Voluntary and Involuntary Postural Responses to Imposed Optic Flow

Transcription

1 Motor Control, 2006, 10, Human Kinetics, Inc. Voluntary and Involuntary Postural Responses to Imposed Optic Flow Thomas A. Stoffregen, Philip Hove, Jennifer Schmit, and Benoît G. Bardy We demonstrated that postural responses to imposed optic flow are to some extent voluntary. In a moving room, participants either stood normally or were instructed to resist any influence of visible motion on their stance. When participants attempted to resist, coupling of body sway with motion of the room was significantly greater than when the eyes were closed, but was significantly reduced relative to coupling in the normal stance condition. The results indicate that the use of imposed optic flow for postural control is not entirely automatic or involuntary. This conclusion motivates a search for non-perceptual factors that may influence the degree to which body sway is coupled to imposed optic flow. Key Words: stance, orientation, visual perception, motor control, perception and action Ordinary body sway consists primarily of small fore-aft motions of the body s center of mass relative to the gravito-inertial force environment. These motions create a slight displacement of the head and eyes relative to the illuminated environment. This displacement produces optic flow. When the ground surface and the illuminated environment are stationary, parameters of the optic flow, such as its direction, amplitude, and frequency, are deterministically related to the corresponding parameters of body sway. For this reason, the optic flow provides information about body sway that can be useful in controlling sway. One prominent method for studying relations between vision and body sway is the moving room paradigm (Lee & Lishman, 1975; Schöner, 1991). In this paradigm, researchers create optic flow that simulates the optical consequences of body sway, and expose standing participants to the simulation. When standing persons are exposed to experimentally-generated optic flow they tend to respond by coupling body sway to the flow (Andersen & Dyre, 1989; Dijkstra, Schöner, & Gielen, 1994; Higgins, Campos, & Kermoian, 1996; Lestienne, Soechting, & Berthoz, 1977; Lee & Lishman, 1975; Stoffregen, 1985, 1986); this is true also for optic flow presented to persons walking on a treadmill (Bardy, Warren, & Kay, Stoffregen is with the School of Kinesiology, University of Minnesota, Minneapolis, MN Hove and Schmit are with the Department of Psychology, University of Cincinnati, Cincinnati, OH Bardy is with the Institut Universitaire de France and the University of Monpellier-1, Faculty of Sport and Movement Sciences, Montpellier, France. 24

2 Postural Responses to Optic Flow , 1999; Warren, Kay, & Yilmaz, 1996). Robust coupling persists across wide variation in parameters of the imposed optic flow, such as its direction and amplitude. Researchers have sought to determine limits to coupling in terms of boundary values of optic flow parameters, such as velocity (Stoffregen, 1986), direction (Warren et al.), visual angle (Andersen & Dyer, 1989; Stoffregen, 1985), and frequency (Dijkstra, Schöner, Giese, & Gielen, 1994). Theories relating posture to vision have focused on the temporal stability between imposed optic flow (often referred to as a driver ) and postural actions (e.g., Dijkstra, Schöner, & Gielen, 1994; Jeka, 1995; Schöner, 1991; Warren et al., 1996). These theories seek to understand and predict deviations from perfect temporal stability, based on manipulations of parameters of imposed optic flow. This approach has been highly successful, and has led to many important insights about perceptual and motor dynamics. A consistent finding of research relating postural control to optic flow is that body sway will become coupled to imposed optic flow having a large visual angle, oscillating in the range of Hz along the body s anterior-posterior axis (e.g., Dijkstra, Schöner, & Gielen, 1994; Dijkstra, Schöner, Giese, & Gielen 1994; Lee & Lishman, 1975; Stoffregen, 1985, 1986). The frequent replication of this effect may suggest that it is a general phenomenon. However, there have been few attempts to determine whether the effect is general across variations in participants intentions. In this study, we assessed its generality across variation in participants intentions. Involuntary Postural Responses to Imposed Optic Flow? Some researchers have argued that the use of imposed optic flow for postural control is automatic, or reflexive. For example, Bronstein and Buckwell (1997, p. 248), argued that postural responses to optic flow will occur when visual-motion stimuli mimic those encountered during body sway, and concluded that postural control relies on processing and motor responses that appear to be automatic. A detailed interpretation of this claim is difficult because Bronstein and Buckwell did not define automatic. An activity may be regarded as automatic if it is involuntary, if it occurs outside of consciousness, if it does not draw on central cognitive resources, if the activity is not affected by variation in other aspects of behavior, and so on. The problem of definition applies also to suggestions that the control of stance is reflexive. White, Post, and Leibowitz (1980) stated that upright stance is controlled reflexively, but did not define this term. One widely accepted aspect of the concept of automaticity relates to our ability to exercise voluntary control over activities or processes: An activity or process may be said to be automatic if we cannot exercise voluntary control over it. To our knowledge, only one experiment has been conducted with the specific intent of determining whether people can voluntarily resist the effects of imposed optic flow. Lee and Lishman (1975) reported an experiment in which participants were instructed; to do [their] best to resist imposed optic flow created by a moving room (Lee & Lishman, p. 90). Despite this instruction, posture was significantly influenced by room movements, leading Lee and Lishman (p. 91) to conclude that the imposed optic flow could not be ignored. We know of no direct replications

