movement about any joint except the ankle. The board was 1.66 m high (including head rest), 0.61 m wide, and had total mass of 12.5 kg. The board was attached to a closed-loop pulley system that allowed “normal postural sway” at the ankle joint unless the experimenter applied a brake that would discretely “lock” or stabilize the board (and thus COM) in place in the sagittal plane without participants’ knowledge (Fig. 1). There were two reasons for fixing subjects to the rigid board within the apparatus. First, it allowed us to minimize the potential for subjects to detect any forces exerted by the apparatus when the brake was applied, by distributing the forces over a large area along the entire dorsal aspect of the body (Fitzpatrick et al., 1992). Second, fixing subjects to the board ensured that postural sway was controlled only about the ankle joint, so that any effects of our manipulation would be measured directly through changes in COP. While we recognize that under normal sway conditions there are also small fluctuations that occur at other joints, the body has still been shown to be controlled as an inverted pendulum using primarily ankle torques in the anterior– posterior direction (Gage et al., 2004). In our original study, we tested the effect of having subjects fixed to the board on postural sway by including a control condition in which subjects stood free from the board and apparatus. While there were slight differences in the amplitude of sway when subjects stood with the board compared to without the board (Carpenter et al., 2010), it is important to note that any effects of having subjects fixed to the board will remain constant across the unlocked and locked conditions, and therefore will have no effect on the primary results of the study. To eliminate any chance that the participant could receive auditory cues to indicate that they had been locked, all participants wore headphones that reduced any noise within the testing area. Participants also wore blinders designed to occlude both horizontal and vertical peripheral vision, yet allow normal vision anteriorly. In all conditions, subjects stood with their arms crossed and feet shoulder width apart on a forceplate (#K00407, Bertec, USA).
COPFB and COMFB), the first 30 s of each trial was used to allow the transient component of sway to stabilize (Carroll and Freedman, 1993), and the following 30 s was used to calculate the mean COP position to be used as the threshold for “locking.” Following the initial 60 s, participants stood freely in the “unlocked” condition for a minimum of 75 s (unlocked) and then the board was locked when the COP was within 2 standard deviations of the calculated mean COP, for a further minimum 75 s (locked). The COP visual feedback was smoothed using a low-pass digital filter with a time constant of 0.2 s to ensure that participants did not respond to very high frequency components within the signal.
Measurements Ground reaction forces and moments were sampled at 100 Hz and low-pass filtered offline using a 5 Hz dual-pass Butterworth filter before calculating COP in the AP direction. From these signals, the root mean square (RMS) of the AP COP displacements (COP RMS) was calculated. Kinematics were sampled at 100 Hz for the duration of each trial using an Optotrak three dimensional optical motion analysis system (Northern Digital Inc., Waterloo, Ontario, Canada) with infrared light emitting diodes placed on the cable, ankle joint, and back of the board at a level that approximated the height of the COM (⬃2⁄3 participants’ height, Winter et al., 1998). Kinematic data were low-pass filtered at 5 Hz with a dual-pass Butterworth filter and used to calculate the angular displacement of the COM in the sagittal plane. From this signal, the RMS of AP COM angular displacement (COM RMS) was calculated. All dependent variables were calculated for each trial over a 60 s period when participants were unlocked and locked. The 60 s measurement periods began 75 s before the beginning and end of the braking period, respectively.
Statistical analysis To ensure the effectiveness of the apparatus was consistent across conditions, we tested whether there were any differences in the sagittal plane displacement of the cable marker during locking in each of the three feedback conditions using a one-way repeated measures ANOVA. To test the hypothesis related to the effect of sensory illusions, we compared the change in COP RMS from unlocked to locked across NOFB and COMFB conditions using a paired t-test. To test the hypothesis related to the influence of motor drift, we compared the change in COP RMS from unlocked to locked across NOFB and COPFB conditions using a paired t-test. In all cases, alpha level was set at 0.05.
Experimental protocol All subjects were fitted into the apparatus and asked to stand as still as possible in four trials. First, participants performed a twominute trial where they were asked to stand as still as possible and to focus on a target located at eye-level 2 m in front of them (CONTROL condition). This trial was used to calculate the peakto-peak amplitude of anterior–posterior (AP) COP and COM marker displacements. Three additional feedback trials, each with a minimum duration of 3.5 min, were randomized and varied in the type of feedback participants received. In one trial, participants received instructions identical to those in the CONTROL condition (no feedback condition—NOFB). In the remaining two trials, participants received real-time COP (COP feedback condition—COPFB) or realtime COM (COM feedback condition—COMFB) feedback on a screen located at eye-level 2 m in front of them. After being familiarized with how to control their COP (or COM), participants were asked to stand for ⬃1 min while the visual feedback screen was adjusted to display their mean COP (or COM) position (scaled to ⫾ 1 peak-to-peak amplitude recorded during the CONTROL condition). Participants were then asked to maintain their COP (or COM) as close as possible to the mean displayed on the screen during the feedback trials. In all feedback conditions (NOFB,
RESULTS The apparatus was effective in minimizing movement of the COM during locking in all three feedback conditions (Figs. 2 and 3, upper panels). When participants were locked, there was no significant difference in the RMS amplitude of cable displacement between the NOFB (0.05⫾0.01 mm), COPFB (0.07⫾0.01 mm), and COMFB (0.06⫾0.01 mm) conditions. Across all three conditions, locking the cable was related to an average RMS and range of COM angular displacement of 0.02⫾0.00° and 0.09⫾0.02°, respectively. In the NOFB condition, 9 of 16 participants had the same or larger COP RMS during the locked compared to the unlocked condition (Figs. 2 and 3, bottom panels), with average COP RMS amplitude of 3.37⫾0.65 mm and 2.47⫾0.35 mm in the locked and unlocked condition, respectively. In contrast, with COM feedback available, 13 of 16 participants had larger COP RMS during the locked
C. D. Murnaghan et al. / Neuroscience 195 (2011) 54 –59
57
Fig. 2. Mean RMS of COM angular displacements (top panel) and AP COP displacements (bottom panel) for all subjects when locked compared to unlocked in each of the three feedback conditions (NOFB, COPFB, and COMFB). Black lines represent subjects who showed increases, while gray lines represent subjects who showed decreases, in COP displacement across locking conditions.
