Neuroscience Letters 581 (2014) 75–79
Contents lists available at ScienceDirect
Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet
Verticality perception during and after galvanic vestibular stimulation Katharina Volkening a,b,c,∗,1 , Jeannine Bergmann b,c,1 , Ingo Keller b,c , Max Wuehr c , Friedemann Müller b,c , Klaus Jahn c,d a
Graduate School of Systemic Neurosciences, Ludwig-Maximilians-University, Großhaderner Str. 2, 82152 Planegg-Martinsried, Germany Schoen Klinik Bad Aibling, Kolbermoorerstr. 72, 83043 Bad Aibling, Germany German Center for Vertigo and Balance Disorders (DSGZ), Ludwig-Maximilians-University, Marchioninistr. 15, 81377 Munich, Germany d Department of Neurology, Ludwig-Maximilians-University, Marchioninistr. 15, 81377 Munich, Germany b c
h i g h l i g h t s • • • •
GVS causes verticality deviations towards the anode during stimulation. After stimulation, the subjective verticals deviate towards the cathode. Aftereffects of GVS (1.5 mA, 20 min) exhibit different types of decay. GVS effects are best assessed with the subjective haptic vertical.
a r t i c l e
i n f o
Article history: Received 4 February 2014 Received in revised form 11 August 2014 Accepted 15 August 2014 Available online 23 August 2014 Keywords: Galvanic vestibular stimulation Subjective vertical Verticality perception
a b s t r a c t The human brain constructs verticality perception by integrating vestibular, somatosensory, and visual information. Here we investigated whether galvanic vestibular stimulation (GVS) has an effect on verticality perception both during and after application, by assessing the subjective verticals (visual, haptic and postural) in healthy subjects at those times. During stimulation the subjective visual vertical and the subjective haptic vertical shifted towards the anode, whereas this shift was reversed towards the cathode in all modalities once stimulation was turned off. Overall, the effects were strongest for the haptic modality. Additional investigation of the time course of GVS-induced changes in the haptic vertical revealed that anodal shifts persisted for the entire 20-min stimulation interval in the majority of subjects. Aftereffects exhibited different types of decay, with a preponderance for an exponential decay. The existence of such reverse effects after stimulation could have implications for GVS-based therapy. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Humans construct and update their sense of verticality by integrating vestibular, somatosensory, and visual input [18]. The internal estimate of verticality can be assessed by different methods, testing preferentially the visual, the tactile and the postural modalities (subjective visual, haptic, and postural vertical). These modalities can be differentially affected in patients with neurological disorders [4,23,33].
∗ Corresponding author at: Schoen Klinik Bad Aibling, Kolbermoorer Str. 72, 83043 Bad Aibling, Germany. Tel.: +49 8061 903 1903. E-mail addresses: [email protected] (K. Volkening), [email protected] (J. Bergmann), [email protected] (I. Keller), [email protected] (F. Müller), [email protected] (K. Jahn). 1 Joint first authors. http://dx.doi.org/10.1016/j.neulet.2014.08.028 0304-3940/© 2014 Elsevier Ireland Ltd. All rights reserved.
Transmastoidal galvanic vestibular stimulation (GVS) acts on afferents from the otoliths and the semicircular canals [8]. It was also shown to affect subjects’ perception of verticality. During stimulation the subjective visual (SVV) and the subjective haptic vertical (SHV) deviate towards the anode [21,22,24]. GVS causes eye torsion and nystagmus via the vestibular-ocular reflex [13] and – with the head upright – body tilt towards the anode via vestibulo-spinal reflexes [31]. However, existing studies on verticality perception have only examined the online effects of GVS, since judgments of verticality were always generated during stimulation intervals. In the oculomotor domain, GVS is known though to elicit reverse responses towards the cathode after stimulation [20,26]. It is not known if these aftereffects exist for the subjective verticals and whether there are any differences between modalities. This is of relevance since the time course and magnitude of effects and aftereffects of GVS on verticality perception might influence responses to
76
K. Volkening et al. / Neuroscience Letters 581 (2014) 75–79
therapeutic interventions. Thus, the purpose of this study was to examine the influence of GVS on different subjective verticals (visual, haptic, and postural) both during and after its application. 2. Materials and methods 2.1. Galvanic vestibular stimulation Bilateral bipolar GVS was delivered by a battery-driven, direct current stimulator (neuroConn Ilmenau, Germany). Electrodes were covered with natrium-chloride soaked sponges (30 cm2 each). Current was ramped up (in steps of 0.1 mA/s) to 1.5 mA and turned off at the end of the stimulation period. 2.2. Assessment of verticality perception 2.2.1. Subjective visual vertical (SVV) The SVV was assessed with the so-called bucket test. The bucket was rotated by the examiner, and the seated subjects indicated when they visually perceived a dark line (13 cm long, 0.3 cm wide, at 23 cm distance) as being vertical [35]. 2.2.2. Subjective haptic vertical (SHV) The SHV was measured with a rod (27 cm long, 1 cm wide) mounted onto a vertical plate 40 cm in front of the subject (see [10] for a similar device). While seated and blindfolded, the subjects’ task was to adjust the wooden rod, with their right hand using a precision grip until they perceived it to be in a vertical position. They were not allowed to touch the device’s plate or the desk. 2.2.3. Subjective postural vertical (SPV) The SPV was measured in the Spacecurl [2], a cardanic suspension apparatus that consists of three concentric rings. The blindfolded subject stood in the centre of the apparatus on a platform that was attached to the midmost ring. The device was tilted in the frontal plane, and subjects had to indicate when they felt they were in an upright position. There was no time limit for individual adjustments. Six adjustments per condition were performed in randomized order of starting positions (for SPV 12◦ , 15◦ & 18◦ ; for SVV and SHV 15◦ , 25◦ & 40◦ ). Half of the adjustments started from a clockwise, half from a counter-clockwise position. The six adjustments were averaged for each condition and modality to calculate the SPV, SVV, and SHV. Data were normalized so that positive values indicated deviations from the earth vertical to the side of the anode, and negative values, deviations in the direction of the cathode. 2.3. Experiments The Ethics Committee of the Ludwig-Maximilians-University Munich approved this study. All subjects provided their written informed consent. 2.3.1. Exp. 1: Manipulation of subjective verticals To investigate online effects and aftereffects of GVS on verticality perception across different modalities, ten healthy subjects participated in experiment 1 (mean age: 59 years, SD: ±6; 5 females). All subjects performed the SPV, the SVV, and the SHV immediately before (baseline), during, and 3 min after a period of GVS. The experiment was conducted on two consecutive days with a fixed sequential order: the SPV on day 1, the SVV and SHV on day 2. The polarity of the GVS current was varied between subjects. Stimulation was applied for the duration of verticality adjustments (4–8 min).