3 26 Stoffregen et al. of this important result. The current study was designed partly to attempt a replication, and partly to extend and clarify the earlier finding. A related issue was addressed by Oullier, Bardy, Stoffregen, & Bootsma (2002), who measured postural control in a moving room. Participants were instructed to simply look at a target on the wall of the room (the Looking task), or were asked to move their heads forward and backward in time with motion of the room, so as to maintain a constant distance between the target and their head (the Tracking task). The cross-correlation between the head and the room was significantly greater during the tracking task than during the looking task (separate measures of hip motion exhibited essentially the same results). These results suggest that coupling of body sway to imposed optic flow can be influenced on a volitional basis. However, Oullier et al. did not ask participants to ignore imposed optic flow. In Lee and Lishman s (1975) study it was not possible to test the hypothesis that participants had partial control over their postural responses to imposed optic flow. This was for two reasons. First, Lee and Lishman used different motions of the room in different conditions. When participants were not instructed to resist, the room was moved in a smooth, sinusoidal manner with amplitude of 0.6 cm but when they were instructed to resist, the room was moved in an irregular manner with amplitude of 5.0 cm. In the present study room movements were identical in all conditions, making it possible to conduct direct comparisons between conditions. Second, the dependent variable used by Lee and Lishman provided a simplified description of the coupling between room motion and body sway. Lee and Lishman computed D, the percentage of time that body movement was in the same direction as room motion. An advantage of the present study is the use of dependent variables (i.e., cross-correlation, phase) that permit analysis of dynamical coupling of body sway to room movement. Cross-correlation resembles Lee and Lishman s D, but is more precise, while phase provides information not captured by D. We measured coupling in conditions in which participants either did or did not attempt to resist the influence of imposed optic flow on their posture. Following Lee and Lishman (1975), we predicted that there would be a significant influence of imposed optic flow on body sway in each experimental condition, which would indicate that participants could not, by an act of will, wholly suppress their postural responses to imposed optic flow. In addition, we compared the two experimental conditions to each other. This comparison allowed us to determine whether participants had the ability partially to resist the imposed optic flow. We predicted that participants would be able to reduce the strength of the coupling through the intention to resist. Participants Method Participants were undergraduate students of the University of Cincinnati who participated in return for course credit. Twenty-seven people participated, ranging in age from 18 to 25, with a mean of 19.8 years, and in height from 1.47 m to 2.08 m, with a mean of 1.70 m. Fifteen people were randomly assigned to the experimental group, and twelve to the control group.