compared to unlocked condition (Figs. 2 and 3, bottom panels), with average COP RMS displacements of 4.36⫾0.88 mm and 1.78⫾0.17 mm, respectively. Although the percent change between locked and unlocked conditions was larger in the COM feedback (183% increase), compared to the no feedback condition (94.7% increase), the differences were not found to be statistically significant between feedback conditions (P⫽0.19).
In the COPFB condition, 15 of 16 participants had larger COP RMS during the locked compared to the unlocked condition (Figs. 2 and 3, bottom panels), with average displacements of 5.56⫾0.83 mm and 1.56⫾0.11 mm in the locked and unlocked conditions, respectively. With COPFB, the percent increase from unlocked to locked was 286%, which was significantly greater (P⫽0.03) than the increase observed in the no feedback condition (94.7% increase).
Fig. 3. COM angular displacements (top panel) and AP COP displacements (bottom panel) from a single representative subject in each of the three feedback conditions over the 60 s periods used for data analysis in the unlocked (gray lines) and locked (black lines) conditions.
58
C. D. Murnaghan et al. / Neuroscience 195 (2011) 54 –59
DISCUSSION Traditional theories suggest that the COP controls, or corrects for, deviations of the COM from a desired position or point of equilibrium. Based on these theories, minimizing or locking COM displacement without participants’ knowledge would be expected to result in a concurrent decrease in COP displacement. Contrary to these theories, the results of the current study converged with prior evidence, and showed that the COP actually increases when participants are locked compared to when they are unlocked and swaying freely. In the present study, we observed increases in RMS of COP displacement in the locked (3.37 mm) compared to unlocked (2.47 mm) condition without visual feedback (NOFB). Likewise, Carpenter et al. (2010) also showed that the RMS of COP displacement increased in the locked (3.25 mm) compared to unlocked (1.74 mm) conditions, respectively. While these results together suggest a possible exploratory role of postural sway, the goals of this study were to rule out whether the increases in COP displacements, observed when the COM was stabilized, could potentially be explained by the effects of sensory illusions or motor drift. Minimizing sensory illusions with real-time sway feedback (Lackner and Taublieb, 1984; Eklund, 1973) had no effect on the way in which COP displacements changed when movements of the COM were stabilized. Increases in COP displacement in the locked compared to unlocked conditions were observed irrespective of whether participants had visual feedback of sway. These results suggest that changes in COP displacements observed in our experiment cannot be explained by the presence of sensory illusions. A similar conclusion has been drawn from prior results. Since sensory illusions are known to be amplified by the removal of vision (Lackner and Taublieb, 1984; Eklund, 1973), it was expected that they would lead to larger changes in COP displacement during locking when subjects stood with eyes closed. However, this was not the case, as similar increases in COP displacements were observed between locked and unlocked conditions independent of whether subjects stood with or without vision (Carpenter et al., 2010). Furthermore, it is likely that any small pressure changes that may have occurred between the body and the apparatus were insufficient to generate sensory illusions. This is because we would expect minimal pressure changes on the trunk for such small disturbances that occur in the current experiment, due to the large surface area of the back of the board and long lever arm of the rigid board when compared to the size and length of the feet (Fitzpatrick et al., 1992). The second goal of the study was to determine if increases in COP displacement with locking (Carpenter et al., 2010) could be explained by the presence of motor drift. We tested this argument by providing participants visual feedback of the COP during both the locked and unlocked conditions, which can minimize the potential for motor drift. Contrary to our hypothesis, we did not observe a decrease in COP displacement in the locked compared with unlocked condition when COP feedback was avail-
