Psychological Research (2007) 71: 192–200 DOI 10.1007/s00426-005-0018-2
O R I GI N A L A R T IC L E
Nicolas Vuillerme Æ Gilel Nafati
How attentional focus on body sway affects postural control during quiet standing
Received: 26 October 2004 / Accepted: 3 August 2005 / Published online: 8 October 2005 Springer-Verlag 2005
Abstract The purpose of this study was to investigate how attentional focus on body sway affects postural control during quiet standing. To address this issue, sixteen young healthy adults were asked to stand upright as immobile as possible on a force platform in both Control and Attention conditions. In the latter condition, participants were instructed to deliberately focus their attention on their body sways and to increase their active intervention into postural control. The critical analysis was focused on elementary motions computed from the centre of pressure (CoP) trajectories: (1) the vertical projection of the centre of gravity (CoGv) and (2) the difference between CoP and CoGv (CoP–CoGv). The former is recognised as an index of performance in this postural task, whilst the latter constitutes a fair expression of the ankle joint stiffness and is linked to the level of neuromuscular activity of the lower limb muscles required for controlling posture. A frequency-domain analysis showed increased amplitudes and frequencies of CoP–CoGv motions in the Attention relative to the Control condition, whereas non-significant changes were observed for the CoGv motions. Altogether, the present findings suggest that attentional focus on body sway, induced by the instructions, promoted the use of less
N. Vuillerme (&) Laboratoire de Mode´lisation des Activite´s Sportives, Universite´ de Savoie, Domaine Universitaire, 73 376 Le Bourget du Lac cedex, France E-mail:
[email protected] Tel.: +33-4-79758115 Fax: +33-4-79758148 N. Vuillerme Laboratoire TIMC-IMAG, UMR CNRS 5525, Equipe AFIRM, Grenoble, France G. Nafati Laboratoire de Plasticite´ et PhysioPathologie de la Motricite´, UMR 6196, CNRS/Universite´ de la Me´diterrane´e, Me´diterrane´e, France
automatic control process and hampered the efficiency for controlling posture during quiet standing.
Introduction In recent years, a growing number of investigations have reported that the participant’s focus of attention, induced by the instructions, can play a significant role in the performance and learning of motor skills (see Wulf & Prinz, 2001, for a review), including those requiring postural control (e.g., McNevin, Shea, & Wulf, 2003; Wulf, Ho¨ß, & Prinz, 1998; Wulf, McNevin, & Shea, 2001; Wulf, Mercer, McNevin, & Guadagnoli, 2004; Wulf, Shea, & Park, 2001). In general, results of these studies suggested that providing instructions that direct the performer’s attention to the effects of his or her movement (‘‘external’’ focus of attention) was more beneficial than directing his or her attention to their own movements (‘‘internal’’ focus of attention). Furthermore, there is some evidence that providing instructions related to the performers’ body movements was not always optimal, and more interestingly, could degrade the execution of automated skills, when compared with no instructions at all (e.g., Wulf et al. 1998; Wulf & Weigelt, 1997). A ‘‘constrained action hypothesis’’ (e.g., McNevin et al. 2003; Wulf et al. 2001; Wulf & Prinz, 2001) has recently been proposed to account for these observations. According to this hypothesis: 1. performers focussing on their body movements, or on the effects that occur in close proximity to their body, would tend to actively intervene in the maintenance of a stable posture more that performers focussing on a more distant effect would do; 2. trying to consciously control one’s movement would constrain the motor system by interfering with automatic motor control processes that would ‘‘normally’’ regulate the movement;
193
