Right cerebral hemisphere specialization for quiet and perturbed body balance control: Evidence from unilateral stroke

Human Movement Science xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Human Movement Science journal homepage: www.elsevier.com/locate/humov

Full Length Article

Right cerebral hemisphere specialization for quiet and perturbed body balance control: Evidence from unilateral stroke ⁎

Corina Aparecida Fernandes, Daniel Boari Coelho , Alessandra Rezende Martinelli, Luis Augusto Teixeira Human Motor Systems Laboratory, School of Physical Education and Sport, University of São Paulo, Brazil

AR TI CLE I NF O

AB S T R A CT

Keywords: Inter-hemispheric asymmetry Cerebral hemisphere damage Sensory manipulation Upright balance Perturbed balance

Our aim in this investigation was to assess the relative importance of each cerebral hemisphere in quiet and perturbed balance, based on uni-hemispheric lesions by stroke. We tested the hypothesis of right cerebral hemisphere specialization for balance control. Groups of damage either to the right (RHD, n = 9) or the left (LHD, n = 7) cerebral hemisphere were compared across tasks requiring quiet balance or body balance recovery following a mechanical perturbation, comparing them to age-matched nondisabled individuals (controls, n = 24). They were evaluated in conditions of full and occluded vision. In Experiment 1, the groups were compared in the task of quiet standing on (A) rigid and (B) malleable surfaces, having as outcome measures center of pressure (CoP) amplitude and velocity sway. In Experiment 2, we evaluated the recovery of body balance following a perturbation inducing forward body oscillation, having as outcome measures CoP displacement, peak hip and ankle rotations and muscular activation of both legs. Results from Experiment 1 showed higher values of CoP sway velocity for RHD in comparison to LHD and controls in the anteroposterior (rigid surface) and mediolateral (malleable surface) directions, while LHD had lower balance stability than the controls only in the mediolateral direction when supported on the rigid surface. In Experiment 2 results showed that RHD led to increased values in comparison to LHD and controls for anteroposterior CoP displacement and velocity, time to CoP direction reversion, hip rotation, and magnitude of muscular activation in the paretic leg, while LHD was found to differ in comparison to controls in magnitude of muscular activation of the paretic leg and amplitude of mediolateral sway only. These results suggest that damage to the right as compared to the left cerebral hemisphere by stroke leads to poorer postural responses both in quiet and perturbed balance. That effect was not altered by manipulation of sensory information. Our findings suggest that the right cerebral hemisphere plays a more prominent role in efferent processes responsible for balance control.

1. Introduction The conceptualization that each cerebral hemisphere is specialized for particular functions of movement control is consensual in the literature. Sainburg (2014) has proposed that in right handers while the left cerebral hemisphere is specialized for predicting the effects of the relationship between muscular and external forces (body-environment dynamics), the right hemisphere is specialized at mechanisms associated with impedance control. Impedance control has been conceptualized as neuromotor mechanisms leading to maintenance of a stable posture and to appropriate movement corrections when the body suffers unanticipated perturbations. Recent ⁎

Corresponding author at: Av. Prof. Mello Moraes, 65, São Paulo, SP 05508-030, Brazil. E-mail address: [email protected] (D.B. Coelho).

http://dx.doi.org/10.1016/j.humov.2017.09.015 Received 5 April 2017; Received in revised form 27 September 2017; Accepted 29 September 2017 0167-9457/ © 2017 Elsevier B.V. All rights reserved.

Please cite this article as: Fernandes, C.A., Human Movement Science (2017), http://dx.doi.org/10.1016/j.humov.2017.09.015

Human Movement Science xxx (xxxx) xxx–xxx

C.A. Fernandes et al.

advancement of Sainburg’s theoretical model has led to the conceptualization of a hybrid movement control, with both cerebral hemispheres participating in the regulation of each limb. From that perspective, hemispheric specialization should be more evident in the contralateral limbs but affecting also (to a lesser extent) the control of ipsilateral body segments as well (Yadav & Sainburg, 2011 2014). Evidence consistent with the notion of specialization of the right cerebral hemisphere for impedance control has been accumulated from investigation in nondisabled individuals by showing that the control of nondominant (left) arm movements is based predominantly on impedance mechanisms to compensate for self- or externally-induced errors in reaching (Schabowsky, Hidler, & Lum, 2007; Yadav & Sainburg, 2011, 2014). This conclusion has been corroborated in similar evaluation of individuals with unilateral cerebral damage (Schaefer, Haaland, & Sainburg, 2007, 2009; Schaefer, Mutha, Haaland, & Sainburg, 2012). Information on hemisphere specialization for balance control has been provided by studies in individuals who suffered unilateral cerebral damage by stroke. Investigations comparing balance control following damage to the right versus left cerebral hemisphere have indicated a functional specialization of the right hemisphere for body balance control. Right hemisphere specialization for balance control has been inferred from assessment of quiet body balance, showing that damage to the right in comparison with the left cerebral hemisphere leads to reduced capacity to shift body weight between the legs (Ishii et al., 2010), poorer body vertical orientation (Perennou et al., 2008), lower scores in qualitative clinical evaluation (Perennou et al., 1999), delayed recovery of independent stance (Bohannon, Smith, & Larkin, 1986; Laufer, Sivan, Schwarzmann, & Sprecher, 2003), increased body sway (Peurala, Kononen, Pitkanen, Sivenius, & Tarkka, 2007; Rode, Tiliket, & Boisson, 1997), and poorer voluntary center of pressure displacement (Ioffe, Chernikova, Umarova, Katsuba, & Kulikov, 2010; Ustinova, Chernikova, Ioffe, & Sliva, 2001). Poor balance control resulting from right hemisphere damage is consistent with the conjecture that right cerebral hemisphere specialization for impedance control favors postural stabilization. Even though previous studies in individuals with unilateral stroke have suggested right cerebral hemisphere specialization for balance control in quiet stance, distinct critical issues have been left untouched thus far. First, a prevalent explanation for decreased balance control from right cerebral hemisphere damage is based on the notion that the right hemisphere is specialized for intersensory integration, leading to poor sensory reweighting when one or more sensory sources are disrupted. Support for this perspective has been provided by results showing that damage of areas in the right but not in the left cerebral hemisphere induced greater sway under visual occlusion in comparison with behavior of undamaged individuals, whereas under full vision balance was not significantly impaired by stroke (Manor et al., 2010). These findings suggest that lateral asymmetries in body balance control are detectable in conditions of vision deprivation, requiring increased weighting of other sensory sources signaling postural sway, while balance performance is similar to that seen in nondisabled individuals under full sensory information (see also Bensoussan et al., 2007; Bonan et al., 2004; Marigold & Eng, 2006b). From this perspective, hemisphere specialization might be associated with sensory processing, with deficit of intersensory reweighting in conditions of sensory deprivation. Then, a comprehensive analysis of right hemisphere specialization for balance control should include evaluation of the role played by sensory processing. Second, the most challenging situation for balance control is represented by unexpected perturbations to stance. This challenge is particularly evident in post-stroke patients (Di Fabio, 1987; Di Fabio, Badke, & Duncan, 1986; Marigold & Eng, 2006a; Pollock, Ivanova, Hunt, & Garland, 2015), with unexpected perturbations being possibly associated with several cases of falls in that population (Mansfield, Inness, Wong, Fraser, & McIlroy, 2013; Salot, Patel, & Bhatt, 2016). While balance control in quiet stance requires continuous low-magnitude postural adjustments, perturbed balance requires fast and strong muscular responses proportional to the magnitude of perturbation to recover balance stability (Azzi, Coelho, & Teixeira, 2017). From this observation, it might be conceived that specific neural mechanisms underlie quiet and perturbed body balance control, with the first relying on continuous feedback processing while the latter being implemented by a burst of activation of postural muscles triggered by different sources of sensory information signaling direction and magnitude of balance loss. In the present study, the issue of cerebral hemisphere specialization for balance control in quiet and perturbed balance was assessed in two experiments, comparing performance between individuals who suffered a unilateral stroke either to the right or to the left cerebral hemisphere. Age-matched neurologically nondisabled individuals served as reference for comparison. In Experiment 1A, we evaluated balance control in quiet stance on a rigid surface, while in Experiment 1B balance was evaluated with participants standing on a malleable surface leading to distortion of tactile afference from the feet soles. In Experiment 2, we evaluated for the first time cerebral hemispheres specialization in reactive postural responses to an unanticipated mechanical perturbation. In the two experiments, the relevance of visual information was assessed by contrasting the conditions of full vision and visual occlusion. Thus, we originally provide here results allowing for evaluation of adequacy of the proposition of right cerebral hemisphere specialization for impedance control, in contrast to specialization of sensory processing, for balance regulation. The proposition of right cerebral hemisphere specialization for impedance control leads to the prediction of poorer performance both in quiet and perturbed balance in individuals with right cerebral hemisphere damage. Alternatively, if hemisphere specialization is associated with deficits in sensory processing, one would expect increased impairment of balance control from right hemisphere damage only in conditions of manipulation of sensory information.