able. In fact, we observed increases in COP displacement with locking that were even greater in the feedback compared with no feedback condition. These results suggest that the increased amplitude of COP displacements observed with locking in this and prior studies cannot be explained by motor drift. Furthermore, there is no evidence that minimizing sway using our apparatus decreases postural sway to the extent that it could create a sensory deprived state, as angular displacements of the COM in the locked condition were still found to be within the known range of perceptual thresholds for stance (Fitzpatrick and McCloskey, 1994) in both the current (average COM angular range across conditions of 0.09⫾0.02°) and previous studies (Carpenter et al., 2010). The current results provide further support for a possible exploratory role of postural sway. The suggestion that there may be an exploratory role for motor variability has been described in a variety of motor control tasks. For example, when fixating the eyes on a visual target, rapid, involuntary eye movements known as microsaccades play a fundamental role in visual perception by ensuring continuous motion of the image projected onto the retina. Without this continuous motion, the visual environment rapidly fades from view (Martinez-Conde et al., 2006; Rolfs, 2009). Therefore, microsaccades ensure constant change in visual input, which is ideal since our nervous system has evolved to detect changes more readily than static stimuli. Similarly in postural control, postural sway is commonly viewed as error or noise in the system (Błaszczyk et al., 2009; Riccio, 1993; Riley et al., 1997; van Emmerik and van Wegen, 2002). However, some studies have theorized that sway variability may represent a perception-action strategy that allows one to acquire essential information about the interaction with the environment (Riccio, 1993; Riley et al., 1997). As described previously (Carpenter et al., 2010), this action may serve as a means to acquire a certain quality and quantity of sensory information. In fact, arguments to this effect have been applied to the larger sway observed in younger children during development, which is thought to facilitate the formation of “a reliable and stable sensorimotor relationship for postural control” (Chen et al., 2008). In this case, postural sway may act as a means of gathering and experiencing a variety of sensory interactions during stance, in order to improve the development of a proficient postural control system. Given that larger sway in children may represent the CNS’s attempt to couple sensorimotor action during stance, it is conceivable that smaller amounts of controlled sway may continue to be used by the CNS throughout life as a means to ensure a constant flow of sensory inputs. Likewise, a return to larger sway in older adults may represent a means for the CNS to generate more sensory information to compensate for the decreased speed and sensitivity of sensory systems in this population. As opposed to a negative consequence of a failing postural control system, the increased magnitude of sway in older populations may be an adaptation of similar origin to the larger sway observed in young children, and may be re-
C. D. Murnaghan et al. / Neuroscience 195 (2011) 54 –59
flective of the CNS’s attempt to re-establish or strengthen sensorimotor control. There are important clinical implications from the results of the current study. Our findings of increased COP displacement, observed despite providing convincing feedback to the subjects, provide further support for an exploratory role of postural sway. In contrast to this view, larger magnitudes of postural sway in populations with balance deficits due to age or disease are considered to be indicative of instability and greater fall risk. Therefore, they are often treated with rehabilitation and preventative strategies aimed at decreasing sway. However, if postural sway is in fact exploratory, and beneficial in acquiring sensory information to develop a better representation of where the body is in space, attempting to decrease sway may oppose the natural adaptations of the CNS that are required to maintain the integrity of the postural control system. Based on conflicting interpretations of increases in postural sway in these populations, it is imperative that we first develop an improved understanding of the origins of postural sway in general. If, as our results suggest, there is support for an exploratory role of postural sway, a serious re-evaluation of current clinical practices designed to deal with balance control deficits due to age or disease may be warranted. Acknowledgments—We gratefully acknowledge the funding support provided by the Natural Sciences and Engineering Research Council of Canada to MG Carpenter, BC Horslen, JT Inglis and CD Murnaghan.
REFERENCES Błaszczyk JW, Cies´linska-Swider J, Plewa M, Zahorska-Markiewicz B, Markiewicz A (2009) Effects of excessive body weight on postural control. J Biomech 42:1295–1300. Burgess PR (1989) Movement cues are necessary for load stabilization against low friction. Soc Neurosci Abstr 15:173. Carpenter MG, Murnaghan CD, Inglis JT (2010) Shifting the balance: evidence of an exploratory role of postural sway. Neuroscience 171:283–291. Carroll JP, Freedman W (1993) Nonstationary properties of postural sway. J Biomech 26:409 – 416. Chen LC, Metcalfe JS, Chang TY, Jeka JJ, Clark JE (2008) The development of infant upright posture: sway less or sway differently? Exp Brain Res 186:293–303.