2.3.2. Exp. 2: Time course A second experiment was designed to determine the time course of online effects and aftereffects of GVS on verticality perception. Since the haptic modality was most responsive to GVS, the time course was studied for the SHV. Fourteen healthy subjects (mean age: 34 years, SD: ±6; 7 females), not included in experiment 1, were tested in experiment 2. The subjects repeatedly performed the SHV during and after a 20-min period of GVS. Six SHV adjustments were performed immediately before starting GVS (baseline), 0.5, 5, 10, 15, and 20 min after starting the stimulation (conditions 1–5) and at the same time points after terminating GVS (conditions 6–10) (total of 66 adjustments). All subjects were stimulated with the cathode over the left and the anode over the right mastoid.
2.4. Data analysis A one-way ANOVA with the within-subject factor modality was used to evaluate differences between modalities at baseline. To determine any differences in verticality adjustments across time points (baseline, during, after GVS), and modalities (SVV, SHV, SPV) a factorial repeated-measures ANOVA was performed with two within-subject factors (time and modality). Another factorial repeated-measures ANOVA (within-subjects factors modality and type of effect) was conducted to compare the magnitudes of online effects and aftereffects across modalities. Effect magnitudes were calculated as absolute differences between baseline and during GVS (online effect) and during and after GVS (aftereffect). If sphericity was violated in an ANOVA, Greenhouse–Geisser correction was applied. In case of significant results subsequent multiple comparisons were performed and Bonferroni corrected. For experiment 2 differences across conditions were analyzed using a one-way repeated-measures ANOVA and subsequent repeated contrasts. Verticality adjustments during (condition 1–5) and after GVS (condition 6–10) were grouped and compared using a paired t-test. A least-squares method was used to estimate which type of model (no decay, linear decay or exponential decay) best fits the individual time course of responses in verticality perception during and after GVS. A decay time constant was calculated for subjects showing an exponential decay. The data were analyzed with Matlab (The Mathworks, Version 2011b) and SPSS Statistics (SPSS Inc., Version 17.0). The significance level for ˛ was set at 0.05.
3. Results 3.1. Exp. 1: Manipulation of subjective verticals There was no significant difference between the modalities at baseline (F(2,27) = 0.056, p = 0.945; mean ± SE: SVV = 0.8 ± 0.5◦ , SHV = 0.6 ± 0.6◦ , SPV = 0.6 ± 0.4◦ ). The factorial repeated-measures ANOVA across all time points, showed a significant main effect of time (F(1.27,11.41) = 20.852, p < 0.001), but not of modality (F(2,18) = 2.082, p = 0.154). Post hoc tests revealed significant differences between the adjustments before and during GVS (p = 0.006) and during and after GVS (p < 0.001), but not between baseline and after GVS (p = 1.000). There was a significant interaction between time and modality (F(2.39,21.50) = 9.265, p = 0.001): Differences between adjustments before and during GVS were smaller for the SPV than for the SVV and SHV; adjustments during and after GVS were also smaller for the SPV than for the SHV. Analysis of the effect magnitudes showed a significant main effect of modality (F(2,18) = 17.260, p < 0.001), but no effect of effect type (online effect vs. aftereffect) (F(1,9) = 0.520, p = 0.489) and no significant interaction (F(1.16,10.42) = 1.506, p = 0.252). Post hoc tests revealed significantly greater online effects and aftereffects for the SHV than
K. Volkening et al. / Neuroscience Letters 581 (2014) 75–79
77
Fig. 1. Online effects and aftereffects of GVS. Adjustments (mean ± standard error, normalized to baseline and side of anode) of the subjective verticals during (SVV: 0.5 ± 0.3◦ , SHV: 2.8 ± 0.5◦ , SPV: −0.1 ± 0.2) and after (SVV: −0.7 ± 0.3◦ , SHV: −2.5 ± 0.4◦ , SPV: −0.5 ± 0.2◦ ) galvanic vestibular stimulation. Positive values indicate a tilt to the side of the anode, negative values a tilt to the side of the cathode. SVV, subjective visual vertical; SHV, subjective haptic vertical, SPV, subjective postural vertical; GVS, galvanic vestibular stimulation.