2. Experiment 1: Quiet standing In this experiment, we evaluated balance control in quiet standing while supported on a rigid (Experiment 1A) and on a malleable (Experiment 1B) surface. Results are presented separately due to the different periods of time participants were able to stand on each surface.

2

Human Movement Science xxx (xxxx) xxx–xxx

C.A. Fernandes et al.

2.1. Material and methods 2.1.1. Participants Participated of this study individuals with single unilateral stroke to the right (n = 9) or left (n = 8) cerebral hemisphere, and age-matched neurologically nondisabled individuals (n = 25). Selection of both post-stroke and control participants was made from the same hospital database, with all of them living in the same region. Post-stroke participants were selected following clinical evaluation and cerebral image analysis by neurologists. Inclusion criteria for hemisphere damage participants were the following: time range following stroke from 6 months to 3 years, mild to moderate functional disability based on the modified Rankin scale (score ≤3) (Wilson et al., 2002), behavioral manifestation of hemiparesis, absence of neurologic or orthopedic diseases other than those induced exclusively by the stroke, absence of damage to the cerebellum or brain stem, and scores 1–2 in the modified Ashworth’s spasticity scale (Ansari, Naghdi, Arab, & Jalaie, 2008). One male participant (left hemisphere stroke) was excluded due to incapacity to perform the experimental tasks under visual occlusion. For the nondisabled participants, there was one exclusion because of detection of stroke symptoms based on questionnaire (cf. Abe, Goulart, Santos Junior, Lotufo, & Bensenor, 2010). We evaluated in this experiment 9 (4 women) individuals who suffered stroke to the right and 7 (3 women) who suffered stroke to the left cerebral hemisphere, and 24 (15 women) nondisabled participants. Their average ages were the following: right hemisphere stroke, 63.77 years (age range 55–83 years, SD = 9.52); left hemisphere stroke, 62.63 years (age range 54–83 years, SD = 9.67), and controls, 58.24 years (age range 46–76 years, SD = 7.15)1. All participants scored over 13 points for the uneducated, 18 for the low and middle school education, and 26 for high school education in the assessment of mental function through the modified minimental state examination (Bertolucci, Brucki, Campacci, & Juliano, 1994). Results from application of the Edinburgh manual dominance inventory (Oldfield, 1971) indicated that participants, with a single exception (right hemisphere stroke), were right handed. Results from application of the Waterloo footedness inventory (Elias, Bryden, & Bulman-Fleming, 1998) indicated that most participants were right footed (n = 7 for right hemisphere lesion, n = 5 for left hemisphere lesion, and n = 20 for controls), with the others being classified as left or mixed footed. Both hand and foot dominance evaluations were for the period preceding stroke for the post-stroke groups. Additional descriptive data for the hemisphere damage participants, including stroke type, damaged areas and Fugl-Meyer scale (Fugl-Meyer, Jääskö, Leyman, Olsson, & Steglind, 1975) are presented in Table 1. The local ethics university committee approved the study protocol for this and for the ensuing experiments. Participants provided informed consent according to the Declaration of Helsinki. 2.1.2. Task and apparatus The experimental task consisted of maintaining quiet balance in upright bipedal support. The feet were oriented approximately in parallel with a distance of 5 cm between the medial malleoli. Both arms were maintained relaxed and parallel to the trunk. Quiet balance was performed under conditions of full vision and visual occlusion. For full vision, participants gazed at a black spot (1-cm diameter) at the top of a vertical shaft, positioned frontally 1.5 m away at the eyes height. For performance under visual occlusion, participants were blindfolded and oriented their head as if gazing at the visual target. The task was performed with the participants supported on the hard surface of a force platform (Model OR6-6-1000; AMTI Corp., Newton, MA) during a time interval of 30 s (Experiment 1A), or on a malleable support surface during a time interval of 15 s (Experiment 1B). The malleable surface corresponded to a 9-mm-thick viscoelastic piece of high-density foam (Tempur, Soft D3110) placed upon the platform surface. 2.1.3. Experimental design and procedures Quiet balance was compared between participants who suffered right (RHD) or left (LHD) hemisphere damage, and the nondisabled controls (CO). Participants were instructed to maintain upright stance as steady as possible during the period of data acquisition, avoiding any movements or body oscillation. They stood barefoot directly on the force platform (Experiment 1A), performing three trials under full vision and three trials under visual occlusion. Trials for each visual condition were blocked, with sequence between conditions alternated across participants within each group. Evaluation was preceded by one familiarization trial for the specific visual condition, with data from these trials discarded from analysis. Experiment 1B was performed 10 min. after the end of Experiment 1A with the same participants. During the interval between the two parts of this experiment participants rested sitting quietly. Procedures were the same between the two parts of the experiment, except for trial duration between surfaces. Sequence of visual conditions was alternated across participants within group. Intertrial intervals within blocks lasted 30 s maintaining upright stance. Intervals between blocks (surface x visual condition) were 2-min. long, during which participants rested sitting. Participants worn a safety harness, attached to the laboratory ceiling by means of cables. Sampling frequency of force platform data acquisition was set at 200 Hz. 2.1.4. Analysis Analysis of balance was based on feet center of pressure (CoP) on the ground, assessing root mean square (RMS) and mean velocity module in the anteroposterior and mediolateral directions. Analysis was made excluding the initial 5 s (Experiment 1A) or 2 s (Experiment 1B) of the acquired data in each trial. Data extraction and processing were made automatically through a MATLAB (Mathwoks®) routine following visual data inspection. Center of pressure data were low-pass filtered through a fourth-order Butterworth filter with a cutoff frequency of 10 Hz. Trials for each visual condition were individually averaged for statistical analysis. 1

Notes: One participant with right hemisphere damage dropped out for Experiment 1B and Experiment 2.

3

Human Movement Science xxx (xxxx) xxx–xxx

C.A. Fernandes et al.

Table 1 Kind of stroke, neural areas damaged and Fugl-Meyer scale score.

Time after stroke, months (SD) Stroke type (n) Ischemic Hemorrhagic Stroke site (n) Parietal Parietal and Temporal Parietal and Frontal Parietal, Temporal and Frontal Parietal, Temporal, Frontal and Basal ganglia Frontal Occipital Occipital and Thalamus Occipital, Temporal and Thalamus Basal ganglia Frontal and Temporal Frontal and Basal ganglia Fugl-Meyer (score) Sensitivity Mobility

Right

Left

17.08 (9.07)

17.44 (8.73)

7 2

7 0

1 1

1 1 1

2 1 1 1

2

1 1 1

1 10 22.5

10 25.5

Stroke sites. Cortical areas: frontal, temporal, parietal, and occipital cortices, temporo-parietal junction. Subcortical: thalamus, basal ganglia, and insula. Maximum score for the Fugl-Meyer scale (Fugl-Meyer, Jääskö, Leyman, Olsson, & Steglind, 1975) is 12 for sensitivity and 34 for mobility.