59
Eklund G (1973) Further studies of vibration-induced effects on balance. Ups J Med Sci 78:65–72. Fitzpatrick RC, McCloskey DI (1994) Proprioceptive, visual and vestibular thresholds for the perception of sway during standing in humans. J Physiol 478:173–186. Fitzpatrick RC, Taylor JL, McCloskey DI (1992) Ankle stiffness of standing humans in response to imperceptible perturbation: reflex and task-dependent components. J Physiol 454:533–547. Gage WH, Winter DA, Frank JS, Adkin AL (2004) Kinematic and kinetic validity of the inverted pendulum model in quiet standing. Gait Posture 19:124 –132. Horak F, MacPherson J (1996) Postural orientation and equilibrium. In: Handbook of physiology (Rowell LB, Shepherd JT, eds). New York: Oxford University Press. Izumizaki M, Tsuge M, Akai L, Proske U, Homma I (2010) The illusion of changed position and movement from vibrating one arm is altered by vision or movement of the other arm. J Physiol 588: 2789 –2800. Lackner JR, Taublieb AB (1984) Influence of vision on vibrationinduced illusions of limb movement. Exp Neurol 85:97–106. Martinez-Conde S, Macknik SL, Troncoso XG, Dyar TA (2006) Microsaccades counteract visual fading during fixation. Neuron 49: 297–305. Maurer C, Peterka RJ (2005) A new interpretation of spontaneous sway measures based on a simple model of human postural control. J Neurophysiol 93(1):189 –200. Riccio GE (1993) In variability and motor control (Newell KM, Corcos DM, eds), pp 317–358. Champaign, IL: Human Kinetics. Riley MA, Mitra S, Stoffregen TA, Turvey MT (1997) Influences of body lean and vision on unperturbed postural sway. Mot Control 1: 229 –246. Rolfs M (2009) Microsaccades: small steps on a long way. Vision Res 49:2415–2441. Rothwell JC, Traub MM, Day BL, Obeso JA, Thomas PK, Marsden CD (1982) Manual motor performance in a deafferented man. Brain 105:515–542. Sanes JN, Mauritz KH, Evarts EV, Dalakas MC, Chu A (1984) Motor deficits in patients with large-fiber sensory neuropathy. Proc Natl Acad Sci U S A 81:979 –982. Sasaki M, Mishima H, Suzuki K, Ohkura M (1995) Observations on micro-exploration in everyday activities. In: Studies in perception and action, Vol. III (Bardy BG, Bootsma RJ, Guiard Y, eds), pp 99 –102. Hillsdale, NJ: Lawrence Erlbaum Associates. van Emmerik RE, van Wegen EE (2002) On the functional aspects of variability in postural control. Exerc Sport Sci Rev 30:177–183. Winter DA, Patla AE, Prince F, Ishac M, Gielo-Perczak K (1998) Stiffness control of balance in quiet standing. J Neurophysiol 80:1211–1221.
(Accepted 9 August 2011) (Available online 12 August 2011)
EXPLORATORY BEHAVIOR DURING STANCE PERSISTS WITH VISUAL FEEDBACK C. D. MURNAGHAN, B. C. HORSLEN, J. T. INGLIS AND M. G. CARPENTER*
or corrects for, deviations of the COM from a desired position or point of equilibrium (Horak and MacPherson, 1996). Therefore, it has been assumed that errors or delays in this control system lead to continuous movement or sway of the body during stance (Maurer and Peterka, 2005), and from this perspective, increases in the amplitude of postural sway are often used as evidence of poor postural control (van Emmerik and van Wegen, 2002). However, recent evidence has begun to challenge the view that postural sway is a consequence of an unstable balance system or errors in feedback control. By using a novel method to minimize or “lock” COM displacements in the anterior–posterior plane during stance without subject awareness, Carpenter et al. (2010) demonstrated that COP amplitude increased compared to when the COM swayed freely. Since locking stabilized movements of the COM, the increase in COP amplitude could not be attributed to increases in movement of the COM. Therefore, it was interpreted that COP displacements during locking may be actively driven by the central nervous system (CNS), potentially as an exploratory means to acquire sensory information. While the conclusions of Carpenter et al. (2010) were consistent with other evidence of exploratory behavior in nonpostural tasks (Burgess, 1989; Martinez-Conde et al., 2006; Sasaki et al., 1995), at least two potential counterarguments had to be acknowledged. One argument was that the apparatus used to stabilize body sway (Fig. 1) may have allowed sensory illusions to emerge from a perceived mismatch between the actual displacement of the body and changes registered by either cutaneous receptors (in the feet and/or dorsal aspects of the body that were in contact with the apparatus) and/or musculotendinous force feedback receptors at the ankle. While it is not possible to accurately account for all of the forces or other sources of feedback that could potentially contribute to sensory illusions, it is possible to reduce the likelihood that sensory illusions will develop. One proven method of reducing sensory illusions is to provide participants with accurate visual input of their actual limb/body position during a task (Lackner and Taublieb, 1984; Izumizaki et al., 2010). Therefore, the first goal of the experiment was to determine if the previously reported increases in COP displacement that occurred following an external stabilization of sway (Carpenter et al., 2010) would still persist when sensory illusions were minimized using real-time visual feedback of postural sway. If there was an influence of sensory illusions, we hypothesized that the effects of locking would be dependent on visual feedback of postural sway, with COP displacements increasing with stabilization under normal
School of Kinesiology, University of British Columbia, Vancouver, BC, Canada, V6T 1Z3
Abstract—Recent evidence showing center of pressure (COP) displacements increase following an external stabilization of the center of mass (COM) supports the theory that postural sway may be exploratory and serve as a means of acquiring sensory information. The aim of the current study was to further test this theory and rule out potential confounding effects of sensory illusions or motor drift on prior observations. Participants stood as still as possible in an apparatus which allowed movements of the COM to be stabilized (“locked”) without subject awareness, and they were provided real-time visual feedback of their COM or COP throughout the trial. If there was an influence of sensory illusions or motor drift, we hypothesized that the change in COP displacement with locking would be reduced when participants were provided visual confirmation of COM stabilization (COM feedback), or when they were aware of the position of the COP throughout the trial (COP feedback). Confirming our previous results, increases in COP displacement were observed when movements of the COM were stabilized. In addition, our results showed that increases in COP displacement could not be explained by the presence of sensory illusions or motor drift, since increases in COP were observed despite being provided convincing evidence that the COM had been stabilized, and when participants were aware of their COP position throughout the trial. These results provide further support for an exploratory role of postural sway. The theoretical basis of current clinical practices designed to deal with balance control deficits due to age or disease is typically based on the opinion that increases in sway are a consequence of a failing balance control system. Our results suggest that this may not be the case, and if sway is in fact exploratory, a serious re-evaluation of current clinical practices may be warranted. © 2011 IBRO. Published by Elsevier Ltd. All rights reserved. Key words: posture, stability, postural sway, neural control, sensorimotor.