4 Postural Responses to Optic Flow 27 For all participants vision was either normal or corrected to normal (with either contact lenses or glasses). No participant reported any history of dizziness or problems with balance. They were naïve to the purpose of the study before participating. All of the participants completed all of the trials. None reported any symptoms of motion sickness. Apparatus Optic flow was generated using a moving room, consisting of a cubical frame, 2.4 m on each side (Figure 1). It was constructed of sheets of plywood fixed to a frame of aluminum beams. There were four sidewalls and a ceiling. The lower part of the back wall was left open to provide access to the room. The interior surfaces of the front and sidewalls were covered with marble-pattern adhesive paper. The room was mounted on wheels, which rested on rails, and was moved by a torque motor under computer control. During data collection, participants stood on the concrete laboratory floor facing the front wall, and the room moved around them. A fluorescent light fixture attached to the center of the ceiling of the moving room provided illumination. The fixture extended only a few centimeters from the ceiling, and the walls were flat. Thus, shadows in the room were minimized. Data on postural motion were collected using an electromagnetic tracking system (Flock of Bird, Ascension Technologies, Inc.). A magnetic field was created by a transmitter located on a stand behind the participant s head. Participants were required to wear a bicycle helmet. One receiver was attached to the helmet and another to the moving room. Motion of the receivers within the magnetic field was detected by the system, sampled at a rate of 50 Hz, and recorded on a computer for later analysis. Figure 1 The moving room. Procedure Participants stood in the moving room in their bare feet, to avoid possible effects of shoe structure (e.g., high heels) on postural sway. Participants entered the room through the opening of the back wall and placed their heels on a marker on the floor (i.e., the feet were side by side) so that they were facing the front wall; the marker

5 28 Stoffregen et al. was 1.5 m from the front wall. Participants were allowed to choose the place they wanted to put their hands (e.g., in front of their body, in their pockets or at the back of their body), but were asked to keep their hands at the same place during trials, and not to move their feet during trials. They were also instructed to try to minimize head movements (i.e., turning the head relative to the body) during trials. The room s motion was a simple 0.2-Hz oscillation, with an amplitude of 2 cm peak to peak. There was no fixation point, or marked target; participants were simply asked to look straight ahead at the front wall. Participants were divided into two groups, the experimental group and the control group. For the experimental group, the experiment consisted of nine trials. Eight were experimental trials and one was the Eyes Closed trial, the purpose of which was to confirm that any effects of room motion on postural sway were due solely to vision. The sequence of experimental trials was the same for each participant. In the first four experimental trials (the Normal condition) participants were instructed only to stand comfortably and look straight ahead. They were not given any instructions relating room motion to body sway. In the second four experimental trials (the Resist condition), participants were reminded that the room would be moving. They were instructed to resist this motion, that is, to try to maintain stationary stance and not to move with the room. The Resist condition was placed second to avoid any possible contamination of postural responses in the Normal condition arising from knowledge of the Resist condition. In the Eyes Closed trial participants were instructed to stand with eyes closed as the room moved around them. The Eyes Closed trial occurred in a randomly chosen position within the sequence of trials. For the control group, the experiment consisted of eight trials. Each trial was in the Normal condition. Data from the control group allowed us to evaluate the hypothesis that coupling would change across trials (i.e., an effect of experience, or fatigue), independent of any other manipulation. For both the experimental and control groups, we recorded anterior-posterior motion of the head and room. Each trial lasted 80 s. The first 20 s were used to begin room motion, with the amplitude ramping gradually from 0 to 2 cm. Data were collected for the latter 60 s of each trial. Design and Data Analysis Two variables were used to describe the relationship between body sway and room motion. These were the maximum cross-correlation between room motion and body sway (cross-correlation), and phase. For each trial, we calculated the cross-correlation of body sway and room motion at each of 101 time lags, from Lag = 50 to Lag = +50 samples (equivalent to lags of up to 1.0 s before and after 0). We then chose the correlation coefficient with the largest absolute value. The value of phase, measured in degrees (from 180 to +180 ), indicates the nature of temporal synchrony between room motion and body sway (in the AP axis). Negative values of phase indicate that body sway lags behind room movement, while positive values indicate that body sway anticipates room motion. A phase of 0 means that that there is perfect temporal synchrony. We used the point estimate of phase (Dijkstra, Schöner, & Gielen, 1994). A Fisher transformation was done on the cross-correlation data to normalize their distribution. Then, a repeated-measures analysis of variance was performed on the cross-correlation data. For cross-correlation data, inferential statistics were