Assumptions of parametric statistics were verified through the Shapiro-Wilks test. Statistical analysis was performed separately for each surface by means of two-way 3 (group: RHD x LHD x CO) x 2 (vision: full x occlusion) ANOVAs with repeated measures on the second factor. Post hoc comparisons were performed through Newman-Keuls procedures, and effect sizes are indicated by partial eta squared (ηp2). Statistical significant effects (P values ≤ .05) are reported only. 2.2. Results In Fig.1 we compare single-trial representative signals of mediolateral CoP sway under full vision between participants of the two hemisphere damage and control groups. That figure represents the lower balance induced by right hemisphere damage in conditions of rigid (Experiment 1A) and malleable (Experiment 1B) surfaces. 2.2.1. Experiment 1A: Rigid surface 2.2.1.1. CoP sway amplitude. Analysis of RMS for anteroposterior CoP sway indicated a significant main effect of vision, F(1, 37) = 16.11, P = .0003, ηp2 = 0.30, which was due to higher values under visual occlusion (M = 4.62 mm, SD = 2.04) in comparison with full vision (M = 3.88 mm, SD = 1.81) (Fig. 2A). Analysis of RMS for mediolateral CoP sway indicated significant main effects of group, F(2, 37) = 6.32. P = .004, ηp2 = 0.25, and vision, F(1, 37) = 7.27, P = .01, ηp2 = 0.16. Post hoc comparisons for the effect of

Fig. 1. Representative individual mediolateral CoP displacement over time in the condition of full vision showing increased oscillation for a participant with right hemisphere damage (RHD) in comparison with left hemisphere damage (LHD) and nondisabled control (CO) participants. Comparisons are shown for the rigid (A, Experiment 1A) and for the malleable (B, Experiment 1B) surfaces.

4

Human Movement Science xxx (xxxx) xxx–xxx

C.A. Fernandes et al.

Fig. 2. Experiment 1A, quiet stance on a rigid surface. Average values (standard errors in vertical bars) for the groups (RHD, right hemisphere damage; LHD, left hemisphere damage; CO, control) under full and occluded vision, for the following variables: CoP root mean square (RMS) for the anteroposterior (A) and mediolateral (B) directions, and CoP mean velocity module for the anteroposterior (C) and mediolateral (D) directions. Single asterisks indicate significant differences from controls, and double asterisks significant differences from LHD and controls (P values < .05).

group showed higher values for both hemisphere damage groups, RHD (M = 3.57 mm, SD = 1.18) and LHD (M = 3.04 mm, SD = 2.96), in comparison with CO (M = 1.68 mm, SD = 0.81), while no significant difference was found between the hemisphere damage groups (Fig. 2B). The effect of vision was due to higher values under visual occlusion (M = 2.45 mm, SD = 1.81) in comparison with full vision (M = 2.24 mm, SD = 1.56). 2.2.1.2. CoP sway velocity. Analysis of anteroposterior velocity of CoP sway indicated significant main effects of group, F(2, 37) = 9.57, P = .0005, ηp2 = 0.34, and vision, F(1, 37) = 72.27, P = .0001, ηp2 = 0.66. Post hoc comparisons for the effect of group showed higher values for RHD (M = 12.91 mm/s, SD = 6.12) in comparison with LHD (M = 9.54 mm/s, SD = 4.44) and CO (M = 6.91 mm/s, SD = 2.35), while no significant difference was found between the latter groups (Fig. 2C). The effect of vision was due to higher values under visual occlusion (M = 10.08 mm/s, SD = 4.75) than under full vision (M = 7.36 mm/s, SD = 3.95). Analysis of mediolateral velocity of CoP sway indicated significant main effects of group, F(2, 37) = 7.50, P = .002, ηp2 = 0.29, and vision, F(1, 37) = 17.36, P = .0002, ηp2 = 0.32. Post hoc comparisons for the effect of group showed higher values for RHD (M = 7.53 mm/s, SD = 4.70) and LHD (M = 5.96 mm/s, SD = 4.95) in comparison with CO (M = 3.26 mm/s, SD = 0.82), while no significant difference was found between the hemisphere damage groups (Fig. 2D). The effect of vision was due to higher values under visual occlusion (M = 5.15 mm/s, SD = 3.25) than under full vision (M = 4.23 mm/s, SD = 3.25). 2.2.2. Experiment 1B: Malleable surface 2.2.2.1. CoP sway amplitude. Analysis of RMS for anteroposterior CoP sway indicated a significant main effect of vision, F(1, 35) = 66.89, P = .0001, ηp2 = 0.66, which was due to higher values under visual occlusion (M = 8.77 mm, SD = 3.16) in comparison with full vision (M = 4.76 mm, SD = 2.01) (Fig. 3A). Analysis of RMS for mediolateral CoP sway indicated significant main effects of group, F(2, 35) = 8.22, P = .001, ηp2 = 0.32, and vision, F(1, 35) = 9.67, P = .004, ηp2 = 0.22. Post hoc comparisons for the effect of group showed higher values for RHD (M = 6.34 mm, SD = 3.17) than for CO (M = 3.87 mm, SD = 1.54), while no significant differences were found in the comparisons with LHD (M = 5.15, SD = 1.78) (Fig. 3B). The effect of vision was due to higher values 5

Human Movement Science xxx (xxxx) xxx–xxx

C.A. Fernandes et al.

Fig. 3. Experiment 1B, quiet stance on a malleable surface. Average values (standard errors in vertical bars) for the groups (RHD, right hemisphere damage; LHD, left hemisphere damage; CO, control) under full and occluded vision, for the following variables: CoP root mean square (RMS) for the anteroposterior (A) and mediolateral (B) directions, and CoP mean velocity module for the anteroposterior (C) and mediolateral (D) directions. Single asterisks indicate significant differences from controls, and double asterisks significant differences from LHD and controls (P values < .05).

under visual occlusion (M = 5.25 mm, SD = 2.47) in comparison with full vision (M = 3.88 mm, SD = 1.58). 2.2.2.2. CoP sway velocity. Analysis of average velocity of anteroposterior CoP sway indicated a significant main effect of vision, F(1, 35) = 54.24, P = .0001, ηp2 = 0.61, which was due to higher values under visual occlusion (M = 24.43 mm/s, SD = 9.18) in comparison with full vision (M = 13.36 mm/s, SD = 4.90) (Fig. 3C). Analysis of average velocity of mediolateral CoP sway indicated significant main effects of group, F(2, 35) = 4.33, P = .02, ηp2 = 0.20, and vision, F(1, 35) = 22.75, P = .0001, ηp2 = 0.39. Post hoc comparisons for the effect of group showed higher values for RHD (M = 13.37 mm/s, SD = 9.06) in comparison with LHD (M = 9.46 mm/s, SD = 4.50) and CO (M = 8.06 mm/s, SD = 3.64), with no significant difference between the latter groups (Fig. 3D). The effect of vision was due to higher values under visual occlusion (M = 11.30 mm/s, SD = 6.45) in comparison with full vision (M = 7.29 mm/s, SD = 3.34). 3. Experiment 2: Perturbed balance 3.1. Material and methods 3.1.1. Tasks and apparatus The experimental task consisted of recovering stable balance following an unanticipated mechanical perturbation. Perturbation was provoked by using a custom-built load release system, as described in the following. As initial position, participants stood upright on the force platform resisting against a load pulling the participant’s body backwards. The arms were maintained relaxed in front of the body in parallel with the trunk. They wore a 20-cm wide harness positioned at the lumbar-sacral region with an electromagnetic system attached to its back side. A steel cable tied to a magnet was linked to the harness by means of an electromagnet. The cable crossed a pulley attached to a height-adjustable support, with a load hung at the end of the cable (Fig. 4A). During the period of load application, participants were asked to lean the body slightly forward to assume a stable body position, keeping legs and trunk in 6

Human Movement Science xxx (xxxx) xxx–xxx

C.A. Fernandes et al.

Fig. 4. Experiment 2, representation of the perturbed balance task, with the load pulling the participants’ trunk backward in the period preceding load release (A). Representative individual signals for right hemisphere damage (RHD) in comparison with left hemisphere damage (LHD) and nondisabled control (CO) participants, showing variation of the following variables over time in the anteroposterior direction following load release (dashed lines): center of pressure (CoP) position (B) and velocity (C), hip (D) and ankle (E) angles, and activation of the muscle gastrocnemius medialis (paretic leg for RHD and LHD).

straight line, with the feet in full contact with the force platform. A remote switch was used to release the load at a time unanticipated by the participant. Following load release, the participant’s body suffered a fast forward oscillation, requiring reactive responses mainly through activation of the posterior muscles of the legs (see Martinelli, Coelho, Magalhaes, Kohn, & Teixeira, 2015, for a detailed description). For evaluation of joint angular displacement following load release, spherical reflective markers, 15-mm in diameter, were attached to joints of interest. Those markers were tracked by means of a set of four optoelectronic cameras (Vicon ®, MX3+). Activation of the muscle gastrocnemius medialis (GM) of both legs was measured by means of wireless surface electrodes (Delsys Trigno Wireless System, Boston, MA, USA). Apparatuses were synchronized by means of the Vicon system.