Postural sway is experienced by all humans when standing quietly, yet the exact nature of postural sway is still unknown. The continuous and random nature of postural sway is commonly thought to result from the interplay between movements of the body or center of mass (COM), and the resultant ground-reaction force acting beneath the feet (center of pressure [COP]), where the COP controls, *Corresponding author. Tel: ⫹1-604-822-8614. E-mail address: [email protected] (M. G. Carpenter). Abbreviations: AP, anterior–posterior; COM, center of mass; COP, center of pressure; RMS, root mean square.
0306-4522/11 $ - see front matter © 2011 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2011.08.020
54
C. D. Murnaghan et al. / Neuroscience 195 (2011) 54 –59
55
Fig. 1. Sagittal plane (A) and overhead (B) views of the apparatus used to minimize or “lock” AP movements of the COM without the subject’s awareness. Participants performed three trials corresponding to each of the feedback conditions. In the NOFB condition, participants stood as still as possible without any visual feedback. In the COPFB and COMFB conditions, participants were provided real-time visual feedback of either the COP or COM throughout the trial on the monitor illustrated in the figure.
conditions, yet decreasing with stabilization when sway feedback was available. A second possible counter-argument to the conclusions of Carpenter et al. (2010) was that the observed increases in COP amplitude with locking were the result of a drift in motor output, as opposed to exploratory behavior. Motor drift is a phenomenon observed in de-afferented individuals during tasks requiring either constant position or constant force output. Without visual feedback, these individuals, over time, have motor output that tends to drift either in a specific or random direction, depending on the goal of the task (Rothwell et al., 1982; Sanes et al., 1984). Assuming that the locking apparatus used by Carpenter et al. (2010) could minimize postural sway to such an extent that proprioceptive inputs were unable to exceed receptor thresholds, it could be argued that subjects may be susceptible to motor drift, in the same way as de-afferented patients. Therefore, the second goal of the study was to determine if the previously reported increases in COP displacement that occurred following an external stabiliza-
tion of sway (Carpenter et al., 2010) still persist with the availability of real-time visual feedback of the COP position. The COP position reflects the net neuromuscular output at the ankle joint (Winter et al., 1998), and when participants are provided with such an indication of motor output, the potential for motor drift is expected to be minimized. If there was an influence of motor drift, we hypothesized that the effects of stabilization would be dependent on visual feedback of COP position, with COP displacements increasing with stabilization under normal conditions, yet decreasing with stabilization when COP feedback was available.
EXPERIMENTAL PROCEDURES Participants Sixteen healthy young adults (eight females; mean⫾SD for age⫽22.9⫾3.9 years; height⫽174.3⫾9.7 cm; weight⫽67.0⫾9.8 kg) volunteered in the study. Each participant provided informed written consent, and the experimental protocol was approved by the
56
C. D. Murnaghan et al. / Neuroscience 195 (2011) 54 –59
Behavioural Research Ethics Board at the University of British Columbia. All subjects were completely naïve to the goals of the experiment and the intended effect of the apparatus on postural sway.
Apparatus Throughout the entire experiment, subjects were fixed firmly to a rigid board that extended from the head to calf, using adjustable straps tightened firmly around the head, shoulders, chest, waist, hips/upper thighs, and upper shank, to prevent movement about any joint except the ankle. The board was 1.66 m high (including head rest), 0.61 m wide, and had total mass of 12.5 kg. The board was attached to a closed-loop pulley system that allowed “normal postural sway” at the ankle joint unless the experimenter applied a brake that would discretely “lock” or stabilize the board (and thus COM) in place in the sagittal plane without participants’ knowledge (Fig. 1). There were two reasons for fixing subjects to the rigid board within the apparatus. First, it allowed us to minimize the potential for subjects to detect any forces exerted by the apparatus when the brake was applied, by distributing the forces over a large area along the entire dorsal aspect of the body (Fitzpatrick et al., 1992). Second, fixing subjects to the board ensured that postural sway was controlled only about the ankle joint, so that any effects of our manipulation would be measured directly through changes in COP. While we recognize that under normal sway conditions there are also small fluctuations that occur at other joints, the body has still been shown to be controlled as an inverted pendulum using primarily ankle torques in the anterior– posterior direction (Gage et al., 2004). In our original study, we tested the effect of having subjects fixed to the board on postural sway by including a control condition in which subjects stood free from the board and apparatus. While there were slight differences in the amplitude of sway when subjects stood with the board compared to without the board (Carpenter et al., 2010), it is important to note that any effects of having subjects fixed to the board will remain constant across the unlocked and locked conditions, and therefore will have no effect on the primary results of the study. To eliminate any chance that the participant could receive auditory cues to indicate that they had been locked, all participants wore headphones that reduced any noise within the testing area. Participants also wore blinders designed to occlude both horizontal and vertical peripheral vision, yet allow normal vision anteriorly. In all conditions, subjects stood with their arms crossed and feet shoulder width apart on a forceplate (#K00407, Bertec, USA).
