Neuroscience Letters 383 (2005) 251255

Alteration of posture-related cortical potentials in mild traumatic brain injury

Semyon Slobounov , Wayne Sebastianelli, Rashanna Moss
The Pennsylvania State University, 19 Recreation Hall, University Park, PA 16802, USA Received 29 January 2005; received in revised form 28 March 2005; accepted 7 April 2005

Abstract This paper presents additional evidence showing the persistent functional decits in concussed athletes as revealed by altered movementrelated cortical potentials (MRCP) preceding whole body postural movements at least 30 days post-injury. Eight student-athletes participated in this study (a) prior to injury; and (b) 3, 10 and 30 days after MTBI. EEG was recorded while subjects produced static balance tasks and dynamic postural movements. All subjects were cleared for sport participation within 10 days post-injury based upon neurological and neuropsychological assessments as well as upon clinical symptoms resolution. There was a persistent reduction of MRCP amplitude prior to initiation of postural movement up to 30 days post-injury, although abnormal postural responses basically recovered within 10 days post-injury. The frontal lobe MRCP effects were larger than posterior areas. This supports the notion that behavioral symptoms resolution may not be indicative of brain injury resolution. Overall, persistent alteration of movement-related cortical potentials after MTBI may indicate residual disturbance of neuronal networks involved in preparation and execution of postural movements and a lower threshold for brain re/injury. 2005 Published by Elsevier Ireland Ltd.
Keywords: Movement-related cortical potentials (MRCP); Mild traumatic brain injury (MTBI); Posture; Balance

There is a growing body of knowledge indicating longlasting residual abnormalities [2,21] including the brain dysfunctions that may persist up to 10 years after mild traumatic brain injury [24]. Athletes with a mild traumatic brain injury (MTBI) often complain of symptoms long after injury, even though their acute clinical abnormalities and measurable cognitive and behavioral decits are cleared. Incomplete recovery after a mild traumatic brain injury may also increase the risk of second impact syndrome [1], which is a signicant health problem faced by the medical community today. Therefore, the ability to identify any residual functional impairment after MTBI is critical to prevention of brain re-injury. In our previous work, we examined the residual effect of MTBI on movement-related cortical potentials (MRCP) preceding and accompanying isometric force production tasks [21]. Signicant behavioral and electro-cortical changes between MTBI subjects and controls were observed especially

Corresponding author. E-mail address: sms18@psu.edu (S. Slobounov).

when the complexity of force production task was increased. Specically, there was a reduction of amplitude of MRCP prior to initiation of a higher degree of complexity force production task. This nding has been conrmed in a more recent line of research demonstrating the temporal course of MRCP in patients with contusions to the prefrontal cortex up-to 52 weeks post-injury [26,27]. Overall, these ndings indicate prolonged malfunctioning of neuronal networks due to the incomplete recovery of the prefrontal cortex after MTBI. Recently, the prominent MRCP preceding and accompanying whole body postural movements have been documented [18]. The observed MRCP waveforms, amplitude and spatial distribution were similar to those reported for voluntary nger [16,11], wrist [22], elbow [5], and/or leg [15] movement. The largest amplitude of MRCP was found at fronto-central electrode sites with maximum at Cz, suggesting the possible source in supplementary motor area (SMA) and the foot area of the sensori-motor cortex responsible for initiation of postural movements. The results from this study conrm the important role of higher cortical structures in regulation of postural equilibrium.

0304-3940/$ see front matter 2005 Published by Elsevier Ireland Ltd. doi:10.1016/j.neulet.2005.04.039