for the SVV (p = 0.009) and the SPV (p = 0.004). Normalized verticality adjustments during and after stimulation are shown in Fig. 1. 3.2. Exp. 2: Time course The average SHV at baseline was −0.6 ± 0.8◦ (mean ± SE). There were significant differences between conditions (F(10,130) = 2.241, p = 0.019). Contrasts revealed differences between baseline and condition 1 (F = 7.279, p = 0.018), condition 5 and 6 (F = 10.852, p = 0.005), and condition 9 and 10 (F = 8.779, p = 0.011) (Fig. 2). Accordingly, the SHV changed when turning stimulation on and off, and the latter effect had decayed after 20 min. Condition 1 and 2 (F = 4.539, p = 0.053), and condition 4 and 5 (F = 3.557, p = 0.082) tended to be different. The SHV adjustments during GVS (conditions 1–5) differed significantly from the adjustments after stimulation (conditions 6–10) (T(69) = 4.398, p < 0.001). As a majority of subjects (n = 8) showed no decay for the online effects, no decay time constant was calculated for the responses during GVS. Half of the subjects showed an exponential decay for the aftereffect with a mean decay time constant of 19.2 ± 1.6 min (for further details see supplementary table). 4. Discussion Our results provide the first direct evidence for the presence and time course of both online effects and aftereffects of GVS on the subjective verticals across different modalities. 4.1. Manipulation of subjective verticals During stimulation subjects’ SVV and SHV – but not SPV – shifted towards the anode. The shift was strongest for the haptic modality. After stimulation was switched off, this effect reversed across all three modalities, and slightly overshot the baseline value in the direction of the cathode for the SVV and the SPV (and also for the SHV in experiment 2). Anodal shifts during GVS are consistent with previous findings [21,24,31] showing that verticality perception can be influenced by manipulating the vestibular input. Former studies demonstrated the SVV to be more malleable by GVS than the SHV, but we found the opposite to be the case [21,31]. This might be due to differences in experimental protocols causing different degrees of ocular torsion. GVS-induced ocular torsion is known to influence the SVV [34], but also to decline over time [19]. While in previous experiments current was applied individually for each trial with SVV adjustments time-locked to current onset, we stimulated continuously across conditions. Thus, GVS-induced ocular torsion was
presumably reduced in our measurements compared to recordings at stimulus onset, resulting in less pronounced effects of GVS on the SVV. For the SHV ocular torsion appears to be rather negligible [6,30]. The SHV is known to be influenced by proprioceptive and tactile cues from the hand, arm and shoulder, as well as from the trunk [9] and the neck [12]. Findings on the role of the vestibular system are mixed. Whereas the SHV of patients with unilateral vestibular nuclear lesions was only marginally affected [6], a study with healthy controls indicated a major role of the vestibular system in SHV adjustments [25]. Our results support this crucial role of the vestibular system. In contrast to the SVV and SHV, we found no significant modulation of the SPV during GVS. Similarly, Bisdorff et al. [3] reported no effect of GVS on the SPV; however, they observed a broader range, within which subjects felt vertical during stimulation. One possible explanation for these findings might be that verticality is processed by distinct pathways in different modalities and thus possibly affected differently by GVS [15]. Another explanation for the limited effect of GVS on the SPV might be the abundance of somatosensory information. Somatosensory information – if available – is known to be crucial and even sufficient for an accurate SPV estimation [1,14,22]. In our set-up, different tactile information was provided by the Spacecurl’s padded restraints around hips and the pressure distribution under the feet. When sensory input from different sources is integrated, possible sensory conflict is resolved by weighting more reliable input stronger [7]. Body tilts generate somatosensory asymmetries that improve the reliability of somatosensory input compared to symmetric signals in upright position [5,22]. Consequently, the somatosensory input in our experiment was presumably considered more reliable for estimating the SPV than the conflicting vestibular input. In contrast to these online effects, we found the reverse effect after stimulation: in all modalities the subjective vertical shifted towards the cathode. Both, online effects and aftereffects were most pronounced for the haptic modality. Aftereffects of GVS have also been reported for the oculomotor domain and body movements [19,20,27,29]. When GVS is turned off, ocular torsion and nystagmus reverse direction and are directed towards the cathode. Studies investigating effects of GVS on body movement found a body sway towards the anode during stimulation [8,31] which reversed direction after stimulation. The magnitude of this aftereffect was significantly larger than the online effect [29]. Similarly, GVS-induced perceptions of rotation were also found to change direction after stimulation. Percepts during and after stimulation had similar peak magnitudes [27].
78
K. Volkening et al. / Neuroscience Letters 581 (2014) 75–79
Fig. 2. Time course of GVS effects. Time course of haptic verticality adjustments (mean ± standard error, normalized to baseline) during and after 20 min of galvanic vestibular stimulation. Times of assessment were 0.5, 5, 10, 15, and 20 min into stimulation (condition 1–5) and at the same time points after stimulation (condition 6–10). Positive values indicate a tilt to the side of the anode, negative values a tilt to the side of the cathode. BSLN, baseline; GVS, galvanic vestibular stimulation; SHV, subjective haptic vertical. Contrasts: *p < 0.05, **p < 0.01.
The influence of GVS on verticality perception was also investigated in right-brain damaged patients. For visuo-tactile adjustments, Saj et al. [24] reported stronger online effects in patients than in healthy subjects. The largest effects were found in patients with spatial neglect who typically show counter-clockwise SVV and SHV deviations [16]. Since these deviations shifted towards the anode during GVS, Saj et al. [24] proposed that right anodal stimulation could alleviate neglect patients’ spatial deficits. However, we found that online effects reversed after switching the stimulation off – with a slight overshoot towards the cathode. Since the time course of these aftereffects is the critical factor in terms of therapy, it was investigated in experiment 2. 4.2. Time course For the SHV, we found that GVS-induced shifts lasted not only during stimulation, but also up to at least 15 min after stimulation. When stimulation started, subjects’ SHV showed an anodal shift that was strongest immediately after the onset of GVS and persisted during the entire 20-min stimulation interval in the majority of subjects. When stimulation was switched off, the SHV reversed direction and shifted towards the cathode. For half of the subjects, this aftereffect showed an exponential decay with a decay time constant of 19.2 min. These findings are in line with studies on GVSinduced eye movements and movement perceptions. Reverse eye