COPFB and COMFB), the first 30 s of each trial was used to allow the transient component of sway to stabilize (Carroll and Freedman, 1993), and the following 30 s was used to calculate the mean COP position to be used as the threshold for “locking.” Following the initial 60 s, participants stood freely in the “unlocked” condition for a minimum of 75 s (unlocked) and then the board was locked when the COP was within 2 standard deviations of the calculated mean COP, for a further minimum 75 s (locked). The COP visual feedback was smoothed using a low-pass digital filter with a time constant of 0.2 s to ensure that participants did not respond to very high frequency components within the signal.
Measurements Ground reaction forces and moments were sampled at 100 Hz and low-pass filtered offline using a 5 Hz dual-pass Butterworth filter before calculating COP in the AP direction. From these signals, the root mean square (RMS) of the AP COP displacements (COP RMS) was calculated. Kinematics were sampled at 100 Hz for the duration of each trial using an Optotrak three dimensional optical motion analysis system (Northern Digital Inc., Waterloo, Ontario, Canada) with infrared light emitting diodes placed on the cable, ankle joint, and back of the board at a level that approximated the height of the COM (⬃2⁄3 participants’ height, Winter et al., 1998). Kinematic data were low-pass filtered at 5 Hz with a dual-pass Butterworth filter and used to calculate the angular displacement of the COM in the sagittal plane. From this signal, the RMS of AP COM angular displacement (COM RMS) was calculated. All dependent variables were calculated for each trial over a 60 s period when participants were unlocked and locked. The 60 s measurement periods began 75 s before the beginning and end of the braking period, respectively.
Statistical analysis To ensure the effectiveness of the apparatus was consistent across conditions, we tested whether there were any differences in the sagittal plane displacement of the cable marker during locking in each of the three feedback conditions using a one-way repeated measures ANOVA. To test the hypothesis related to the effect of sensory illusions, we compared the change in COP RMS from unlocked to locked across NOFB and COMFB conditions using a paired t-test. To test the hypothesis related to the influence of motor drift, we compared the change in COP RMS from unlocked to locked across NOFB and COPFB conditions using a paired t-test. In all cases, alpha level was set at 0.05.
Experimental protocol All subjects were fitted into the apparatus and asked to stand as still as possible in four trials. First, participants performed a twominute trial where they were asked to stand as still as possible and to focus on a target located at eye-level 2 m in front of them (CONTROL condition). This trial was used to calculate the peakto-peak amplitude of anterior–posterior (AP) COP and COM marker displacements. Three additional feedback trials, each with a minimum duration of 3.5 min, were randomized and varied in the type of feedback participants received. In one trial, participants received instructions identical to those in the CONTROL condition (no feedback condition—NOFB). In the remaining two trials, participants received real-time COP (COP feedback condition—COPFB) or realtime COM (COM feedback condition—COMFB) feedback on a screen located at eye-level 2 m in front of them. After being familiarized with how to control their COP (or COM), participants were asked to stand for ⬃1 min while the visual feedback screen was adjusted to display their mean COP (or COM) position (scaled to ⫾ 1 peak-to-peak amplitude recorded during the CONTROL condition). Participants were then asked to maintain their COP (or COM) as close as possible to the mean displayed on the screen during the feedback trials. In all feedback conditions (NOFB,
RESULTS The apparatus was effective in minimizing movement of the COM during locking in all three feedback conditions (Figs. 2 and 3, upper panels). When participants were locked, there was no significant difference in the RMS amplitude of cable displacement between the NOFB (0.05⫾0.01 mm), COPFB (0.07⫾0.01 mm), and COMFB (0.06⫾0.01 mm) conditions. Across all three conditions, locking the cable was related to an average RMS and range of COM angular displacement of 0.02⫾0.00° and 0.09⫾0.02°, respectively. In the NOFB condition, 9 of 16 participants had the same or larger COP RMS during the locked compared to the unlocked condition (Figs. 2 and 3, bottom panels), with average COP RMS amplitude of 3.37⫾0.65 mm and 2.47⫾0.35 mm in the locked and unlocked condition, respectively. In contrast, with COM feedback available, 13 of 16 participants had larger COP RMS during the locked
C. D. Murnaghan et al. / Neuroscience 195 (2011) 54 –59
57
Fig. 2. Mean RMS of COM angular displacements (top panel) and AP COP displacements (bottom panel) for all subjects when locked compared to unlocked in each of the three feedback conditions (NOFB, COPFB, and COMFB). Black lines represent subjects who showed increases, while gray lines represent subjects who showed decreases, in COP displacement across locking conditions.
compared to unlocked condition (Figs. 2 and 3, bottom panels), with average COP RMS displacements of 4.36⫾0.88 mm and 1.78⫾0.17 mm, respectively. Although the percent change between locked and unlocked conditions was larger in the COM feedback (183% increase), compared to the no feedback condition (94.7% increase), the differences were not found to be statistically significant between feedback conditions (P⫽0.19).