3. focussing on the movement effect would allow the motor system to more naturally self-organise, unconstrained by the interference caused by conscious control attempts and would result in a more effective postural performance. Note that this latter assumption has been confirmed in a recent experiment evidencing reduced attentional demands, assessed by shorter probe reaction time (RT) (e.g., Abernethy, 1988; Kahneman, 1973), for participants balancing on the stabilometer with an external focus of attention (i.e. focus on the markers attached to the stabilometer platform) compared to those with an internal focus of attention (i.e. focus on their feet) (Wulf et al. 2001). Interestingly, most of the studies mentioned above have reported effects of different attentional foci on postural task performance, i.e. in terms of using outcome measures (e.g., Magill, 2001), such as the movements of the stabilometer platform participants were required to minimise from the horizontal. On the contrary, little is known on how the performer’s attention focus affects the control of posture, in terms of performance production measures (e.g., Magill, 2001). In this context, the purpose of the present experiment was to investigate how internal attentional focus on body sway control, induced by the instructions given to the participants, affects not only postural performance during quiet standing (i.e. outcome measures), but also neuromuscular requirements for ensuring standing control (i.e. performance production measures). To achieve this goal, an experimental protocol consisting in the evaluation of postural control during bipedal quiet standing was chosen for the purpose of the use of a discriminating method of analysis of the centre of foot pressure (CoP) trajectories (Rougier & Caron, 2000). This method consisted in the dissociation of CoP trajectories into two superimposed elementary components: the centre of gravity vertical projection (CoGv) and the difference between the centre of foot pressure and the centre of gravity vertical projection (CoP–CoGv) displacements, presenting specific attributes in undisturbed stance control. On the one hand, the CoGv, corresponding to the barycentre of the centres of mass from the different segments, is thought to be the controlled variable in postural control and is recognised as an index of postural performance in this particular task (e.g., Cle´ment, Gurfinkel, Lestienne, Lipshits, & Popov, 1984; Gollhofer, Horstmann, Berger, & Dietz, 1989; Horstmann & Dietz, 1990; Massion, 1994; Winter, 1995; Winter, Patla, Prince, Ishac, & Gielo-Perczak, 1998). On the other hand, as long as upright standing can be modelled as a one-link inverted pendulum (e.g., Winter et al. 1998; Gage, Winter, Frank, & Adkin, 2004), the CoP–CoGv constitutes a fair expression of the ankle joint stiffness (Caron, Gelat, Rougier, & Blanchi, 2000; Winter et al. 1998) and is assumed to be linked to the level of neuromuscular activity of the lower limb muscles required for controlling undisturbed upright stance (e.g.,
Nault, Allard, Hinse, Le Blanc, Caron, Labelle, & Sadeghi, 2002; Rougier, 2003; Rougier, Burdet, Farenc, & Berger, 2001, 2004). Following the ‘‘constrained action hypothesis’’ (e.g., McNevin et al. 2003; Wulf et al. 2001; Wulf & Prinz, 2001) according to which an attentional focus directed toward the body movements (i.e. postural sway) engages participants to exert a relatively more conscious control over the regulatory processes involved in balance control, it was hypothesised that attentional focus on body sway would impair postural control. Precisely, the analysis of both the CoGv and CoP–CoGv motions should indicate to which extent internal attentional focus induced by the instructions given to the participants would modify (1) postural performance and (2) the resultant joint stiffness and neuromuscular requirements for ensuring standing control, respectively.