movement responses were reported to last up to 6 min post a 5-min stimulation interval [19]. Response strength was linearly related to current intensity. Similarly, rotation perceptions during and after stimulation had comparable peak magnitudes and time courses of decay [27]. GVS-induced aftereffects appear to be directly related to the stimulation interval. Until now, however, this has not been explicitly confirmed for the subjective verticals. The longer-lasting reversed tilts of the subjective verticals that occurred after the application of GVS reflect adaptive mechanisms during prolonged stimulation. This may have implications for the use of GVS in the rehabilitation of distorted verticality perception. Based on earlier studies, GVS was proposed as a therapy for pusher behaviour and spatial neglect [17,24]. Evidence on its effectiveness, however, is lacking. The placement of the electrodes for therapy has so far been based on the effects observed during stimulation. Thus, neglect patients’ counter-clockwise verticality deviations were ameliorated when the anode was on the right and the cathode on the left mastoid [24]. Our finding that the perceived verticals deviate to the side of the cathode after stimulation suggests that the effect after stimulation should also be taken into account when using GVS as a therapeutic tool. In some cases it might be useful to reconsider the placement of the electrodes and setting of parameters. Our data showed a decay of the aftereffect in the majority of subjects, between-subject variability was however high. Thus further investigation of long-term
K. Volkening et al. / Neuroscience Letters 581 (2014) 75–79
aftereffects of GVS is needed, especially of the impact of repetitive stimulation. This pilot study has a few limitations. One is that we did not control the subjects’ trunk and head positions during verticality adjustments. Even though GVS is known to affect head and trunk orientation, its influence on verticality adjustments is controversial [11,22,28,31]. Another limitation is that the order of the different verticality assessments in experiment 1 was not randomized. The SVV and SHV adjustments were performed in a fixed sequential order on day 2. This might have biased the results, e.g. by adding up the stimulation effects, thus affecting the measures on day 2 more than those on day 1 (SPV). However, a recent study provides evidence against this, as repetitive GVS of up to 10 sessions was not more effective than a single session of GVS for improving neglect [32]. 5. Conclusions Our results provide the first direct evidence that GVS has both online effects and aftereffects on the subjective verticals. While online effects generally persisted during the entire 20-min stimulation interval, aftereffects showed different types of decay, with a preponderance for an exponential one. Reversed tilts of the subjective verticals after the termination of GVS might have implications for the rehabilitation of spatial orientation deficits. Further studies are needed to investigate such aftereffects over a longer timescale, after repetitive stimulation and in brain-damaged patients. Authors’ contribution KV and JB (joint first authors) designed the study, recruited the participants, collected the data, performed analyses and interpretation of data and wrote the article. IK, FM, JK obtained funding, contributed to the interpretation of the results and revised the article. MW performed part of the data analyses, wrote the corresponding text and contributed to the interpretation of the results. Acknowledgements The authors thank J. Benson for copyediting the manuscript. This work was supported by funds from the German Federal Ministry of Education and Research (BMBF IFB 01EO0901). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.neulet. 2014.08.028. References [1] J. Barra, A. Marquer, R. Joassin, C. Reymond, L. Metge, V. Chauvineau, D. Perennou, Humans use internal models to construct and update a sense of verticality, Brain 133 (2010) 3552–3563. [2] J. Bergmann, M.A. Ciupe, C. Krewer, A. Schepermann, F. Müller, K. Jahn, E. Koenig, Intrarater- und Interrater-Reliabilität des Spacecurls als Messinstrument der subjektiven posturalen Vertikalen, Neurol. Rehabil. 6 (2012) 417. [3] A.R. Bisdorff, C.J. Wolsley, D. Anastasopoulos, A.M. Bronstein, M.A. Gresty, The perception of body verticality (subjective postural vertical) in peripheral and central vestibular disorders, Brain 119 (Pt 5) (1996) 1523–1534. [4] I.V. Bonan, A. Marquer, S. Eskiizmirliler, A.P. Yelnik, P.P. Vidal, Sensory reweighting in controls and stroke patients, Clin. Neurophysiol. 124 (2012) 713–722. [5] A.M. Bronstein, M. Guerraz, Visual-vestibular control of posture and gait: physiological mechanisms and disorders, Curr. Opin. Neurol. 12 (1999) 5–11. [6] A.M. Bronstein, D.A. Perennou, M. Guerraz, D. Playford, P. Rudge, Dissociation of visual and haptic vertical in two patients with vestibular nuclear lesions, Neurology 61 (2003) 1260–1262.
79
[7] M.O. Ernst, H.H. Bulthoff, Merging the senses into a robust percept, Trends Cogn. Sci. 8 (2004) 162–169. [8] R.C. Fitzpatrick, B.L. Day, Probing the human vestibular system with galvanic stimulation, J. Appl. Physiol. 96 (2004) 2301–2316. [9] B. Fourre, B. Isableu, D. Bernardin, M. Gueguen, G. Giraudet, N. Vuillerme, C. Pagano, M.A. Amorim, The role of body centre of mass on haptic subjective vertical, Neurosci. Lett. 465 (2009) 230–234. [10] J. Funk, K. Finke, H.J. Muller, R. Preger, G. Kerkhoff, Systematic biases in the tactile perception of the subjective vertical in patients with unilateral neglect and the influence of upright vs. supine posture, Neuropsychologia 48 (2010) 298–308. [11] J. Funk, K. Finke, H.J. Muller, K.S. Utz, G. Kerkhoff, Effects of lateral head inclination on multimodal spatial orientation judgments in neglect: evidence for impaired spatial orientation constancy, Neuropsychologia 48 (2010) 1616–1627. [12] M. Guerraz, M. Luyat, D. Poquin, T. Ohlmann, The role of neck afferents in subjective orientation in the visual and tactile sensory modalities, Acta Otolaryngol. 