In the COPFB condition, 15 of 16 participants had larger COP RMS during the locked compared to the unlocked condition (Figs. 2 and 3, bottom panels), with average displacements of 5.56⫾0.83 mm and 1.56⫾0.11 mm in the locked and unlocked conditions, respectively. With COPFB, the percent increase from unlocked to locked was 286%, which was significantly greater (P⫽0.03) than the increase observed in the no feedback condition (94.7% increase).
Fig. 3. COM angular displacements (top panel) and AP COP displacements (bottom panel) from a single representative subject in each of the three feedback conditions over the 60 s periods used for data analysis in the unlocked (gray lines) and locked (black lines) conditions.
58
C. D. Murnaghan et al. / Neuroscience 195 (2011) 54 –59
DISCUSSION Traditional theories suggest that the COP controls, or corrects for, deviations of the COM from a desired position or point of equilibrium. Based on these theories, minimizing or locking COM displacement without participants’ knowledge would be expected to result in a concurrent decrease in COP displacement. Contrary to these theories, the results of the current study converged with prior evidence, and showed that the COP actually increases when participants are locked compared to when they are unlocked and swaying freely. In the present study, we observed increases in RMS of COP displacement in the locked (3.37 mm) compared to unlocked (2.47 mm) condition without visual feedback (NOFB). Likewise, Carpenter et al. (2010) also showed that the RMS of COP displacement increased in the locked (3.25 mm) compared to unlocked (1.74 mm) conditions, respectively. While these results together suggest a possible exploratory role of postural sway, the goals of this study were to rule out whether the increases in COP displacements, observed when the COM was stabilized, could potentially be explained by the effects of sensory illusions or motor drift. Minimizing sensory illusions with real-time sway feedback (Lackner and Taublieb, 1984; Eklund, 1973) had no effect on the way in which COP displacements changed when movements of the COM were stabilized. Increases in COP displacement in the locked compared to unlocked conditions were observed irrespective of whether participants had visual feedback of sway. These results suggest that changes in COP displacements observed in our experiment cannot be explained by the presence of sensory illusions. A similar conclusion has been drawn from prior results. Since sensory illusions are known to be amplified by the removal of vision (Lackner and Taublieb, 1984; Eklund, 1973), it was expected that they would lead to larger changes in COP displacement during locking when subjects stood with eyes closed. However, this was not the case, as similar increases in COP displacements were observed between locked and unlocked conditions independent of whether subjects stood with or without vision (Carpenter et al., 2010). Furthermore, it is likely that any small pressure changes that may have occurred between the body and the apparatus were insufficient to generate sensory illusions. This is because we would expect minimal pressure changes on the trunk for such small disturbances that occur in the current experiment, due to the large surface area of the back of the board and long lever arm of the rigid board when compared to the size and length of the feet (Fitzpatrick et al., 1992). The second goal of the study was to determine if increases in COP displacement with locking (Carpenter et al., 2010) could be explained by the presence of motor drift. We tested this argument by providing participants visual feedback of the COP during both the locked and unlocked conditions, which can minimize the potential for motor drift. Contrary to our hypothesis, we did not observe a decrease in COP displacement in the locked compared with unlocked condition when COP feedback was avail-
able. In fact, we observed increases in COP displacement with locking that were even greater in the feedback compared with no feedback condition. These results suggest that the increased amplitude of COP displacements observed with locking in this and prior studies cannot be explained by motor drift. Furthermore, there is no evidence that minimizing sway using our apparatus decreases postural sway to the extent that it could create a sensory deprived state, as angular displacements of the COM in the locked condition were still found to be within the known range of perceptual thresholds for stance (Fitzpatrick and McCloskey, 1994) in both the current (average COM angular range across conditions of 0.09⫾0.02°) and previous studies (Carpenter et al., 2010). The current results provide further support for a possible exploratory role of postural sway. The suggestion that there may be an exploratory role for motor variability has been described in a variety of motor control tasks. For example, when fixating the eyes on a visual target, rapid, involuntary eye movements known as microsaccades play a fundamental role in visual perception by ensuring continuous motion of the image projected onto the retina. Without this continuous motion, the visual environment rapidly fades from view (Martinez-Conde et al., 2006; Rolfs, 2009). Therefore, microsaccades ensure constant change in visual input, which is ideal since our nervous system has evolved to detect changes more readily than static stimuli. Similarly in postural control, postural sway is commonly viewed as error or noise in the system (Błaszczyk et al., 2009; Riccio, 1993; Riley et al., 1997; van Emmerik and van Wegen, 2002). However, some studies have theorized that sway variability may represent a perception-action strategy that allows one to acquire essential information about the interaction with the environment (Riccio, 1993; Riley et al., 1997). As described previously (Carpenter et al., 2010), this action may serve as a means to acquire a certain quality and quantity of sensory information. In fact, arguments to this effect have been applied to the larger sway observed in younger children during development, which is thought to facilitate the formation of “a reliable and stable sensorimotor relationship for postural control” (Chen et al., 2008). In this case, postural sway may act as a means of gathering and experiencing a variety of sensory interactions during stance, in order to improve the development of a proficient postural control system. Given that larger sway in children may represent the CNS’s attempt to couple sensorimotor action during stance, it is conceivable that smaller amounts of controlled sway may continue to be used by the CNS throughout life as a means to ensure a constant flow of sensory inputs. Likewise, a return to larger sway in older adults may represent a means for the CNS to generate more sensory information to compensate for the decreased speed and sensitivity of sensory systems in this population. As opposed to a negative consequence of a failing postural control system, the increased magnitude of sway in older populations may be an adaptation of similar origin to the larger sway observed in young children, and may be re-
C. D. Murnaghan et al. / Neuroscience 195 (2011) 54 –59
flective of the CNS’s attempt to re-establish or strengthen sensorimotor control. There are important clinical implications from the results of the current study. Our findings of increased COP displacement, observed despite providing convincing feedback to the subjects, provide further support for an exploratory role of postural sway. In contrast to this view, larger magnitudes of postural sway in populations with balance deficits due to age or disease are considered to be indicative of instability and greater fall risk. Therefore, they are often treated with rehabilitation and preventative strategies aimed at decreasing sway. However, if postural sway is in fact exploratory, and beneficial in acquiring sensory information to develop a better representation of where the body is in space, attempting to decrease sway may oppose the natural adaptations of the CNS that are required to maintain the integrity of the postural control system. Based on conflicting interpretations of increases in postural sway in these populations, it is imperative that we first develop an improved understanding of the origins of postural sway in general. If, as our results suggest, there is support for an exploratory role of postural sway, a serious re-evaluation of current clinical practices designed to deal with balance control deficits due to age or disease may be warranted. Acknowledgments—We gratefully acknowledge the funding support provided by the Natural Sciences and Engineering Research Council of Canada to MG Carpenter, BC Horslen, JT Inglis and CD Murnaghan.