3.1.2. Experimental design and procedures Participants of both hemisphere damage groups, RHD (n = 8) and LHD (n = 7), and controls (n = 24) were tested under full vision and visual occlusion, with sequence of visual conditions alternated across participants within group. For familiarization, the perturbation task was performed preliminarily with lower loads. The first trial was performed with the load of 1 kg, the second with 2 kg, and the third with 5% of the personal body mass (3–5 kg across participants). Following those familiarization trials (discarded from analysis), they were evaluated by using the 5% load, performing three trials in block under each visual condition. Reflective markers for kinematic analysis were attached over the body having the following anatomical points as reference: fifth 7

Human Movement Science xxx (xxxx) xxx–xxx

C.A. Fernandes et al.

metatarsophalangeal joint, lateral malleolus, lateral knee joint center, greater trochanter, and acromion. Markers were attached to the nonparetic side in the hemisphere damage participants, and to the right body side in the controls. Fixation of the markers was made directly on the skin, except by the marker for the greater trochanter. That marker was attached firmly on a belt worn over the clothes. Electrodes were attached bilaterally to the GM muscle, following the SENIAM recommendations for electromyography (EMG) measurements (http://www.seniam.org/). In cases that the participant moved their feet is response to balance perturbation that trial was rejected and immediately repeated. Intertrial intervals within visual conditions were 30-s long, time during which participants rested standing upright on the force platform. Between visual conditions, participants rested sitting during an interval of 2 min. Sampling frequency of data acquisition was set at 200 Hz for the force platform and cameras, and at 2000 Hz for EMG. 3.1.3. Analysis Analysis was conducted in the period immediately following load release. Dependent variables were the following: for the anteroposterior direction, amplitude of CoP displacement following load release (a); maximum CoP velocity (b); time of CoP reversal (c), measured from the time of load release until the time of CoP peak displacement; for the mediolateral direction, amplitude between right-left peaks of COP displacement in the interval of 1 s following perturbation onset (d); peak rotation amplitudes at the hip (e) and ankle (f); latency of muscular activation onset (g), having as criterion the sustained increment of values above two standard deviations in the period of 200 ms preceding perturbation; and magnitude of muscular activation (h), given by the normalized integral for the time interval of 75 ms following muscle activation onset. Electromyographic signals were band-pass filtered online between 20 and 400 Hz, rectified, and then Butterworth low-pass filtered offline at 30 Hz. Data was normalized by the respective maximum muscular activation across experimental conditions. Center of pressure and kinematic signals were digitally low-pass filtered with a fourth-order Butterworth filter adopting a cutoff frequency of 10 Hz. Statistical analysis for CoP and kinematic data was made by means of two-way 3 (group: RHD × LHD × CO) × 2 (vision: vision x occlusion) ANOVAs with repeated measures on the second factor. For EMG analysis, half of participants of the control group had their right leg labeled as paretic and the left leg as nonparetic, while for the other half the opposite labeling was used. Legs labeling across control participants was randomly assigned. Analysis of EMG-related variables was made by means of three-way 3 (group) × 2 (vision) × 2 (leg: paretic x nonparetic) ANOVAs with repeated measures on the last two factors. Post hoc comparisons were performed through Newman-Keuls procedures, and effect sizes are indicated by partial eta squared (ηp2). Statistical significant effects (P values ≤ 0.05) are reported only. 3.2. Results A few trials were repeated due to a compensatory step in response to perturbation: RHD = 2, LHD = 3 and CO = 2. Signals from individual trials are presented in Fig. 4 (panels B-F) to show the distinctive characteristics between participants of the RHD, LHD and CO groups. In that figure, we compare groups’ representative signals for reactive postural responses to load release (time of load release indicated by vertical dashed lines) for anteroposterior CoP displacement (B) and velocity (C), hip (D) and ankle (E) rotations, and activation of the GM muscle (F, paretic leg for RHD). Analyses of average values of CoP displacement in the anteroposterior and mediolateral directions and amplitude of joints rotations are presented in Table 2. 3.2.1. Anterior CoP displacement Analysis of amplitude of anterior CoP displacement indicated average values between 10.76 and 11.22 cm across groups and visual conditions, but no significant differences were found in the respective analysis. Analysis of peak CoP velocity showed a significant main effect of group, F(2, 36)=5.97, P = .006, ηp2=0.25. Post hoc comparisons indicated lower values for RHD Table 2 Experiment 2, center of pressure and joint rotation related variables. Full vision

CoPap amplitude (cm) CoPap max velocity (cm/s) CoPap reversal time (ms) CoPml amplitude (cm) Hip rotation (°) Ankle rotation (°)

Visual occlusion

RHD

LHD

CO

RHD

LHD

CO

10.82 (0.57) 62.90** (3.69) 720.16** (48.84) 0.19* (0.05) 3.63** (0.94) 3.63 (0.77)

10.91 (0.16) 84.91 (6.18) 587.89 (64.20) 0.20* (0.04) 1.90 (0.40) 2.91 (0.45)

11.11 (0.40) 83.18 (4.76) 582.60 (40.31) 0.12 (0.02) 1.68 (0.46) 2.29 (0.58)

11.22 (0.51) 70.25** (5.12) 668.35** (79.85) 0.22* (0.05) 3.71** (0.94) 4.19 (0.69)

10.76 (0.37) 87.20 (7.73) 538.07 (26.89) 0.23* (0.05) 2.58 (0.66) 3.69 (0.66)

11.00 (0.42) 84.73 (6.46) 565.13 (45.44) 0.12 (0.02) 1.80 (0.53) 2.61 (0.66)

Average values (standard errors in parenthesis) for the groups (RHD, right hemisphere damage; LHD, left hemisphere damage; CO, control), under full vision and visual occlusion. Single asterisks indicate significant differences from controls, and double asterisks significant differences from LHD and controls (P values ≤ .05).

8

Human Movement Science xxx (xxxx) xxx–xxx

C.A. Fernandes et al.