120 (2000) 735–738. [13] K. Jahn, A. Naessl, M. Strupp, E. Schneider, T. Brandt, M. Dieterich, Torsional eye movement responses to monaural and binaural galvanic vestibular stimulation: side-to-side asymmetries, Ann. N. Y. Acad. Sci. 1004 (2003) 485–489. [14] R. Joassin, V. Bonniaud, J. Barra, A. Marquer, D. Perennou, Somaesthetic perception of the vertical in spinal cord injured patients: a clinical study, Ann. Phys. Rehabil. Med. 53 (2010) 568–574. [15] H.O. Karnath, S. Ferber, J. Dichgans, The origin of contraversive pushing: evidence for a second graviceptive system in humans, Neurology 55 (2000) 1298–1304. [16] G. Kerkhoff, Multimodal spatial orientation deficits in left-sided visual neglect, Neuropsychologia 37 (1999) 1387–1405. [17] C. Krewer, K. Riess, J. Bergmann, F. Muller, K. Jahn, E. Koenig, Immediate effectiveness of single-session therapeutic interventions in pusher behaviour, Gait Posture 37 (2013) 246–250. [18] C. Lopez, O. Blanke, The thalamocortical vestibular system in animals and humans, Brain Res. Rev. 67 (2011) 119–146. [19] H.G. MacDougall, A.E. Brizuela, A.M. Burgess, I.S. Curthoys, Between-subject variability and within-subject reliability of the human eye-movement response to bilateral galvanic (DC) vestibular stimulation, Exp. Brain Res. 144 (2002) 69–78. [20] H.G. MacDougall, A.E. Brizuela, I.S. Curthoys, Linearity, symmetry and additivity of the human eye-movement response to maintained unilateral and bilateral surface galvanic (DC) vestibular stimulation, Exp. Brain Res. 148 (2003) 166–175. [21] F. Mars, K. Popov, J.L. Vercher, Supramodal effects of galvanic vestibular stimulation on the subjective vertical, Neuroreport 12 (2001) 2991–2994. [22] F. Mars, J.L. Vercher, K. Popov, Dissociation between subjective vertical and subjective body orientation elicited by galvanic vestibular stimulation, Brain Res. Bull. 65 (2005) 77–86. [23] D. Perennou, C. Piscicelli, G. Barbieri, M. Jaeger, A. Marquer, J. Barra, Measuring verticality perception after stroke: why and how? Neurophysiol. Clin. 44 (2014) 25–32. [24] A. Saj, J. Honore, M. Rousseaux, Perception of the vertical in patients with right hemispheric lesion: effect of galvanic vestibular stimulation, Neuropsychologia 44 (2006) 1509–1512. [25] J.R. Schuler, C.J. Bockisch, D. Straumann, A.A. Tarnutzer, Precision and accuracy of the subjective haptic vertical in the roll plane, BMC Neurosci. 11 (2010) 83. [26] A. Severac Cauquil, M. Faldon, K. Popov, B.L. Day, A.M. Bronstein, Short-latency eye movements evoked by near-threshold galvanic vestibular stimulation, Exp. Brain Res. 148 (2003) 414–418. [27] R.J. St George, B.L. Day, R.C. Fitzpatrick, Adaptation of vestibular signals for self-motion perception, J. Physiol. 589 (2011) 843–853. [28] A.A. Tarnutzer, C.J. Bockisch, D. Straumann, Roll-dependent modulation of the subjective visual vertical: contributions of head- and trunk-based signals, J. Neurophysiol. 103 (2010) 934–941. [29] C.M. Tax, A.P. Bom, R.L. Taylor, N. Todd, K.K. Cho, R.C. Fitzpatrick, M.S. Welgampola, The galvanic whole-body sway response in health and disease, Clin. Neurophysiol. 124 (2013) 2036–2045. [30] S.W. Wade, I.S. Curthoys, The effect of ocular torsional position on perception of the roll-tilt of visual stimuli, Vision Res. 37 (1997) 1071–1078. [31] D.L. Wardman, J.L. Taylor, R.C. Fitzpatrick, Effects of galvanic vestibular stimulation on human posture and perception while standing, J. Physiol. 551 (2003) 1033–1042. [32] D. Wilkinson, O. Zubko, M. Sakel, S. Coulton, T. Higgins, P. Pullicino, Galvanic vestibular stimulation in hemi-spatial neglect, Front. Integr. Neurosci. 8 (2014) 4. [33] A.P. Yelnik, F.O. Lebreton, I.V. Bonan, F.M. Colle, F.A. Meurin, J.P. Guichard, E. Vicaut, Perception of verticality after recent cerebral hemispheric stroke, Stroke 33 (2002) 2247–2253. [34] R. Zink, S.F. Bucher, A. Weiss, T. Brandt, M. Dieterich, Effects of galvanic vestibular stimulation on otolithic and semicircular canal eye movements and perceived vertical, Electroencephalogr. Clin. Neurophysiol. 107 (1998) 200–205. [35] A. Zwergal, N. Rettinger, C. Frenzel, M. Dieterich, T. Brandt, M. Strupp, A bucket of static vestibular function, Neurology 72 (2009) 1689–1692.
Contents lists available at ScienceDirect
Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet
Verticality perception during and after galvanic vestibular stimulation Katharina Volkening a,b,c,∗,1 , Jeannine Bergmann b,c,1 , Ingo Keller b,c , Max Wuehr c , Friedemann Müller b,c , Klaus Jahn c,d a
Graduate School of Systemic Neurosciences, Ludwig-Maximilians-University, Großhaderner Str. 2, 82152 Planegg-Martinsried, Germany Schoen Klinik Bad Aibling, Kolbermoorerstr. 72, 83043 Bad Aibling, Germany German Center for Vertigo and Balance Disorders (DSGZ), Ludwig-Maximilians-University, Marchioninistr. 15, 81377 Munich, Germany d Department of Neurology, Ludwig-Maximilians-University, Marchioninistr. 15, 81377 Munich, Germany b c
h i g h l i g h t s • • • •
GVS causes verticality deviations towards the anode during stimulation. After stimulation, the subjective verticals deviate towards the cathode. Aftereffects of GVS (1.5 mA, 20 min) exhibit different types of decay. GVS effects are best assessed with the subjective haptic vertical.
a r t i c l e
i n f o
Article history: Received 4 February 2014 Received in revised form 11 August 2014 Accepted 15 August 2014 Available online 23 August 2014 Keywords: Galvanic vestibular stimulation Subjective vertical Verticality perception
a b s t r a c t The human brain constructs verticality perception by integrating vestibular, somatosensory, and visual information. Here we investigated whether galvanic vestibular stimulation (GVS) has an effect on verticality perception both during and after application, by assessing the subjective verticals (visual, haptic and postural) in healthy subjects at those times. During stimulation the subjective visual vertical and the subjective haptic vertical shifted towards the anode, whereas this shift was reversed towards the cathode in all modalities once stimulation was turned off. Overall, the effects were strongest for the haptic modality. Additional investigation of the time course of GVS-induced changes in the haptic vertical revealed that anodal shifts persisted for the entire 20-min stimulation interval in the majority of subjects. Aftereffects exhibited different types of decay, with a preponderance for an exponential decay. The existence of such reverse effects after stimulation could have implications for GVS-based therapy. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Humans construct and update their sense of verticality by integrating vestibular, somatosensory, and visual input [18]. The internal estimate of verticality can be assessed by different methods, testing preferentially the visual, the tactile and the postural modalities (subjective visual, haptic, and postural vertical). These modalities can be differentially affected in patients with neurological disorders [4,23,33].