Method Participants Sixteen young healthy adults (mean age = 26.0±4.4 years; mean body weight = 73.1±10.0 kg; mean height = 178.4±7.8 cm) participated in the experiment. They gave written consent to the experimental procedure as required by the Helsinki declaration (1964) and the local Ethics Committee. None of the participants presented any known motor problems, neurological disorders or vestibular impairment. Finally, all participants had normal or corrected-to-normal vision. Task and procedures Participants stood barefoot as immobile as possible on a force platform (Equi+, model PF01, Aix les Bains, France), in a natural position (feet abducted at 30, heels separated by 3 cm), their arms hanging loosely by their sides, looking at an eye level black cross (20·25 cm) located 1.20 m ahead. Signals from the force platform were sampled at 64 Hz, amplified and converted from analogue to digital form. The CoP trajectories were processed, as seen above, in different ways. In the coordinate system used, ML and AP characterise mediolateral and antero-posterior axes, respectively. Two experimental conditions were presented. In the so-called Control condition, participants stood upright without any specific instruction concerning their attentional focus of attention. In the co-called Attention condition, participants were instructed to deliberately focus their attention on their body sways and to increase their active intervention into postural control. This instruction was always given before the trial was initiated. This condition was intended for participants to consciously intervene in postural control. Each experimental condition included three trials of 32 s, which has been shown to be sufficient for reliable postural sway measures
194
(LeClair & Riach, 1996). In addition, a longer time recording in our mind might have distorted the postural behaviour by introducing some spurious effects such as mental fatigue in the Attention condition. Rests of 1 and of about 5 min were allowed between each trial and each condition, respectively. The order of presentation of the two Control and Attention conditions was counterbalanced between participants. To ensure that (1) participants actually adopt the attentional focus they were instructed to use and (2) the instructions given to the participants were successful to induce the participants to allocate more attention into the postural task in the Attention than in the Control condition, two methodological precautionary measures were taken. On the one hand, after the completion of each trial, participants were asked to provide subjective ratings of degree of attentional focus and active involvement into body sway control. The ratings were made using a 7-point numerical rating scale anchored with (1) ‘‘completely uninvolved, not trying hard at all’’ to (7) ‘‘extremely involved, trying as hard as possible’’. On the other hand, a preliminary control experiment, performed by 8 of the 16 participants involved in the main experiment, was designed to investigate whether the Attention condition promotes less automatic control process than does the Control condition proposed in the main experiment. A secondary probe RT task was used to determine the amount of attention required performing the postural task in the two experimental conditions. Participants were asked to respond vocally as quickly as possible to an unpredictable auditory stimulus while performing the postural task in the two Control and Attention conditions. For each trial, five auditory stimuli separated by at least 2 s were presented randomly. The number and timing of the stimuli delivered were similar for each experimental condition. These experimental conditions, including three trials of 32 s, were successively and randomly performed. Within this so-called ‘‘dual-task paradigm’’, any change in RT presumably would reflect changes in the resources necessary for performing the postural task (e.g., Abernethy, 1988; Kahneman, 1973). If the Attention condition promotes the use of more conscious postural control and less automatic control process, performance in this condition should require more attention and therefore should yield slower probe RTs than performance in the Control condition (Wulf et al. 2001). Signal processing The signal processing used in this study have been detailed in previous reports (e.g., Rougier, 2003; Rougier et al. 2001, 2004; Rougier & Caron, 2000; Rougier & Farenc, 2000). Only the main points will thus be given here. CoGv and CoP–CoGv motions were determined from the CoP trajectories computed from the force platform. A relationship between the amplitude ratio of the vertical projection of the centre of gravity (CoGv) and CoP