252

S. Slobounov et al. / Neuroscience Letters 383 (2005) 251255

Persistent balance problems in acute traumatic brain injury have been reported in numerous studies [4,6]. Postural instability in MTBI subjects has been attributed to disruption of brain functions within the brainstem, cerebellum and frontal lobe [6,7], but there have been no direct examinations of this hypothesis. We hypothesize that observed postural instability in MTBI subjects may be related to impaired cortical functions for balance. As stated by Jahanshahi and Hallett [8], the movement-related cortical potentials are not static, but may be altered as a function of recovery from brain injury. Accordingly, the primary aim of this study was (a) to examine MRCP associated with initiation of postural movement prior to mild traumatic brain injury; and (b) to identify residual alterations of MRCP associated with production of postural movement in individuals suffering from MTBI. To our knowledge this is the rst study examining MRCP in the temporal course prior to and right after mild traumatic brain injury. A total of 48 subjects were initially recruited for this experiment. All subjects were Pennsylvania State University athletes at risk for traumatic brain injury (collegiate football, ice hockey and rugby players), male, aged between 18 and 25 years (mean = 20.95 years). None of these subjects had a concussion history at the time of baseline testing. Eight of these subjects suffered a grade 12 MTBI within six months after baseline testing, as assessed by a team physician. These subjects were tested again on day 3, day 10 and day 30 postinjury. All subjects were asymptomatic at day 10 of testing and were cleared for sport participation based upon neurological and neuropsychological assessments as well as based upon clinical symptoms resolution. A detailed description of inclusion criteria for MTBI can be found [21]. Consent forms approved by the Institutional Review Board of the Pennsylvania State University were obtained from each subject prior to each testing. EEG recordings were taken under three experimental conditions: static postural taskseyes open (EO) standing, eyes closed (EC) standing; and dynamic tasksthe whole body anteriorposterior (AP) postural movements. All standing trials were performed using a bipedal stance on an AMTI force plate. For the static standing trials, subjects were instructed to remain as still as possible for 30 s (EEG data during static tasks will be fully discussed in our future reports). For the dynamic AP task, subjects were requested to produce self-initiated discrete whole body postural movement in the forward direction. Subjects were instructed to sway forward as far as they could to the limits of their respective stability boundary at comfortable speed without moving their feet. Subjects were instructed to produce postural sways with opened eyes at a self-paced rate of approximately once every 10 s. Subjects performed 60 postural sways in each session. There were two sessions for this task condition. For the whole body AP postural movement, subjects were instructed to lean forward and backward with maximal range of motion predominantly at the ankle joints. A detailed description of the experimental procedure and setup can be found [21,22]. The continuous EEG activity from the scalp was recorded

at 19 sites: FP1, FP2, FZ, F3, F4, F7, F8, CZ, C3, C4, T3, T4, T7, T8, PZ, P3, P4, O1, O2 according to the International 1020 system [9a]. The ground electrode was located 10% anterior to FZ, linked earlobes served as references and electrode impedances were below 10 k . The Ag/AgCl electrodes mounted in a Quickcap Electrode Helmet (NeuroScan, Inc., El Paso, TX, USA) were used in this study. EEG signals were recorded using a programmable DC coupled broadband SynAmps amplier (NeuroScan, Inc.). The EEG signals were amplied (gain 1000, range set for 55 mV) and bandpass ltered in the dc to 70 Hz frequency range. The EEG data were sampled at 250 Hz, using a separate 16-bit analog-todigital converter for each channel. Data were collected using NeuroScans Scan 4.2 software package and written to and stored on a Dell Precision 530 computer running an Intel XEON processor. A detailed description of EEG setup can be found [21,25]. Postural data was collected using an AMTI (Advanced Mechanical Technologies, Inc., Watertown, MA, OR6-7-1000) force plate. Data were collected at a 90 Hz sampling rate, sent to an AMTI Model ADI-32 interface box, processed through the ANTI MiniAmp MSA-6 amplier, and synchronized with EEG records. The EEG signals were rst corrected for eye movements (ocular artifact reduction option of NeuroScans Scan 4.1 software). The Fy signal from the force plate indicating the initiation of forward sway was used as the trigger, and epochs were established 2500 ms before and 5000 ms after its onset. The baseline was derived from the average of the segment from 1500 to 1200 ms before the trigger point for each channel. Each epoch was visually inspected and those containing artifacts were removed. At least 50 trials were averaged for each condition. The amplitude of MRCP was measured as: (i) the mean negativity measured between 600 and 500 ms prior to force onset referred to as the Bereitschaftspotential (BP600 to 500 ) reecting the cortical activation associated with the early stages for postural movement preparation; (ii) the mean negativity measured between 100 ms prior to force onset and force onset referred to as the motor potential (MP100 to 0 ) reecting the cortical activation associated with later stages for postural movement preparation [10,11], (iii) the mean negativity measured from force-onset to 500 ms of movement production referred to as the movement monitoring potential (MMP) [5]. The MRCPs were measured at all of the above mentioned electrode sites representing the frontal, central, and parietal cortical areas. The analysis for the force plate data was performed using AMTIs Biodaq analysis program, version 1.0 software package. Center of pressure (COP) measures were assessed in both eyes open and eyes closed conditions in both static and dynamic postural tasks. COP was measured as 95% ellipse area in inches. A repeated-measures ANOVA was conducted to test for the effects of testing date (injury condition) and vision conditions on COP measures. The changes of the MRCP subcomponents amplitude in the temporal course prior to and after brain injury (main effect for the factor