∗ Corresponding author at: Schoen Klinik Bad Aibling, Kolbermoorer Str. 72, 83043 Bad Aibling, Germany. Tel.: +49 8061 903 1903. E-mail addresses: [email protected] (K. Volkening), [email protected] (J. Bergmann), [email protected] (I. Keller), [email protected] (F. Müller), [email protected] (K. Jahn). 1 Joint first authors. http://dx.doi.org/10.1016/j.neulet.2014.08.028 0304-3940/© 2014 Elsevier Ireland Ltd. All rights reserved.
Transmastoidal galvanic vestibular stimulation (GVS) acts on afferents from the otoliths and the semicircular canals [8]. It was also shown to affect subjects’ perception of verticality. During stimulation the subjective visual (SVV) and the subjective haptic vertical (SHV) deviate towards the anode [21,22,24]. GVS causes eye torsion and nystagmus via the vestibular-ocular reflex [13] and – with the head upright – body tilt towards the anode via vestibulo-spinal reflexes [31]. However, existing studies on verticality perception have only examined the online effects of GVS, since judgments of verticality were always generated during stimulation intervals. In the oculomotor domain, GVS is known though to elicit reverse responses towards the cathode after stimulation [20,26]. It is not known if these aftereffects exist for the subjective verticals and whether there are any differences between modalities. This is of relevance since the time course and magnitude of effects and aftereffects of GVS on verticality perception might influence responses to
76
K. Volkening et al. / Neuroscience Letters 581 (2014) 75–79
therapeutic interventions. Thus, the purpose of this study was to examine the influence of GVS on different subjective verticals (visual, haptic, and postural) both during and after its application. 2. Materials and methods 2.1. Galvanic vestibular stimulation Bilateral bipolar GVS was delivered by a battery-driven, direct current stimulator (neuroConn Ilmenau, Germany). Electrodes were covered with natrium-chloride soaked sponges (30 cm2 each). Current was ramped up (in steps of 0.1 mA/s) to 1.5 mA and turned off at the end of the stimulation period. 2.2. Assessment of verticality perception 2.2.1. Subjective visual vertical (SVV) The SVV was assessed with the so-called bucket test. The bucket was rotated by the examiner, and the seated subjects indicated when they visually perceived a dark line (13 cm long, 0.3 cm wide, at 23 cm distance) as being vertical [35]. 2.2.2. Subjective haptic vertical (SHV) The SHV was measured with a rod (27 cm long, 1 cm wide) mounted onto a vertical plate 40 cm in front of the subject (see [10] for a similar device). While seated and blindfolded, the subjects’ task was to adjust the wooden rod, with their right hand using a precision grip until they perceived it to be in a vertical position. They were not allowed to touch the device’s plate or the desk. 2.2.3. Subjective postural vertical (SPV) The SPV was measured in the Spacecurl [2], a cardanic suspension apparatus that consists of three concentric rings. The blindfolded subject stood in the centre of the apparatus on a platform that was attached to the midmost ring. The device was tilted in the frontal plane, and subjects had to indicate when they felt they were in an upright position. There was no time limit for individual adjustments. Six adjustments per condition were performed in randomized order of starting positions (for SPV 12◦ , 15◦ & 18◦ ; for SVV and SHV 15◦ , 25◦ & 40◦ ). Half of the adjustments started from a clockwise, half from a counter-clockwise position. The six adjustments were averaged for each condition and modality to calculate the SPV, SVV, and SHV. Data were normalized so that positive values indicated deviations from the earth vertical to the side of the anode, and negative values, deviations in the direction of the cathode. 2.3. Experiments The Ethics Committee of the Ludwig-Maximilians-University Munich approved this study. All subjects provided their written informed consent. 2.3.1. Exp. 1: Manipulation of subjective verticals To investigate online effects and aftereffects of GVS on verticality perception across different modalities, ten healthy subjects participated in experiment 1 (mean age: 59 years, SD: ±6; 5 females). All subjects performed the SPV, the SVV, and the SHV immediately before (baseline), during, and 3 min after a period of GVS. The experiment was conducted on two consecutive days with a fixed sequential order: the SPV on day 1, the SVV and SHV on day 2. The polarity of the GVS current was varied between subjects. Stimulation was applied for the duration of verticality adjustments (4–8 min).
2.3.2. Exp. 2: Time course A second experiment was designed to determine the time course of online effects and aftereffects of GVS on verticality perception. Since the haptic modality was most responsive to GVS, the time course was studied for the SHV. Fourteen healthy subjects (mean age: 34 years, SD: ±6; 7 females), not included in experiment 1, were tested in experiment 2. The subjects repeatedly performed the SHV during and after a 20-min period of GVS. Six SHV adjustments were performed immediately before starting GVS (baseline), 0.5, 5, 10, 15, and 20 min after starting the stimulation (conditions 1–5) and at the same time points after terminating GVS (conditions 6–10) (total of 66 adjustments). All subjects were stimulated with the cathode over the left and the anode over the right mastoid.
2.4. Data analysis A one-way ANOVA with the within-subject factor modality was used to evaluate differences between modalities at baseline. To determine any differences in verticality adjustments across time points (baseline, during, after GVS), and modalities (SVV, SHV, SPV) a factorial repeated-measures ANOVA was performed with two within-subject factors (time and modality). Another factorial repeated-measures ANOVA (within-subjects factors modality and type of effect) was conducted to compare the magnitudes of online effects and aftereffects across modalities. Effect magnitudes were calculated as absolute differences between baseline and during GVS (online effect) and during and after GVS (aftereffect). If sphericity was violated in an ANOVA, Greenhouse–Geisser correction was applied. In case of significant results subsequent multiple comparisons were performed and Bonferroni corrected. For experiment 2 differences across conditions were analyzed using a one-way repeated-measures ANOVA and subsequent repeated contrasts. Verticality adjustments during (condition 1–5) and after GVS (condition 6–10) were grouped and compared using a paired t-test. A least-squares method was used to estimate which type of model (no decay, linear decay or exponential decay) best fits the individual time course of responses in verticality perception during and after GVS. A decay time constant was calculated for subjects showing an exponential decay. The data were analyzed with Matlab (The Mathworks, Version 2011b) and SPSS Statistics (SPSS Inc., Version 17.0). The significance level for ˛ was set at 0.05.