6 Postural Responses to Optic Flow 29 computed on the Fisher-transformed means; untransformed means are reported in the text and tables. Measures of effect size are reported in terms of the percentage of the total variance that was accounted for by each effect. The phase data were analyzed using circular statistics (Batschelet, 1981). The Watson-Williams F-test, used for circular statistics, requires a separate analysis for each main effect, and does not permit analysis of interactions. Results Data from the experimental and control groups were analyzed separately, in each case using a one-within design. For the experimental group, the within-groups variable was experimental condition, which had three levels: Eyes Closed, Normal, and Resist. For the control group, the within-groups variable was trial block (trials 1-4 versus trials 5-8). Generally, the results from the experimental group indicated that participants were able, by an act of will, to reduce significantly (but not eliminate) coupling of body sway with imposed optic flow. In the control group coupling did not change across trial blocks, indicating that the condition effects in the experimental group can be taken at face value. Table 1 Means and Standard Deviations (In Parenthesis) Showing Coupling Between Motion of the Room and the Head, for the Experimental Group. For Phase, Circular Standard Deviations Are Given. Cross-correlation Phase Eyes closed (0.147) (90.13 ) Normal (0.122) (20.86 ) Resist (0.095) (17.86 ) Experimental Group The data for the experimental group are summarized in Table 1. Cross-correlation. We found a significant main effect of condition, F(2, 28) = , p <.05, accounting for 84.6% of the variance (using the partial eta-squared statistic). Planned comparisons were done to examine differences between conditions. In the Normal condition, cross-correlation was significantly greater than in the Eyes Closed condition, t(14) = 7.869, p <.05. Similarly, cross-correlation in the Resist condition was greater than in the eyes closed condition, t(14) = 5.756, p <.05. These results indicate that in the experimental conditions participants sway was coupled to the optic flow created by room motion. When instructed to resist, participants could not completely suppress their postural responses to the imposed optic flow, consistent with the findings of Lee and Lishman (1975). However, the instruction to resist did produce a significant reduction in coupling.

7 30 Stoffregen et al. There was a difference between experimental conditions, with cross-correlation being higher in the Normal condition than in the Resist condition, t(14) = 5.492, p <.05. Phase. Watson-Williams F-tests for two circular means showed that the mean phase in the Eyes Closed condition differed from the mean phase in each of the experimental conditions (eyes closed vs. normal: F(1, 73) = 9.32; eyes closed vs. resist: F(1,73) = 9.92, each p <.05), indicating that motion of the room had a visual effect on temporal synchrony. In experimental conditions, phase was close to 10, indicating that participants were following the motion of the room and were not anticipating, consistent with earlier results (e.g., Bardy et al., 1996, 1999; Dijkstra, Schöner, & Gielen, 1994; Oullier, Bardy, Stoffregen, & Bootsma, 2002). The difference between the Normal and Resist conditions was not significant, Watson-Williams F(1, 118) < 1. Sway Amplitude. We were concerned about the possibility that in the Resist condition participants might reduce coupling with imposed optic flow simply by reducing postural sway, in general. Such an outcome would be relatively uninteresting, because it would suggest that participants simply tried to reject all available information (i.e., to stiffen up), rather than to ignore selectively the imposed optic flow. Much more interesting would be the possibility that participants might selectively ignore the imposed optic flow (as instructed), while maintaining their overall level of postural activity. To evaluate this possibility, we contrasted the amplitude of sway in the Normal and Resist conditions, measuring amplitude relative to the earth, rather than relative to room motion. Sway amplitude was estimated by the standard deviation of body position in the anterior-posterior axis (e.g., Stoffregen, Smart, Bardy, & Pagulayan, 1999). The mean values of the standard deviation of body position for the Normal and Resist conditions were cm, and cm, respectively. These did not differ [(t(14) = 0.291, p >.05)]. The insignificant difference in the standard deviation of sway between the Normal and Resist conditions indicates that in the Resist condition participants did not reduce their overall level of body sway which, in turn, implies that the significant reduction in coupling in the Resist condition resulted from a selective resistance to imposed optic flow, per se. Control Group Data for the control group are summarized in Table 2. With the control group (as with the experimental group; Table 1), we compared the mean of trials 1-4 with the mean of trials 5-8, for cross-correlation and phase. There were no significant differences between trials 1-4 and 5-8. Table 2 Means (Standard Deviations) for the Control Group Cross-correlation Phase Trials (0.189) (40.70 ) Trials (0.219) (23.10 )