S. Slobounov et al. / Neuroscience Letters 383 (2005) 251255

253

testing day) at certain brain areas (main effect for the factor electrode site) within subjects were subjected to repeatedmeasures ANOVA. The variable groupings were also used in a sectioned analysis that involved grouping the electrode sites based on anatomical location to detect differences between general functional areas within the brain. The anterior areas of the brain, but not the posterior areas of the brain, are highly vulnerable to damage after MTBI, resulting in long-term cognitive and behavioral impairment [23]. Therefore, the electrode frontback groupings are used to nd possible differences between the anterior (F3, Fz, F4), central (C3, Cz, C4), and posterior (P3, Pz, P4) sites. The condence interval for all ANOVAs that were conducted was set at 95%. When using repeated-measures ANOVA, all F-ratios were assessed using degrees of freedom corrected with the GreenhouseGeisser procedures [9b] for controlling Type1 error. For the static balance tasks, the main effect of testing date was signicant (p < 0.05). There was also an interaction of vision and testing day (p < 0.05), with signicantly larger area of COP on day 3 post-injury (see Fig. 1A). Independent-samples t-tests revealed signicant differences between baseline testing and day 3 post-injury for both vision conditions (p < 0.05). Also, signicant differences were observed between baseline testing and day 10 post-injury only for closed eyes conditions (p < 0.05). No differences were observed between baseline testing and day 30 post-injury

regardless of vision conditions (p > 0.05). Overall (a) the effect of injury on balance during static tasks was most obvious on day 3 post-injury; and (b) vision availability was a signicant factor inuencing postural stability at least within 10 days post-injury; and (c) balance problems during static tasks are fully cleared within 30 days post-injury. For the dynamic postural task, the main effect of testing day was signicant (p < 0.05), indicating the subjects reduced mobility at least within 10 days post-injury (see Fig. 1B). Independent-sample t-tests revealed signicant differences between baseline testing and day 3 and day 10 post-injury (p < 0.05). No differences were observed between baseline testing and day 30 post-injury (p > 0.05). Prior to brain injury, a slowly rising dc negativity (BP600 to 500 ) was observed in all subjects under study, starting 1500 ms prior to initiation of postural sway predominantly at anterior and central electrode grouping. From about 200 ms prior to the onset of postural sway, the amplitude (MP100 to 0 ) increased more rapidly, and was maintained throughout the duration of postural sway (MMP). The maximal negativities of MP100 to 0 were observed at Cz electrode site (see Fig. 2). Three days after injury, the BP600 to 500 component was absent and negativity departed from baseline 250 ms prior to initiation of postural sway, although the MP100 to 0 component was detected. MMP component did not show as a pronounced negativity. On day 10 and day 30 post-injury, the BP600 to 500 component still was not pronounced and the amplitude of MP 100 to 0 gradually increased, but not reach baseline. MMP component was pronounced, but also did not return to baseline. For all three MRCP components under study (BP600 to 500 ; MP100 to 0 ; MMP), repeated-measures ANOVA revealed that the main effect for the factor of testing day was signicant (p < 0.01) and none of these components reached baseline level within 30 days post-injury. Also, the ANOVA revealed a signicant main effect for the factor electrode site (grouping, p < 0.01), suggesting the

Fig. 1. (A) Temporal course of the area of center of pressure (CP) during static postural task performance with eyes open and eyes closed conditions. X-axis: 1, baseline testing; 2, day 3 post-injury; 3, day 10 post-injury; 4, day 30 post-injury. (B) Areas of the center of pressure (CP) during dynamic AP postural task performance. X-axis similar to (A).