3. Results 3.1. Exp. 1: Manipulation of subjective verticals There was no significant difference between the modalities at baseline (F(2,27) = 0.056, p = 0.945; mean ± SE: SVV = 0.8 ± 0.5◦ , SHV = 0.6 ± 0.6◦ , SPV = 0.6 ± 0.4◦ ). The factorial repeated-measures ANOVA across all time points, showed a significant main effect of time (F(1.27,11.41) = 20.852, p < 0.001), but not of modality (F(2,18) = 2.082, p = 0.154). Post hoc tests revealed significant differences between the adjustments before and during GVS (p = 0.006) and during and after GVS (p < 0.001), but not between baseline and after GVS (p = 1.000). There was a significant interaction between time and modality (F(2.39,21.50) = 9.265, p = 0.001): Differences between adjustments before and during GVS were smaller for the SPV than for the SVV and SHV; adjustments during and after GVS were also smaller for the SPV than for the SHV. Analysis of the effect magnitudes showed a significant main effect of modality (F(2,18) = 17.260, p < 0.001), but no effect of effect type (online effect vs. aftereffect) (F(1,9) = 0.520, p = 0.489) and no significant interaction (F(1.16,10.42) = 1.506, p = 0.252). Post hoc tests revealed significantly greater online effects and aftereffects for the SHV than
K. Volkening et al. / Neuroscience Letters 581 (2014) 75–79
77
Fig. 1. Online effects and aftereffects of GVS. Adjustments (mean ± standard error, normalized to baseline and side of anode) of the subjective verticals during (SVV: 0.5 ± 0.3◦ , SHV: 2.8 ± 0.5◦ , SPV: −0.1 ± 0.2) and after (SVV: −0.7 ± 0.3◦ , SHV: −2.5 ± 0.4◦ , SPV: −0.5 ± 0.2◦ ) galvanic vestibular stimulation. Positive values indicate a tilt to the side of the anode, negative values a tilt to the side of the cathode. SVV, subjective visual vertical; SHV, subjective haptic vertical, SPV, subjective postural vertical; GVS, galvanic vestibular stimulation.
for the SVV (p = 0.009) and the SPV (p = 0.004). Normalized verticality adjustments during and after stimulation are shown in Fig. 1. 3.2. Exp. 2: Time course The average SHV at baseline was −0.6 ± 0.8◦ (mean ± SE). There were significant differences between conditions (F(10,130) = 2.241, p = 0.019). Contrasts revealed differences between baseline and condition 1 (F = 7.279, p = 0.018), condition 5 and 6 (F = 10.852, p = 0.005), and condition 9 and 10 (F = 8.779, p = 0.011) (Fig. 2). Accordingly, the SHV changed when turning stimulation on and off, and the latter effect had decayed after 20 min. Condition 1 and 2 (F = 4.539, p = 0.053), and condition 4 and 5 (F = 3.557, p = 0.082) tended to be different. The SHV adjustments during GVS (conditions 1–5) differed significantly from the adjustments after stimulation (conditions 6–10) (T(69) = 4.398, p < 0.001). As a majority of subjects (n = 8) showed no decay for the online effects, no decay time constant was calculated for the responses during GVS. Half of the subjects showed an exponential decay for the aftereffect with a mean decay time constant of 19.2 ± 1.6 min (for further details see supplementary table). 4. Discussion Our results provide the first direct evidence for the presence and time course of both online effects and aftereffects of GVS on the subjective verticals across different modalities. 4.1. Manipulation of subjective verticals During stimulation subjects’ SVV and SHV – but not SPV – shifted towards the anode. The shift was strongest for the haptic modality. After stimulation was switched off, this effect reversed across all three modalities, and slightly overshot the baseline value in the direction of the cathode for the SVV and the SPV (and also for the SHV in experiment 2). Anodal shifts during GVS are consistent with previous findings [21,24,31] showing that verticality perception can be influenced by manipulating the vestibular input. Former studies demonstrated the SVV to be more malleable by GVS than the SHV, but we found the opposite to be the case [21,31]. This might be due to differences in experimental protocols causing different degrees of ocular torsion. GVS-induced ocular torsion is known to influence the SVV [34], but also to decline over time [19]. While in previous experiments current was applied individually for each trial with SVV adjustments time-locked to current onset, we stimulated continuously across conditions. Thus, GVS-induced ocular torsion was
presumably reduced in our measurements compared to recordings at stimulus onset, resulting in less pronounced effects of GVS on the SVV. For the SHV ocular torsion appears to be rather negligible [6,30]. The SHV is known to be influenced by proprioceptive and tactile cues from the hand, arm and shoulder, as well as from the trunk [9] and the neck [12]. Findings on the role of the vestibular system are mixed. Whereas the SHV of patients with unilateral vestibular nuclear lesions was only marginally affected [6], a study with healthy controls indicated a major role of the vestibular system in SHV adjustments [25]. Our results support this crucial role of the vestibular system. In contrast to the SVV and SHV, we found no significant modulation of the SPV during GVS. Similarly, Bisdorff et al. [3] reported no effect of GVS on the SPV; however, they observed a broader range, within which subjects felt vertical during stimulation. One possible explanation for these findings might be that verticality is processed by distinct pathways in different modalities and thus possibly affected differently by GVS [15]. Another explanation for the limited effect of GVS on the SPV might be the abundance of somatosensory information. Somatosensory information – if available – is known to be crucial and even sufficient for an accurate SPV estimation [1,14,22]. In our set-up, different tactile information was provided by the Spacecurl’s padded restraints around hips and the pressure distribution under the feet. When sensory input from different sources is integrated, possible sensory conflict is resolved by weighting more reliable input stronger [7]. Body tilts generate somatosensory asymmetries that improve the reliability of somatosensory input compared to symmetric signals in upright position [5,22]. Consequently, the somatosensory input in our experiment was presumably considered more reliable for estimating the SPV than the conflicting vestibular input. In contrast to these online effects, we found the reverse effect after stimulation: in all modalities the subjective vertical shifted towards the cathode. Both, online effects and aftereffects were most pronounced for the haptic modality. Aftereffects of GVS have also been reported for the oculomotor domain and body movements [19,20,27,29]. When GVS is turned off, ocular torsion and nystagmus reverse direction and are directed towards the cathode. Studies investigating effects of GVS on body movement found a body sway towards the anode during stimulation [8,31] which reversed direction after stimulation. The magnitude of this aftereffect was significantly larger than the online effect [29]. Similarly, GVS-induced perceptions of rotation were also found to change direction after stimulation. Percepts during and after stimulation had similar peak magnitudes [27].