(M = 66.57 cm/s, SD = 12.77) in comparison with LHD (M = 86.05 cm/s, SD = 17.83) and CO (M = 83.95 cm/s, SD = 14.87), with no significant difference between the latter groups. Analysis of time of direction reversal of CoP displacement showed a significant main effect of group, F(2, 36) = 3.31, P = .05, ηp2 = 0.16. Post hoc comparisons revealed longer times for RHD (M = 694.26 ms, SD = 182.83) in comparison with LHD (M = 562.09 ms, SD = 127.75) and CO (M = 573.86 ms, SD = 112.78), with no significant difference between the latter groups. 3.2.2. Mediolateral CoP displacement Analysis of peak-to-peak mediolateral CoP displacement indicated a significant main effect of group, F(2, 36) = 5.63, P = .007, ηp2 = 0.24. Post hoc comparisons indicated increased values for RHD (M = 0.20 cm, SD = 0.14) and LHD (M = 0.21 cm, SD = 0.12) in comparison to CO (M = 0.12 cm, SD = 0.05), with no significant difference between the post-stroke groups. 3.2.3. Joints rotation amplitude Kinematic analysis of amplitude of hip rotation showed a significant main effect of group, F(2, 36) = 4.45, P = .02, ηp2 = 0.20. Post hoc comparisons indicated larger hip rotation amplitude for RHD (M = 3.66°, SD = 2.67) in comparison with LHD (M = 2.39°, SD = 1.43) and CO (M = 1.74°, SD = 1.31), with no significant difference between the latter groups. Analysis of amplitude of ankle rotation indicated a significant main effect of vision, F(1, 36) = 7.11, P = .01, ηp2 = 0.16. Effect of vision was due to increased ankle rotation amplitude under visual occlusion (M = 3.13°, SD = 1.98) in comparison with full vision (M = 2.68°, SD = 1.67). 3.2.4. Muscular activation EMG recordings were unreliable due to noise artifact in one of the legs for one male participant of RHD and for four participants of CO. This way, EMG analysis was performed on 34 participants (RHD, n = 7, 4 women; LHD, n = 7, 3 women; CO, n = 20, 12 women). Results for latency of muscular activation onset showed a significant main effect of leg, F(1, 31) = 24.30, P = .0001, ηp2 = 0.44, and a significant group by leg interaction, F(2, 31) = 7.80, P = .0002, ηp2=0.33. A nonsignificant trend toward a main effect of group, F(2, 31) = 2.76, P = .08, suggests overall longer latency for RHD. Decomposition of the group by leg interaction into its simple components indicated that both hemisphere damage groups had longer latencies in the paretic than in the nonparetic leg; the paretic leg of RHD, but not of LHD, presented longer latencies than the corresponding leg of controls (Fig. 5A). Analysis of magnitude of muscular activation indicated a significant group × leg interaction, F (2, 31) = 6.82, P = .004, ηp2 = 0.31. Decomposition of the interaction into its simple components showed significantly lower activation of the RHD’s paretic leg in comparison with the corresponding leg of LHD, P = .05, and CO, P = .0005, and lower activation of that leg in LHD than in CO, P = .05. Results also showed lower muscular activation in both hemisphere damage groups in the paretic than in the nonparetic leg, RHD, P = .001; LHD, P = .04; no significant differences were found in the comparisons between groups for the nonparetic leg (Fig. 5B). 3.3. Discussion In the present experiments, we compared performance in quiet and perturbed balance between individuals who suffered damage either to the right or the left cerebral hemisphere by stroke, having performance of nondisabled individuals as reference. Results from both quiet and perturbed balance evaluations converged to show that individuals who suffered right hemisphere damage had poorer balance control not only in comparison with nondisabled individuals but also in comparison with those who suffered damage to the left cerebral hemisphere. Impoverished balance stability of RHD was manifested in increased velocity of CoP sway in quiet balance, and in many parameters assessed in perturbed balance indicating a reduced capacity to respond appropriately to unanticipated body oscillations. Longer time to CoP reversal in this group is consistent with their reduced magnitude of muscular activation of the paretic leg and lower velocity of CoP displacement. Left hemisphere damage, on the other hand, led to reduced balance control deficits, with their performance being similar to controls in several variables associated with quiet and mainly perturbed balance. These results suggest that the right cerebral hemisphere is more specialized than the left for body balance control, having preeminence in both stabilization of quiet balance and in generating automatic postural responses to recover body balance following an unanticipated perturbation. Lack of interactions associated with sensory manipulation suggests that right cerebral hemisphere specialization for balance control is not due to its supposed superior capacity of somatosensory processing. These results are the first to show that right cerebral hemisphere damage by stroke leads to a more dramatic impairment of postural responses both in quiet and perturbed body balance control. 3.3.1. Is hemispheric specialization due to sensory processing? Investigation of balance control in nondisabled individuals has shown that weights given to different sensory sources signaling balance stability change dynamically (Jeka & Lackner, 1994; Peterka & Loughlin, 2004). Vision and somatosensory information have been suggested to interchange primacy in balance control depending on their reliability (Oie, Kiemel, & Jeka, 2002). Reduced tactile and joint position sensation of the lower limbs has been found to be associated with balance instability both in conditions of full sensory information and under sensory restriction (Lord, Clark, & Webster, 1991; Niam, Cheung, Sullivan, Kent, & Gu, 1999). As an additional point on this matter relevant for understanding interhemispheric asymmetry, Goble et al. (2011) showed results suggesting right cerebral hemisphere specialization for processing of proprioceptive information associated with body balance stability. From these observations, one could conjecture that decreased balance stability following right hemisphere damage is due to a reduced capacity of processing proprioceptive information, which would become more evident under visual deprivation (Bonan et al., 2004; Manor et al., 2010). In these cases, failure of intersensory reweighting in the absence of vision might be due to a decreased capacity to 9

Human Movement Science xxx (xxxx) xxx–xxx

C.A. Fernandes et al.

Fig. 5. Experiment 2, average values (standard errors in vertical bars) for the groups (RHD, right hemisphere damage; LHD, left hemisphere damage; CO, control), in the conditions of full vision and visual occlusion for the following variables: onset latency (A) and normalized magnitude (B) of activation of the muscle gastrocnemius medialis (GM). Results are shown separately for the nonparetic and paretic legs for participants with hemisphere damage and for the corresponding legs of controls. Single asterisks indicate significant differences from controls, and double asterisks significant differences from LHD and controls (significant intra-group between-leg differences for the hemisphere damage groups are not indicated), P values ≤ .05.

process proprioception in the damaged right cerebral hemisphere. Based on this rationale, it might be expected that under visual occlusion balance control deficits due to right hemisphere damage would be revealed, with magnification of the balance deficits observed in the condition of full sensory information. Contradictory to this expectation, our results showed that poor balance control in individuals with right cerebral hemisphere damage was not associated with sensory manipulation. Vision deprivation in quiet balance (Experiment 1A) increased amplitude and velocity of body balance sway in similar proportions across the hemisphere damage and control groups, with no interactions that might indicate a differential impairment of balance control induced by absence of vision in individuals with right hemisphere damage. Increasing the difficulty of tactile and proprioceptive processing from the feet soles and ankles by testing balance on a malleable surface (Experiment 1B) also failed to reveal a distinctive effect of right hemisphere damage in comparison with the other groups. In the task of balance recovery from a mechanical perturbation (Experiment 2) the same conclusion was drawn, with manipulation of vision being ineffective to produce differential effects between the groups in comparison with full vision. Proportional decline of balance stability under visual occlusion and somatosensory distortion across groups suggests that right cerebral hemisphere damage did not impair the capacity of reweighting sensorimotor integration to modulate the use of afference from somatosensory receptors located on the feet soles (Meyer, Oddsson, & De Luca, 2004), and in muscles spindles of the ankles (Thompson, Belanger, & Fung, 2011) and of the upper legs or trunk (Bloem, Allum, Carpenter, Verschuuren, & Honegger, 2002) to stabilize upright balance. An alternative argument could be made that increased deficits of balance control observed in individuals with right hemisphere damage might be associated with hemineglect (cf. Perennou et al., 1999). Supporting that argument, previous research has shown a 10

Human Movement Science xxx (xxxx) xxx–xxx

C.A. Fernandes et al.