8 Postural Responses to Optic Flow 31 Discussion There are several important results. First, participants who were instructed to stand comfortably exhibited strong coupling of body sway with the moving room, replicating several previous studies (e.g., Lee & Lishman, 1975; Stoffregen, 1985). Second, participants were not able to eliminate postural responses to imposed optic flow when they made a deliberate attempt to do so, replicating Lee and Lishman. This is shown by the fact that the value of cross-correlation in the Resist condition was significantly greater than in the Eyes Closed condition, while the value of phase in the Resist condition differed from phase in the Eyes Closed condition. Third, despite being unable to eliminate completely coupling of body sway with the imposed optic flow, participants were able, by an act of will, significantly to reduce the strength of the coupling. This is demonstrated by the fact that in the experimental group cross-correlation was significantly lower in the Resist condition than in Normal condition. This third result sheds new light on the results and interpretation of Lee and Lishman (1975). By an act of will people can reduce coupling of body sway with imposed optic flow. Our finding that coupling of body sway with imposed optic flow can be influenced by a person s conscious intention suggests that it may be inappropriate to refer to the coupling as being automatic or reflexive (e.g., Bronstein & Buckwell, 1997; Sveistrup & Woollacott, 1996; White et al., 1980). This is consistent with Woollacott and Shumway-Cook (2002), who concluded that postural control is not a reflexive, automatic behavior. More generally, the data suggest that Lee and Lishman s conclusion that imposed optic flow could not be ignored may be too strong. The data are consistent with a more conservative interpretation, for example, that imposed optic flow can be ignored partially, but not entirely. This interpretation raises questions about the circumstances that might influence the extent to which body sway is influenced by imposed optic flow. For three decades, one of the major methods used in the study of postural control has been measurement of postural responses to optic flow created by experimenters and imposed on standing participants. The prevalence of this method is due in part to the fact that imposed optic flow tends to elicit robust postural responses (e.g., Lee & Lishman, 1975; Schmuckler, 1997; Stoffregen, 1985; Van Asten et al., 1988). In recent years, this moving room paradigm has become the primary locus of research examining the dynamics of perception-action coupling in postural control (Dijkstra, Schöner, & Gielen, 1994; Jeka, 1995; Schöner, 1991; Warren et al., 1996). It has become commonplace to assume that the influence of imposed optic flow on the control of stance is automatic, meaning either that it is involuntary, that it is independent of other tasks or activities, or both (e.g., Bronstein & Buckwell, 1997). The present findings support the idea that postural motion is generated actively (e.g., Dijkstra, Schöner, & Gielen, 1994; Jeka, 1995; Warren et al., 1996), but cast this idea in a new light. It has been suggested that the dynamics of postural control change as the behavioral situation changes, (Giese, Dijkstra, Schöner, & Gielen, 1996, p. 428). However, the kinds of changes that have been incorporated in recent models have been limited to changes in the available sensory information, (Giese et al., p. 428). Our findings indicate that the active control of posture may not be driven by changes in information in any simple sense. We observed

9 32 Stoffregen et al. changes in postural control when there was no variation in optic flow, but only a variation in participants intentions. The influence of conscious intention on postural responses to imposed optic flow may be consistent with a broad range of recent research which has demonstrated that postural control can be influenced by perceptual and cognitive factors that are nominally distinct from postural activity, such as performance of visual and memory tasks (e.g., Riley, Baker, & Schmit, 2003; Stoffregen, Pagulayan, Bardy, & Hettinger, 2000; Stoffregen et al., 1999; Woollacott and Shumway-Cook, 2002). Conclusion We have demonstrated that the use of imposed optic flow for the control of standing posture is not automatic. Participants were able to reduce the influence of imposed optic flow on stance through a conscious act of will, indicating at least some role for consciousness in the perceptual control of stance. The present study suggests that postural control is not automatically driven by optic flow. This result raises questions about the factors that may influence the extent to which sway is coupled to imposed optic flow. This is a subject for future research. Author s Note Portions of the data were reported at the XI International Conference on Perception and Action, Storrs, CT, June Acknowledgments Preparation of this article was supported by Enactive Interfaces, a network of excellence (IST contract #002114) of the Commission of the European Community, and by the National Science Foundation (INT , BCS ), with additional support from the University of Paris XI, and the Institut Universitaire de France. We extend our grateful thanks to Nat Hemasilpin for motion control programming and control systems engineering, to Olivier Oullier for laboratory assistance, and to Agatha Ritchey, Sarah Donahue, Melissa Watson, and Justin Schneider, for help with data collection and analysis. References Andersen, G.J., & Dyre, B.P. (1989). Spatial orientation from optic flow in the central visual field. Perception & Psychophysics, 45, Bardy, B.G., Warren, W.H., & Kay, B. (1996). Motion parallax is used to control postural sway during walking. Experimental Brain Research, 11, Bardy, B.G., Warren, W.H., & Kay, B. (1999). The role of central and peripheral vision in postural control during walking. Perception & Psychophysics, 61, Batschelet, E. (1981). Circular statistics in biology. New York: Academic Press. Bronstein, A.M., & Buckwell, D. (1997). Automatic control of postural sway by visual motion parallax. Experimental Brain Research, 113, Dijkstra, T.M.H., Schöner, G., & Gielen, C.C.A.M. (1994). Temporal stability of the actionperception cycle for postural control in a moving visual environment. Experimental Brain Research, 97,