Fig. 2. Grand averaged waveforms preceding and accompanying anteriorposterior (AP) postural movement at Cz electrode site. Green line: baseline testing; blue line: 3 days post-injury; red line: 10 days post-injury; brown line: 30 days post-injury, vertical line: postural sway onset. (For interpretation of the references to colour in this gure legend, the reader is referred to the web version of the article.)

254

S. Slobounov et al. / Neuroscience Letters 383 (2005) 251255

alteration of MRCP components predominantly at the anterior and central areas. For all MRCP components under study, repeated-measures ANOVA revealed a signicant alteration of MRCP at anterior (F3, Fz, F4) and central (C3, Cz, C4) electrode sites, p < 0.01, but not at posterior sites (P3, Pz, P4). We designed this study to examine the temporal course of postural decit and underlying functional alterations in the brain as a result of mild traumatic brain injury. It should be noted that no concussion symptoms were reported during the pre-testing interview, baseline testing, or 10 days post-injury, when the athletes were cleared for sport participation determined by standard clinical symptom resolution. Three main ndings of interest can be summarized: (1) Indices of postural instability during static tasks were most pronounced at day 3 post-injury and basically resolved within 10 days postinjury, although, vision availability may inuence postural stability in concussed subjects beyond 10 days post-injury, (2) symptoms of reduced dynamic stability may be observed at least within10 days post-injury if more challenging postural tasks are introduced, and (3) insufcient functioning of neural mechanisms involved in postural control, as evidenced by alteration of movement-related cortical potentials, is present 30 days post-injury. Systematic increase of the COP areas during static tasks, as an indication of postural instability [19], was observed at 10 days post-injury, especially when vision was not available. Moreover, abnormal postural symptoms may not be resolved within 37 days post-injury, as previously suggested [6], if more challenging dynamic postural tasks are introduced. These ndings are consistent with our previous study demonstrating the presence of various motor deciencies in MTBI subjects up to 9 months post-injury when the complexity of motor task was increased [21,25]. These new ndings are also in agreement with previous neuropsychological research suggesting that the differences in error and reaction time (RT) tests between concussed and control subjects became more obvious when behavioral task demands increased [7]. Observed symptoms of postural instability during static tasks under no vision condition may occur because of excessive reliance of concussed individuals on visual input [13] while performing postural tasks. Overall, a negative effect of MTBI on postural stability (which attributed to a sensory interaction problem using visual, vestibular and somatosensory systems [6]), may be present far beyond 7 days post-injury if properly assessed. There are several specic ndings of interest regarding the temporal course of MRCP alteration as a result of MTBI. First, BP600 to 500 component, traditionally reported as an index of preparation for self-initiated movement [8,10], was absent on day 3 post-injury and did not return to baseline level within 30 days post-injury. This may indicate insufcient brain resource allocation [3] and/or resource mobilization [12] to initiate whole body postural sway within stability region. Posture recovery, an essential skill in any sport, has been shown to be a demanding task requiring cognitive work [6]. In balance symptomatic individuals, the inability to focus