larger proportion of hemineglect in the visual and tactile dimensions following acute damage to the right in comparison to the left cerebral hemisphere (Stone, Halligan, & Greenwood, 1993). On the other hand, evidence has been presented that while damage to the right cerebral hemisphere leads to poor balance control, no correlation has been found between hemineglect and quiet or dynamic balance stability (Ishii et al., 2010). Based on this finding, it seems improbable that hemineglect might underlie our results showing increased balance deficits from right cerebral damage. 3.3.2. Cortically-mediated balance control Lack of effect of sensory manipulation in the current series of experiments suggests that the poorer control of individuals with right hemisphere damage both in quiet and perturbed balance is due predominantly to response generation rather than sensory processing. We draw attention to the fact that the groups with right or left hemisphere damage achieved similar scores in clinical scales, while neural areas lesioned by stroke were distinct within groups. Thus, it seems that group differences in the evaluation of balance control were not due to groups dissimilarities in terms of magnitude of behavioral dysfunction or location of the damage at specific cortical sites of major importance for balance control. The fact that the lesioned sites of the right hemisphere leading to poor balance control were distinct across participants suggest that corrective responses are organized in a highly distributed network involving cortical (e.g., Mihara et al., 2012; Moro et al., 2014) and subcortical (e.g., Ferraye et al., 2014; Ouchi, Okada, Yoshikawa, Nobezawa, & Futatsubashi, 1999) areas. From this perspective, damage to delimited cortical sites seems to have the potential to affect the output of this large hemispheric network specialized in the regulation of balance stability. The primary motor cortex (Bolton, Williams, Staines, & McIlroy, 2012; Taube et al., 2006), prefrontal cortex (Mihara, Miyai, Hatakenaka, Kubota, & Sakoda, 2008; Mihara et al., 2012; Moro et al., 2014), and the supplementary motor area (Ferraye et al., 2014; Fujimoto et al., 2014; Marlin, Mochizuki, Staines, & McIlroy, 2014; Mierau, Hulsdunker, & Struder, 2015) have been found to participate in the circuitry dedicated to the regulation of body balance. Participation of higher order neural structures in balance control can be conceived to allow for appropriate selection, coordination and scalability of muscular responses to maintain balance stability or to recover body balance following an unexpected perturbation (see Teixeira, 2017, for a review). Increased hip mobilization to respond to the mechanical perturbation in right hemisphere damage (Experiment 2) indicates an adaptive strategy employed to deal with perturbations potentially more destabilizing of balance for those individuals. Such modulatory postural responses require adjustments in the coordination between body segments and in the magnitude of activation of the different muscle groups for the generation of a suitable response to the perturbation (de Lima-Pardini et al., 2014). Coordinated postural adjustments are conceived to require participation of higher order centers of control, rather than peripheral reflex circuitries exclusively. In line with this proposition, Mochizuki, Boe, Marlin, and McIlRoy (2010) found that amplitude of negative potential signals from electroencephalography scales to magnitude of balance perturbation (see also Varghese et al., 2014), with further investigation showing that signals amplitudes are larger at frontocentral cortical sites (Mochizuki, Sibley, Cheung, & McIlroy, 2009). Additional research has shown that challenging quiet standing tasks induce higher neural activation in different cortical areas (Hulsdunker, Mierau, Neeb, Kleinoder, & Struder, 2015; Varghese, Beyer, Williams, Miyasike-daSilva, & McIlroy, 2015), and reduced intracortical inhibition (Papegaaij et al., 2016). Those findings converge to indicate the relevance of cortical structures in the adaptive control of automatic postural responses to self- and externallygenerated perturbations. Our results suggest that the right cerebral hemisphere plays a prominent role in the control of postural responses to both kinds of perturbation. 3.3.3. On the hypothesis of right hemisphere specialization for impedance control The right cerebral hemisphere has been theorized to be specialized for impedance control, based on fast feedback mechanisms, ensuring steady state stability and positional stabilization in response to unanticipated mechanical perturbations (Sainburg, 2014; Yadav & Sainburg, 2011; Yadav & Sainburg, 2014). Steady state stability is an important property allowing for maintenance of stable balance in quiet stance, whereas positional stabilization in response to perturbation is critical for generating appropriate reactive responses for recovering a stable vertical orientation of the body following an unanticipated mechanical perturbation to stance. Our results are consistent with those predictions by showing that damage to the right cerebral hemisphere by stroke led to poorer balance control both in quiet and perturbed balance. Impedance control has been proposed to be implemented through different mechanisms underlying the stiffness- and viscous-like properties of the movements (Shadmehr & Arbib, 1992). More recently, modulation of impedance has been proposed to be made by means of adjusting proprioceptive reflex gains and thresholds (Kurtzer, 2015; Mutha, Boulinguez, & Sainburg, 2008; Pruszynski & Scott, 2012). Pruszynski and Scott (2012) have theorized that long-latency reflexes are modulated online through a neural network comprising cortical areas for movement regulation. From this model, it has been conceptualized that reflex circuitries are continuously adjusted by descending signals to incorporate higher order output into balance control (Coelho & Teixeira, 2017). Evidence that impedance control is impaired by brain damage has been provided by findings of deficit in long-latency stretch reflex amplitude adaptation in stroke survivors in the performance of a task demanding positional stability (Trumbower, Finley, Shemmell, Honeycutt, & Perreault, 2013). This latter finding suggests that uni-hemispheric cortical damage leads to a reduced capability to regulate reflexes to deal with postural perturbations in tasks requiring fast changes in muscular activation to recover positional stability. Yadav and Sainburg (Sainburg, 2014; Yadav & Sainburg, 2011, 2014) have proposed a hybrid model of hemispheric specialization for impedance control, with more evident manifestation of that specialization in the control of the contralateral limbs, but with its manifestation in the ipsilateral body segments as well. At this point, our results (Experiment 2) are in disagreement with the model’s prediction, as we found a similar pattern of muscular activation between the RHD’s nonparetic leg and the homologous leg of LHD and controls. However, some points should be considered about this finding. First, using a single force platform was a limitation in our analysis, preventing the evaluation of forces applied individually through each foot to the ground to keep balance stability (cf. 11

Human Movement Science xxx (xxxx) xxx–xxx

C.A. Fernandes et al.

Mansfield, Danells, Inness, Mochizuki, & McIlroy, 2011). While muscular activation is an indicator of the integrity of the motor system to respond to a perturbation, variables from EMG recordings do not allow for a conclusion about the effective participation of each leg for balance recovery. Second, the delayed and weaker activation of RHD’s paretic leg in our test required compensatory stronger responses with the nonparetic leg to recover balance following the perturbation. The compensatory activation of the ipsilesional leg has been suggested from previous results showing lower balance stability in situations that hemisphere damage patients failed to compensate for the control deficits of the paretic leg with increased activation of the nonparetic leg (Genthon et al., 2008). Further evidence of between-leg compensation following stroke has been found in the task of fast arm raising while standing, with compensation for weaker activation of the paretic leg muscles by increased magnitude of muscular activation in the nonparetic leg (Garland, Willems, Ivanova, & Miller, 2003). As an additional point on this matter, it can be expected that in individuals with stroke body weight is borne predominantly on the nonparetic leg (Mansfield et al., 2013) requiring stronger responses with that leg to a sudden balance perturbation. Considering this between-leg compensation over time, it is plausible that the apparent normal-like pattern of muscular activation of the nonparetic leg in the individuals with hemisphere damage results from use-induced adaptation (cf. Fujimoto et al., 2014) promoted by daily living standing following the stroke. Thus, a possible interpretation for the lack of effect of right hemisphere damage on ipsilateral muscular activation is that the nonparetic leg compensates for the weaker activation of the paretic leg, which our results showed to be particularly low and delayed in individuals with right hemisphere damage. From this interpretation, the acute increased body balance demand on the nonparetic leg and use-induced adaptation over time may underlie the apparent normal-like muscular activation of this leg in the situation of perturbed stance, leading to results different from those expected from the hybrid model of hemisphere specialization (Sainburg, 2014; Yadav & Sainburg, 2011, 2014). In this sense, it should be noticed that while in manual tasks movements of each arm can be assessed individually without mutual interference, in stance control the two legs work in collaboration to maintain or recover body balance. From this rationale, it becomes apparent that whereas in unimanual movements ipsilesional deficits can be observed in movement control (e.g., Schaefer, Haaland, & Sainburg, 2007), analysis of standing seems to be insensitive to reveal ipsilesional deficits in the lower limbs in chronic hemisphere damage patients. To attenuate the impact of these factors in the interpretation of results in future studies, we suggest that the evaluation of balance control be made in the acute phase following stroke, preventing extensive daily adaptation. 3.3.4. Conclusions As a conclusion, results from the two experiments reported in the present study showed that damage to the right in comparison with the left cerebral hemisphere by stroke led to increased impairment of balance control both in quiet and perturbed contexts. Manipulation of sensory information had a similar effect across hemisphere damage and nondisabled control participants, suggesting that impoverished postural control due to cerebral damage by stroke is not due to a reduced capacity of reweighing sensory information. We interpret our results as indicating specialization of the right cerebral hemisphere for body balance control, consistently with the theorization of preeminence of this hemisphere for impedance regulation (Sainburg, 2014). We propose that cerebral hemisphere specialization should not be understood as an absolute dominance of the right over the left cerebral hemisphere for balance control, but as a functional superiority of the right hemisphere at processing functions relevant for balance control leading to its greater proportional influence on stabilization and recovery of upright balance. As an observation, care should be exercised in the comparison of our results with other investigations using displacement of the base of support to generate balance perturbation. As our experimental setup required participants to lean against a pulling load before perturbation, reactive responses may differ regarding responses to perturbations applied while one keeps a neutral unloaded standing posture. As a major limitation of our investigation, the cerebral damage groups were relatively limited in size. So, our conclusions should be considered cautiously before additional investigation is conducted on increased numbers of participants. Additionally, estimation of latency of muscular activation onset is dependent on baseline variability, which may be different between groups. For this reason, conclusions based on this variable should take this point into consideration. Acknowledgments The authors would like to thank Paulo Lotufo, Isabela Benseñor e Alessandra Goulart for their technical support, and to Leon Barbosa for his assistance in data collection. This study was funded through grants provided by the São Paulo Research Foundation, Brazil (FAPESP, CAF #2011/16667-9, ARM #2011/18173-3, and LAT #2012/05769-8), and by the Brazilian Council of Science and Technology (CNPq, LAT #302628/2013-4). References Abe, I. M., Goulart, A. C., Santos Junior, W. R., Lotufo, P. A., & Bensenor, I. M. (2010). Validation of a stroke symptom questionnaire for epidemiological surveys. Sao Paulo Medical Journal, 128, 225–231. Ansari, N. N., Naghdi, S., Arab, T. K., & Jalaie, S. (2008). The interrater and intrarater reliability of the Modified Ashworth Scale in the assessment of muscle spasticity: Limb and muscle group effect. NeuroRehabilitation, 23, 231–237. Azzi, N. M., Coelho, D. B., & Teixeira, L. A. (2017). Automatic postural responses are generated according to feet orientation and perturbation magnitude. Gait and Posture, 57, 172–176. Bensoussan, L., Viton, J. M., Schieppati, M., Collado, H., Milhe de Bovis, V., Mesure, S., & Delarque, A. (2007). Changes in postural control in hemiplegic patients after stroke performing a dual task. Archives of Physical Medicine and Rehabilitation, 88, 1009–1015. Bertolucci, P. H., Brucki, S. M., Campacci, S. R., & Juliano, Y. (1994). The Mini-Mental State Examination in a general population: Impact of educational status. Arquivos de Neuropsiquiatria, 52, 1–7. Bloem, B. R., Allum, J. H., Carpenter, M. G., Verschuuren, J. J., & Honegger, F. (2002). Triggering of balance corrections and compensatory strategies in a patient with