motions (CoGv /CoP) and sway frequencies allows determining the CoGv and consequently CoP–CoGv motions. Body sways being particularly reduced, standing still can therefore theoretically be modelled as a one-link inverted pendulum (e.g., Winter et al. 1998; Gage et al. 2004), where CoGv and CoP behave as periodic functions in phase with each other. The method, initially proposed by Breniere (1996) for gait and then extended to standing posture by Caron, Faure and Breniere (1997), is given by the following formula: CoGv X2 ¼ 2 0 2 ; CoP ðX0 þ X Þ where W=2 p f is the pulsation (rad s1) and W 0= [mgh / (IG+mh2)]1/2 (Hz), termed natural body frequency, is a biomechanical constant relative to the anthropometry of the subject (m, g, h, IG: mass of the subject, gravity acceleration, distance from CG to the ground, and moment of body inertia around the ML or AP axis with respect to the CG). Depending on the axis, two distinct relationships were used to characterise the subject’s anthropometry since the moments of inertia are slightly different. According to Ledebt and Breniere (1994), these moments of inertia can be deduced from the following relationships: IG ML ¼ 0:0572 mHs2 and IG AP ¼ 0:0533 mHs2 ; where Hs represents the height of the subjects. From this CoGv /CoP relationship, it is therefore relevant to consider that CoP oscillations operating over too high frequencies would not incur appreciable CoGv movements. The principle of this model is that the body constitutes a low-pass filter, which would explain the loss in amplitude observed between CoP and CoGv as the sway frequency increases. As previously detailed (e.g., Rougier, 2003; Rougier et al. 2001, 2004; Rougier & Caron, 2000; Rougier & Farenc, 2000), the CoGv calculation consists in multiplying the data, transformed in the frequential domain through a fast fourier transform (FFT), by the above-mentioned filter and recovering to the time domain by processing an inverse FFT (Fig. 1). Dependent variables RT (in ms) served for determining the attentional demand associated with regulating postural sway in both Attention and Control conditions. RT was defined as the temporal interval between the presentation of the auditory stimulus and the subjects’ verbal responses. CoGv and CoP–CoGv motions were processed through a frequency-domain analysis, issued from the fast fourier transform decomposition, including the calculation of root mean square (RMS in mm), mean and median frequencies (Mean F and Median F in Hz, respectively) parameters aimed at characterising the mean spectral decompositions of the sway motions on specific bandwidths (0–0.5 Hz for CoGv and 0–3 Hz for
195
ML
AP
5 mm
AP 5 mm
FFT
0.25
0.00 3 Amplitude (mm)
CoP-CoGv spectrum
0
3
CoP spectrum
ML
3
CoGv spectrum
1
CoG/CoP filter 0 0
0
0
Frequency (Hz)
3
0
3
0
0.5
iFFT
CoGv motion
CoP-CoGv motion
Fig. 1 Method used to calculate CoP–CoGv and CoGv motions from the CoP trajectories. The CoP trajectories can be depicted as a function of time along each ML and AP axis (upper part, left side) or combined to be displayed over the plane of support (upper part, right side). In order to obtain, for instance along the AP axis, CoGv motions, and consequently the difference CoP–CoGv, a mathematical low-pass filter expressing an amplitude ratio between CoGv and CoP as a function of the movement frequency, is used (middle part). To this aim, the CoP displacements are processed through a fast fourier transform (FFT) in order to obtain the amplitude
CoP–CoGv) (e.g., Rougier, 2003; Rougier et al. 2001, 2004; Rougier & Farenc, 2000).
distribution as a function of the frequency (left side, from upper to lower part). Once this CoP spectrum is obtained, a multiplication with the aforementioned filter will give the CoGv spectrum and a subtraction will give the CoP–CoGv spectrum (middle part, from left to right side). At this stage, through an inverse FFT (iFFT), it is possible to return to the temporal domain and obtain CoGv (right side, from upper to lower part) and consequently CoP–CoGv motions (lower part, from right to left side). (Adapted from Rougier, Burdet, Farenc, & Berger, 2001)
Results Control experiment
Statistical analysis One-way analyses of variance (ANOVAs) [two conditions (Control vs Attention)] were applied to the data. Level of significance was set at 0.05.
Slower probe reaction times were observed in the Attention than in the Control condition (332±11 vs 308±10 ms; F(1,7)=16.22, P0.05). Postural data from the two Order conditions were then pooled and analysed using one-way ANOVAs (two Conditions (Control vs Attention). Figure 2 illustrates representative frequential decomposition spectra for both CoGv (Fig. 2a, b) and CoP–CoGv (Fig. 2c, d) motions obtained in the two Control and Attention conditions from a typical participant. Analysis of the CoGv motions showed, along both the ML and AP axes, non significant changes for the RMS (F(1,15)=0.01, P>0.05; Fig. 3a and F(1,15)=0.30, P>0.05; Fig. 3b, respectively), Mean F (F(1,15)=0.91, P>0.05; Fig. 3c and F(1,15)=0.16, P>0.05; Fig. 3d, respectively) and Median F (F(1,15)=0.51, P>0.05; Fig. 3e and F(1,15)=0.16, P>0.05; Fig. 3f, respectively) between the two Control and Attention conditions. Analysis of the CoP–CoGv motions showed, along both the ML and AP axes, increased RMS (F(1,15)=10.75, P