attention on the task of postural recovery and efciently recruit the cognitive resources needed for this task may be reected in the reduced amplitude of BP600 to 500, especially in the acute stage of MTBI. A steady increase of the BP component after one-year post-TBI was reported by Wiese et al. [27], suggesting the enhanced cognitive resources during the preparation of self-initiated nger movements, partly due to recovery of frontocortical systems. The long-term temporal course of whole body posture-related cortical potentials and underlying behavioral symptoms in subjects suffering from MTBI is awaiting future experimentations. Second, numerous studies support the hypothesis that the early BP600 to 500 reects general aspects of voluntary movement preparation and is less sensitive to specic movement parameters [10,11]. In contrast, motor-task specic late components of MRCP reect the central neural drive or scaling of control signal-excitation pulse from the cortex to the involved muscle groups [11,20]. The late, MP100 to 0 component of MRCP, although pronounced, was signicantly reduced on day 3 post-injury and never returned to baseline level. Collectively the data from this study suggest that the residual abnormality of cortical control of postural movements in MTBI subjects is preserved up to at least 30 days after injury. Third, previous studies in humans have reported that the pre-movement cortical negativity has a tendency to remain or even increase throughout the movement duration [5,14]. It was also reported that the longer the movement trial, the greater the MRCP duration [14]. Accordingly, this later component of MRCP was called MMP. Thus, the magnitude and duration of MMP may directly reect the efcacy of regulatory mechanisms involved in motor task production [17]. The MMP amplitude signicantly dropped on day 3 postinjury and remained lower than the baseline level during the entire testing period. This may indicate residual functional decits of higher cortical structures regulating posture and equilibrium. Finally, the alteration of MRCP was predominantly observed at the frontal and central regions of the brain with a maximum at Cz electrode site. It is well documented that frontal areas of the brain are highly vulnerable to damage after mild traumatic brain injury, resulting in long-term behavioral impairment [23]. Reduced and delayed onset of MRCP in concussed subjects may result from deranged neuronal input from the prefrontal cortex into SMA. On the functional level, this may correspond to a loss of motor preplanning caused by frontal lobe damage [26]. In conclusion, the fact that no behavioral signs of postural abnormality were observed on day 30 post-injury, while underlying cerebral alteration of postural control were still present may be explained by the enormous plasticity at different levels of the CNS allowing compensation for decient motor functions. Specic mechanisms responsible for compensatory postural responses are awaiting future examinations. The results from this report support the notion that behavioral symptoms resolution may not be indicative

S. Slobounov et al. / Neuroscience Letters 383 (2005) 251255

255

of brain injury resolution. As a result, the athletes who prematurely return to play based solely on symptom resolution criteria are highly susceptible to future and possibly more severe brain injuries. There is no universal agreement on concussion grading and return-to-play criteria after concussion. However, recent evidence in clinical practice indicates underestimation of the amount of time it takes to recover brain functions from concussion. Accordingly, the alteration of MRCPs associated with postural movement clearly observed within 30 days post-injury could potentially be considered within the scope of existing grading scales and return-to-play criteria.

[12]

[13]

[14]

[15]

References
[1] J.T. Barth, J.R. Freeman, D.K. Boshek, R.N. Varney, Accelerationdeceleration sport-related concussion: the gravity of it all, J. Athl. Train. 36 (3) (2001) 253256. [2] K. Chen, K.M. Johnston, S. Frey, M. Petrides, K. Worsley, A. Ptito, Functional abnormalities in symptomatic concussed athletes: an fMRI study, Neuroimage 22 (2004) 6882. [3] A.S. Gevins, G.M. Zeitlin, J.C. Doyle, C.D. Yingling, R.E. Schaffer, E. Callway, C.L. Yeager, EEG correlates of higher cognitive functions, Science 203 (1979) 665668. [4] A. Geurts, J. Knoop, J. van Limbeek, Is postural control associated with mental functioning in the persistent postconcussion syndrome? Arch. Phys. Rehabil. 80 (1999) 144149. [5] G. Gr unewald, E. Gr unewald-Zuberbier, Cerebral potentials during skilled slow positioning movements, Biol. Psychiatry 31 (1983) 7178. [6] K.M. Guskiewicz, Assessment of postural stability following sportrelated concussion, Curr. Sport. Med. Rep. 2 (1) (2003) 2430. [7] H. Hugenholtz, D.T. Stuss, L.L. Stethem, M.T. Richard, How long does it take to recover from a mild concussion? Neurosurgery 22 (5) (1988) 853858. [8] M. Jahanshahi, M. Hallett (Eds.), The Bereitschaftpotential: Movement-related Cortical Potentials, Kruger Academic/Plenum Press, New York, 2003, pp. 113. [9a] H.H. Jasper, The 10-20 electrode system of the International Federation, Electroencephalogr. Clin. Neurophysiol. 10 (1958) 370375. [9b] J.R. Jennings, M.J. Cohen, D.S. Ruchkin, A.J. Fridlund, Editorial policy on analysis of variance with repeated measures, Psychophysics 24 (1987) 478487. [10] H.H. Kornhuber, L. Deecke, Hirnpotentialanderungen bei Willkurbewegungen und passiven Bewegungen des Menschen, Bereitschaftspotential und reafferente Potential. P ugers Archiv f ur die Gesamte Physiologie des Menschen und der Tiere 284 (1965) 117. [11] R. Kristeva, D. Cheyne, W. Lang, G. Lindengen, L. Deecke, Movement-related potentials accompanying unilateral and bilateral