REFERENCES Błaszczyk JW, Cies´linska-Swider J, Plewa M, Zahorska-Markiewicz B, Markiewicz A (2009) Effects of excessive body weight on postural control. J Biomech 42:1295–1300. Burgess PR (1989) Movement cues are necessary for load stabilization against low friction. Soc Neurosci Abstr 15:173. Carpenter MG, Murnaghan CD, Inglis JT (2010) Shifting the balance: evidence of an exploratory role of postural sway. Neuroscience 171:283–291. Carroll JP, Freedman W (1993) Nonstationary properties of postural sway. J Biomech 26:409 – 416. Chen LC, Metcalfe JS, Chang TY, Jeka JJ, Clark JE (2008) The development of infant upright posture: sway less or sway differently? Exp Brain Res 186:293–303.
59
Eklund G (1973) Further studies of vibration-induced effects on balance. Ups J Med Sci 78:65–72. Fitzpatrick RC, McCloskey DI (1994) Proprioceptive, visual and vestibular thresholds for the perception of sway during standing in humans. J Physiol 478:173–186. Fitzpatrick RC, Taylor JL, McCloskey DI (1992) Ankle stiffness of standing humans in response to imperceptible perturbation: reflex and task-dependent components. J Physiol 454:533–547. Gage WH, Winter DA, Frank JS, Adkin AL (2004) Kinematic and kinetic validity of the inverted pendulum model in quiet standing. Gait Posture 19:124 –132. Horak F, MacPherson J (1996) Postural orientation and equilibrium. In: Handbook of physiology (Rowell LB, Shepherd JT, eds). New York: Oxford University Press. Izumizaki M, Tsuge M, Akai L, Proske U, Homma I (2010) The illusion of changed position and movement from vibrating one arm is altered by vision or movement of the other arm. J Physiol 588: 2789 –2800. Lackner JR, Taublieb AB (1984) Influence of vision on vibrationinduced illusions of limb movement. Exp Neurol 85:97–106. Martinez-Conde S, Macknik SL, Troncoso XG, Dyar TA (2006) Microsaccades counteract visual fading during fixation. Neuron 49: 297–305. Maurer C, Peterka RJ (2005) A new interpretation of spontaneous sway measures based on a simple model of human postural control. J Neurophysiol 93(1):189 –200. Riccio GE (1993) In variability and motor control (Newell KM, Corcos DM, eds), pp 317–358. Champaign, IL: Human Kinetics. Riley MA, Mitra S, Stoffregen TA, Turvey MT (1997) Influences of body lean and vision on unperturbed postural sway. Mot Control 1: 229 –246. Rolfs M (2009) Microsaccades: small steps on a long way. Vision Res 49:2415–2441. Rothwell JC, Traub MM, Day BL, Obeso JA, Thomas PK, Marsden CD (1982) Manual motor performance in a deafferented man. Brain 105:515–542. Sanes JN, Mauritz KH, Evarts EV, Dalakas MC, Chu A (1984) Motor deficits in patients with large-fiber sensory neuropathy. Proc Natl Acad Sci U S A 81:979 –982. Sasaki M, Mishima H, Suzuki K, Ohkura M (1995) Observations on micro-exploration in everyday activities. In: Studies in perception and action, Vol. III (Bardy BG, Bootsma RJ, Guiard Y, eds), pp 99 –102. Hillsdale, NJ: Lawrence Erlbaum Associates. van Emmerik RE, van Wegen EE (2002) On the functional aspects of variability in postural control. Exerc Sport Sci Rev 30:177–183. Winter DA, Patla AE, Prince F, Ishac M, Gielo-Perczak K (1998) Stiffness control of balance in quiet standing. J Neurophysiol 80:1211–1221.
(Accepted 9 August 2011) (Available online 12 August 2011)