10 Postural Responses to Optic Flow 33 Dijkstra, T.M.H., Schöner, G., Giese, M.A., & Gielen, C.C.A.M. (1994). Frequency dependence of the action-perception cycle for postural control in a moving visual environment: Relative phase dynamics. Biological Cybernetics, 71, Giese, M.A., Dijkstra, T.M.H., Schöner, G., & Gielen, C.C.A.M. (1996). Identification of the nonlinear state space dynamics of the action-perception cycle for visually induced postural sway. Biological Cybernetics, 74, Higgins, C.I., Campos, J.J., & Kermoian, R. (1996). Effect of self-produced locomotion on infant postural compensation to optic flow. Developmental Psychology, 32, Jeka, J. (1995). Is servo-theory the language of human postural control? Ecological Psychology, 7, Lee, D.N., & Lishman, J.R. (1975). Visual proprioceptive control of stance. Journal of Human Movement Studies, 1, Lestienne, F., Soechting, J., & Berthoz, A. (1977). Postural readjustments induced by linear motion of visual scenes. Experimental Brain Research, 28, Oullier, O., Bardy, B.G., Stoffregen, T.A., & Bootsma, R.J. (2002). Postural coordination in looking and tracking tasks. Human Movement Science, 21, Riley, M.A., Baker, A.A., & Schmit, J.M. (2003). Inverse relation between postural variability and difficulty of a concurrent short-term memory task. Brain Research Bulletin, 62, Schmuckler, M.A. (1997). Children s postural sway in response to low- and high-frequency visual information for oscillation. Journal of Experimental Psychology: Human Perception and Performance, 23, Schöner, G. (1991). Dynamic theory of action-perception patterns: The moving room paradigm. Biological Cybernetics, 64, Stoffregen, T.A. (1985). Flow structure versus retinal location in the optical control of stance. Journal of Experimental Psychology: Human Perception and Performance, 11, Stoffregen, T.A. (1986). The role of optical velocity in the control of stance. Perception & Psychophysics, 39, Stoffregen, T.A., Pagulayan, R.J., Bardy, B.G., & Hettinger, L.J. (2000). Modulating postural control to facilitate visual performance. Human Movement Science, 19, Stoffregen, T.A., Smart, L.J., Bardy, B.G., & Pagulayan, R.J. (1999). Postural stabilization of looking. Journal of Experimental Psychology: Human Perception & Performance, 25, Sveistrup, H., & Woollacott, M.H. (1996). Longitudinal development of the automatic postural response in infants. Journal of Motor Behavior, 28, Van Asten, W.N.J.C., Gielen, C.C.A.M., & Denier van der Gon, J.J. (1988). Postural adjustments induced by simulated motion of differently structured environments. Experimental Brain Research, 73, Warren, W.H., Kay, B., & Yilmaz, E. (1996). Visual control of posture during walking: Functional specificity. Journal of Experimental Psychology: Human Perception and Performance, 22, White, K.D., Post, R.B., & Leibowitz, H.W. (1980). Saccadic eye movements and body sway. Science, 208, Woollacott, M., & Shumway-Cook, A. (2002). Attention and the control of posture and gait: A review of an emerging area of research. Gait and Posture, 16, 1-14.