12

Human Movement Science xxx (xxxx) xxx–xxx

C.A. Fernandes et al.

total leg proprioceptive loss. Experimental Brain Research, 142, 91–107. Bohannon, R. W., Smith, M. B., & Larkin, P. A. (1986). Relationship between independent sitting balance and side of hemiparesis. Physical Therapy, 66, 944–945. Bolton, D. A., Williams, L., Staines, W. R., & McIlroy, W. E. (2012). Contribution of primary motor cortex to compensatory balance reactions. BMC Neuroscience, 13, 102. Bonan, I. V., Colle, F. M., Guichard, J. P., Vicaut, E., Eisenfisz, M., Tran Ba Huy, P., & Yelnik, A. P. (2004). Reliance on visual information after stroke. Part I: Balance on dynamic posturography. Archives of Physical Medicine and Rehabilitation, 85, 268–273. Coelho, D. B., & Teixeira, L. A. (2017). Cognition and balance control: Does processing of explicit contextual cues of impending perturbations modulate automatic postural responses? Experimental Brain Research, 235, 2375–2390. de Lima-Pardini, A. C., Coelho, D. B., Silva, M. B., Azzi, N. M., Martinelli, A. R., Horak, F. B., & Teixeira, L. A. (2014). Aging increases flexibility of postural reactive responses based on constraints imposed by a manual task. Frontiers in Aging Neuroscience, 6, 327. Di Fabio, R. P. (1987). Lower extremity antagonist muscle response following standing perturbation in subjects with cerebrovascular disease. Brain Research, 406, 43–51. Di Fabio, R. P., Badke, M. B., & Duncan, P. W. (1986). Adapting human postural reflexes following localized cerebrovascular lesion: Analysis of bilateral long latency responses. Brain Research, 363, 257–264. Elias, L. J., Bryden, M. P., & Bulman-Fleming, M. B. (1998). Footedness is a better predictor than is handedness of emotional lateralization. Neuropsychologia, 36, 37–43. Ferraye, M. U., Debû, B., Heil, L., Carpenter, M., Bloem, B. R., & Toni, I. (2014). Using motor imagery to study the neural substrates of dynamic balance. PLoS One, 9, e91183. Fugl-Meyer, A. R., Jääskö, L., Leyman, I., Olsson, S., & Steglind, S. (1975). The post-stroke hemiplegic patient. I. A method for evaluation of physical performance. Scandinavian Journal of Rehabilitation Medicine. Supplement, 7, 13–31. Fujimoto, H., Mihara, M., Hattori, N., Hatakenaka, M., Kawano, T., Yagura, H., ... Mochizuki, H. (2014). Cortical changes underlying balance recovery in patients with hemiplegic stroke. NeuroImage, 85, 547–554. Garland, S. J., Willems, D. A., Ivanova, T. D., & Miller, K. J. (2003). Recovery of standing balance and functional mobility after stroke. Archives of Physical Medicine and Rehabilitation, 84, 1753–1759. Genthon, N., Rougier, P., Gissot, A. S., Froger, J., Pélissier, J., & Pérennou, D. (2008). Contribution of each lower limb to upright standing in stroke patients. Stroke, 39, 1793–1799. Goble, D. J., Coxon, J. P., Van Impe, A., Geurts, M., Doumas, M., Wenderoth, N., & Swinnen, S. P. (2011). Brain activity during ankle proprioceptive stimulation predicts balance performance in young and older adults. The Journal of Neuroscience, 31, 16344–16352. Hulsdunker, T., Mierau, A., Neeb, C., Kleinoder, H., & Struder, H. K. (2015). Cortical processes associated with continuous balance control as revealed by EEG spectral power. Neuroscience Letters, 592, 1–5. Ioffe, M. E., Chernikova, L. A., Umarova, R. M., Katsuba, N. A., & Kulikov, M. A. (2010). Learning postural tasks in hemiparetic patients with lesions of left versus right hemisphere. Experimental Brain Research, 201, 753–761. Ishii, F., Matsukawa, N., Horiba, M., Yamanaka, T., Hattori, M., Wada, I., & Ojika, K. (2010). Impaired ability to shift weight onto the non-paretic leg in right-cortical brain-damaged patients. Clinical Neurology and Neurosurgery, 112, 406–412. Jeka, J. J., & Lackner, J. R. (1994). Fingertip contact influences human postural control. Experimental Brain Research, 100, 495–502. Kurtzer, I. L. (2015). Long-latency reflexes account for limb biomechanics through several supraspinal pathways. Frontiers in Integrative Neuroscience, 8, 99. Laufer, Y., Sivan, D., Schwarzmann, R., & Sprecher, E. (2003). Standing balance and functional recovery of patients with right and left hemiparesis in the early stages of rehabilitation. Neurorehabilitation and Neural Repair, 17, 207–213. Lord, S. R., Clark, R. D., & Webster, I. W. (1991). Postural stability and associated physiological factors in a population of aged persons. Journal of Gerontology, 46, M69–M76. Manor, B., Hu, K., Zhao, P., Selim, M., Alsop, D., Novak, P., ... Novak, V. (2010). Altered control of postural sway following cerebral infarction: A cross-sectional analysis. Neurology, 74, 458–464. Mansfield, A., Danells, C. J., Inness, E., Mochizuki, G., & McIlroy, W. E. (2011). Between-limb synchronization for control of standing balance in individuals with stroke. Clinical Biomechanics, 26, 312–317. Mansfield, A., Danells, C. J., Zettel, J. L., Black, S. E., & McIlroy, W. E. (2013). Determinants and consequences for standing balance of spontaneous weight-bearing on the paretic side among individuals with chronic stroke. Gait & Posture, 38, 428–432. Mansfield, A., Inness, E. L., Wong, J. S., Fraser, J. E., & McIlroy, W. E. (2013). Is impaired control of reactive stepping related to falls during inpatient stroke rehabilitation? Neurorehabilitation and Neural Repair, 27, 526–533. Marigold, D. S., & Eng, J. J. (2006a). Altered timing of postural reflexes contributes to falling in persons with chronic stroke. Experimental Brain Research, 171, 459–468. Marigold, D. S., & Eng, J. J. (2006b). The relationship of asymmetric weight-bearing with postural sway and visual reliance in stroke. Gait & Posture, 23, 249–255. Marlin, A., Mochizuki, G., Staines, W. R., & McIlroy, W. E. (2014). Localizing evoked cortical activity associated with balance reactions: Does the anterior cingulate play a role? Journal of Neurophysiology, 111, 2634–2643. Martinelli, A. R., Coelho, D. B., Magalhaes, F. H., Kohn, A. F., & Teixeira, L. A. (2015). Light touch modulates balance recovery following perturbation: From fast response to stance restabilization. Experimental Brain Research, 233, 1399–1408. Meyer, P. F., Oddsson, L. I., & De Luca, C. J. (2004). Reduced plantar sensitivity alters postural responses to lateral perturbations of balance. Experimental Brain Research, 157, 526–536. Mierau, A., Hulsdunker, T., & Struder, H. K. (2015). Changes in cortical activity associated with adaptive behavior during repeated balance perturbation of unpredictable timing. Frontiers in Behavioral Neuroscience, 9, 1–12. Mihara, M., Miyai, I., Hatakenaka, M., Kubota, K., & Sakoda, S. (2008). Role of the prefrontal cortex in human balance control. NeuroImage, 43, 329–336. Mihara, M., Miyai, I., Hattori, N., Hatakenaka, M., Yagura, H., Kawano, T., & Kubota, K. (2012). Cortical control of postural balance in patients with hemiplegic stroke. NeuroReport, 23, 314–319. Mochizuki, G., Boe, S., Marlin, A., & McIlRoy, W. E. (2010). Perturbation-evoked cortical activity reflects both the context and consequence of postural instability. Neuroscience, 170, 599–609. Mochizuki, G., Sibley, K. M., Cheung, H. J., & McIlroy, W. E. (2009). Cortical activity prior to predictable postural instability: Is there a difference between selfinitiated and externally-initiated perturbations? Brain Research, 1279, 29–36. Moro, S. B., Bisconti, S., Muthalib, M., Spezialetti, M., Cutini, S., Ferrari, M., ... Quaresima, V. (2014). A semi-immersive virtual reality incremental swing balance task activates prefrontal cortex: A functional near-infrared spectroscopy study. Neuroimage, 85, 451–460. Mutha, P. K., Boulinguez, P., & Sainburg, R. L. (2008). Visual modulation of proprioceptive reflexes during movement. Brain Research, 1246, 54–69. Niam, S., Cheung, W., Sullivan, P. E., Kent, S., & Gu, X. (1999). Balance and physical impairments after stroke. Archives of Physical Medicine and Rehabilitation, 80, 1227–1233. Oie, K. S., Kiemel, T., & Jeka, J. J. (2002). Multisensory fusion: Simultaneous re-weighting of vision and touch for the control of human posture. Brain Research. Cognitive Brain Research, 14, 164–176. Oldfield, R. C. (1971). The assessment and analysis of handedness: The Edinburgh inventory. Neuropsychologia, 9, 97–113. Ouchi, Y., Okada, H., Yoshikawa, E., Nobezawa, S., & Futatsubashi, M. (1999). Brain activation during maintenance of standing postures in humans. Brain, 122(Pt 2), 329–338. Papegaaij, S., Taube, W., van Keeken, H. G., Otten, E., Baudry, S., & Hortobagyi, T. (2016). Postural challenge affects motor cortical activity in young and old adults. Experimental Gerontology, 73, 78–85. Perennou, D. A., Benaim, C., Rouget, E., Rousseaux, M., Blard, J. M., & Pelissier, J. (1999). Postural balance following stroke: Towards a disadvantage of the right brain-damaged hemisphere. Revue Neurologique, 155, 281–290.