[16]

[17]

[18]

[19] [20]

[21]

[22]

[23] [24]

[25]

[26]

[27]

nger movements with different inertial loads, Electoencephalogr. Clin. Neurophysiol. 75 (1990) 410418. W.C. McCallum, Human slow potential research: a review, in: W.C. McCallum, S.H. curry (Eds.), Slow Potential Changes in the Human Brain, Plenum Press, NY, 1993, pp. 112. P. Madeleine, H. Prietzel, H. Svarrer, L. Arendt-Nielsen, Quantitative posturagraphy in altered sensory conditions: a way to assess balance instability in patients with chronic whiplash injury, Arch. Phys. Med. Rehab. 85 (3) (2004) 432438. J. Niemann, T. Winker, A. Hufschgmidt, C.H. Lucking, The inuence of hand movement on cortical negative DC potentials, in: H.J. Heinze, T.F. Munte, G.R. Mangunm (Eds.), Cognitive Psychophysiology, Birkhauser, Boston, 1994, pp. 265287. K. Saitou, Y. Washimi, Y. Koike, A. Takahashi, Y. Kaneoke, Slow negative cortical potential preceding the onset of postural adjustment, Electoencephalogr. Clin. Neurophysiol. 98 (1996) 449455. H. Shibasaki, G. Barrett, E. Halliday, A.M. Halliday, Components of the movement-related cortical potential and their scalp topography, Electroencephalogr. Clin. Neurophysiol. 49 (1980) 213226. S. Slobounov, M. Hallett, K. Newell, Perceived effort in force production as reected in motor-related cortical potentials, Clin. Neurophysiol. 155 (2004) 23912402. S. Slobounov, M. Hallett, S. Stanhope, H. Shibasaki, Role of cerebral cortex in human postural control: an EEG study, Clin. Neurophysiol. 116 (2) (2005) 215223. S. Slobounov, S. Moose, E. Slobounova, K. Newell, Aging and time to instability in posture, J. Gerontol. Biol. Sci 53A1 (1998) B71B78. S. Slobounov, W. Ray, Movement-related potentials with reference to isometric force output in discrete and repetitive tasks, Exp. Brain. Res. 123 (4) (1998) 461473. S. Slobounov, W. Sebastianelli, R. Simon, Neurophysiological and behavioral concomitants of mild brain injury in college athletes, Clin. Neurophysiol. 113 (2002) 185193. S. Slobounov, R. Simon, R. Tutwiler, EEG correlates of wrist kinematics as revealed by averaging techniques and Morlet wavelet transforms, Mot. Control 45 (2000) 247272. D. Stuss, R. Knight, Principles of Frontal Lobe Function, University Press, Oxford, 2002. R.W. Thatcher, C. Biver, J.F. Gomez, D. North, R. Curtin, R.A. Walker, A. Salazar, Estimation of the EEG power spectrum using MRI T2 relaxation time in traumatic brain injury, Electroencephalogr. Clin. Neurophysiol. 112 (2001) 17291745. J. Thompson, W. Sebastianelli, S. Slobounov, EEG and postural correlates of mild traumatic brain injury in athletes, Neurosci. Lett. 337 (2005) 158163. H. Wiese, P. Stude, K. Nebel, D. Osenberg, W. Ischwbeck, D. Stolke, H.C. Diener, M. Keidel, Impaired movement-related potentials in acute frontal traumatic brain injury, Clin. Neurophysiol. 115 (2004) 289298. H. Wiese, P. Stude, K. Nebel, D. Osenberg, W. Ischwbeck, D. Stolke, H.C. Diener, M. Keidel, Recover of movement-related potentials in the temporal course after prefrontal traumatic brain injury: a followup study, Clin. Neurophysiol. 115 (2004) 26772692.