13

Human Movement Science xxx (xxxx) xxx–xxx

C.A. Fernandes et al.

Perennou, D. A., Mazibrada, G., Chauvineau, V., Greenwood, R., Rothwell, J., Gresty, M. A., & Bronstein, A. M. (2008). Lateropulsion, pushing and verticality perception in hemisphere stroke: A causal relationship? Brain, 131, 2401–2413. Peterka, R. J., & Loughlin, P. J. (2004). Dynamic regulation of sensorimotor integration in human postural control. Journal of Neurophysiology, 91, 410–423. Peurala, S. H., Kononen, P., Pitkanen, K., Sivenius, J., & Tarkka, I. M. (2007). Postural instability in patients with chronic stroke. Restorative Neurology and Neuroscience, 25, 101–108. Pollock, C. L., Ivanova, T. D., Hunt, M. A., & Garland, S. J. (2015). Behavior of medial gastrocnemius motor units during postural reactions to external perturbations after stroke. Clinical Neurophysiology, 126, 1951–1958. Pruszynski, J. A., & Scott, S. H. (2012). Optimal feedback control and the long-latency stretch response. Experimental Brain Research, 218, 341–359. Rode, G., Tiliket, C., & Boisson, D. (1997). Predominance of postural imbalance in left hemiparetic patients. Scandinavian Journal of Rehabilitation Medicine, 29, 11–16. Sainburg, R. L. (2014). Convergent models of handedness and brain lateralization. Frontiers in Psychology, 5, 1092. Salot, P., Patel, P., & Bhatt, T. (2016). Reactive balance in individuals with chronic stroke: Biomechanical factors related to perturbation-induced backward falling. Physical Therapy, 96, 338–347. Schabowsky, C. N., Hidler, J. M., & Lum, P. S. (2007). Greater reliance on impedance control in the nondominant arm compared with the dominant arm when adapting to a novel dynamic environment. Experimental Brain Research, 182, 567–577. Schaefer, S. Y., Haaland, K. Y., & Sainburg, R. L. (2007). Ipsilesional motor deficits following stroke reflect hemispheric specializations for movement control. Brain, 130, 2146–2158. Schaefer, S. Y., Haaland, K. Y., & Sainburg, R. L. (2009). Dissociation of initial trajectory and final position errors during visuomotor adaptation following unilateral stroke. Brain Research, 1298, 78–91. Schaefer, S. Y., Mutha, P. K., Haaland, K. Y., & Sainburg, R. L. (2012). Hemispheric specialization for movement control produces dissociable differences in online corrections after stroke. Cerebral Cortex, 22, 1407–1419. Shadmehr, R., & Arbib, M. A. (1992). A mathematical analysis of the force-stiffness characteristics of muscles in control of a single joint system. Biological Cybernetics, 66, 463–477. Stone, S., Halligan, P., & Greenwood, R. (1993). The incidence of neglect phenomena and related disorders in patients with an acute right or left hemisphere stroke. Age and Ageing, 22, 46–52. Taube, W., Schubert, M., Gruber, M., Beck, S., Faist, M., & Gollhofer, A. (2006). Direct corticospinal pathways contribute to neuromuscular control of perturbed stance. Journal of Applied Physiology, 101, 420–429. Teixeira, L. A. (2017). Interlateral asymmetries of body balance control resulting from cerebral stroke. Locomotion and posture in older adults (pp. 291–305). Springer. Thompson, C., Belanger, M., & Fung, J. (2011). Effects of plantar cutaneo-muscular and tendon vibration on posture and balance during quiet and perturbed stance. Human Movement Science, 30, 153–171. Trumbower, R. D., Finley, J. M., Shemmell, J. B., Honeycutt, C. F., & Perreault, E. J. (2013). Bilateral impairments in task-dependent modulation of the long-latency stretch reflex following stroke. Clinical Neurophysiology, 124, 1373–1380. Ustinova, K. I., Chernikova, L. A., Ioffe, M. E., & Sliva, S. S. (2001). Impairment of learning the voluntary control of posture in patients with cortical lesions of different locations: The cortical mechanisms of pose regulation. Neuroscience and Behavioral Physiology, 31, 259–267. Varghese, J. P., Beyer, K. B., Williams, L., Miyasike-daSilva, V., & McIlroy, W. E. (2015). Standing still: Is there a role for the cortex? Neuroscience Letters, 590, 18–23. Varghese, J. P., Marlin, A., Beyer, K. B., Staines, W. R., Mochizuki, G., & McIlroy, W. E. (2014). Frequency characteristics of cortical activity associated with perturbations to upright stability. Neuroscience Letters, 578, 33–38. Wilson, J. T., Hareendran, A., Grant, M., Baird, T., Schulz, U. G., Muir, K. W., & Bone, I. (2002). Improving the assessment of outcomes in stroke: Use of a structured interview to assign grades on the modified Rankin Scale. Stroke, 33, 2243–2246. Yadav, V., & Sainburg, R. L. (2011). Motor lateralization is characterized by a serial hybrid control scheme. Neuroscience, 196, 153–167. Yadav, V., & Sainburg, R. L. (2014). Handedness can be explained by a serial hybrid control scheme. Neuroscience, 278, 385–396.

14