78
K. Volkening et al. / Neuroscience Letters 581 (2014) 75–79
Fig. 2. Time course of GVS effects. Time course of haptic verticality adjustments (mean ± standard error, normalized to baseline) during and after 20 min of galvanic vestibular stimulation. Times of assessment were 0.5, 5, 10, 15, and 20 min into stimulation (condition 1–5) and at the same time points after stimulation (condition 6–10). Positive values indicate a tilt to the side of the anode, negative values a tilt to the side of the cathode. BSLN, baseline; GVS, galvanic vestibular stimulation; SHV, subjective haptic vertical. Contrasts: *p < 0.05, **p < 0.01.
The influence of GVS on verticality perception was also investigated in right-brain damaged patients. For visuo-tactile adjustments, Saj et al. [24] reported stronger online effects in patients than in healthy subjects. The largest effects were found in patients with spatial neglect who typically show counter-clockwise SVV and SHV deviations [16]. Since these deviations shifted towards the anode during GVS, Saj et al. [24] proposed that right anodal stimulation could alleviate neglect patients’ spatial deficits. However, we found that online effects reversed after switching the stimulation off – with a slight overshoot towards the cathode. Since the time course of these aftereffects is the critical factor in terms of therapy, it was investigated in experiment 2. 4.2. Time course For the SHV, we found that GVS-induced shifts lasted not only during stimulation, but also up to at least 15 min after stimulation. When stimulation started, subjects’ SHV showed an anodal shift that was strongest immediately after the onset of GVS and persisted during the entire 20-min stimulation interval in the majority of subjects. When stimulation was switched off, the SHV reversed direction and shifted towards the cathode. For half of the subjects, this aftereffect showed an exponential decay with a decay time constant of 19.2 min. These findings are in line with studies on GVSinduced eye movements and movement perceptions. Reverse eye
movement responses were reported to last up to 6 min post a 5-min stimulation interval [19]. Response strength was linearly related to current intensity. Similarly, rotation perceptions during and after stimulation had comparable peak magnitudes and time courses of decay [27]. GVS-induced aftereffects appear to be directly related to the stimulation interval. Until now, however, this has not been explicitly confirmed for the subjective verticals. The longer-lasting reversed tilts of the subjective verticals that occurred after the application of GVS reflect adaptive mechanisms during prolonged stimulation. This may have implications for the use of GVS in the rehabilitation of distorted verticality perception. Based on earlier studies, GVS was proposed as a therapy for pusher behaviour and spatial neglect [17,24]. Evidence on its effectiveness, however, is lacking. The placement of the electrodes for therapy has so far been based on the effects observed during stimulation. Thus, neglect patients’ counter-clockwise verticality deviations were ameliorated when the anode was on the right and the cathode on the left mastoid [24]. Our finding that the perceived verticals deviate to the side of the cathode after stimulation suggests that the effect after stimulation should also be taken into account when using GVS as a therapeutic tool. In some cases it might be useful to reconsider the placement of the electrodes and setting of parameters. Our data showed a decay of the aftereffect in the majority of subjects, between-subject variability was however high. Thus further investigation of long-term
K. Volkening et al. / Neuroscience Letters 581 (2014) 75–79
aftereffects of GVS is needed, especially of the impact of repetitive stimulation. This pilot study has a few limitations. One is that we did not control the subjects’ trunk and head positions during verticality adjustments. Even though GVS is known to affect head and trunk orientation, its influence on verticality adjustments is controversial [11,22,28,31]. Another limitation is that the order of the different verticality assessments in experiment 1 was not randomized. The SVV and SHV adjustments were performed in a fixed sequential order on day 2. This might have biased the results, e.g. by adding up the stimulation effects, thus affecting the measures on day 2 more than those on day 1 (SPV). However, a recent study provides evidence against this, as repetitive GVS of up to 10 sessions was not more effective than a single session of GVS for improving neglect [32]. 5. Conclusions Our results provide the first direct evidence that GVS has both online effects and aftereffects on the subjective verticals. While online effects generally persisted during the entire 20-min stimulation interval, aftereffects showed different types of decay, with a preponderance for an exponential one. Reversed tilts of the subjective verticals after the termination of GVS might have implications for the rehabilitation of spatial orientation deficits. Further studies are needed to investigate such aftereffects over a longer timescale, after repetitive stimulation and in brain-damaged patients. Authors’ contribution KV and JB (joint first authors) designed the study, recruited the participants, collected the data, performed analyses and interpretation of data and wrote the article. IK, FM, JK obtained funding, contributed to the interpretation of the results and revised the article. MW performed part of the data analyses, wrote the corresponding text and contributed to the interpretation of the results. Acknowledgements The authors thank J. Benson for copyediting the manuscript. This work was supported by funds from the German Federal Ministry of Education and Research (BMBF IFB 01EO0901). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.neulet. 2014.08.028. References [1] J. Barra, A. Marquer, R. Joassin, C. Reymond, L. Metge, V. Chauvineau, D. Perennou, Humans use internal models to construct and update a sense of verticality, Brain 133 (2010) 3552–3563. [2] J. Bergmann, M.A. Ciupe, C. Krewer, A. Schepermann, F. Müller, K. Jahn, E. Koenig, Intrarater- und Interrater-Reliabilität des Spacecurls als Messinstrument der subjektiven posturalen Vertikalen, Neurol. Rehabil. 6 (2012) 417. [3] A.R. Bisdorff, C.J. Wolsley, D. Anastasopoulos, A.M. Bronstein, M.A. Gresty, The perception of body verticality (subjective postural vertical) in peripheral and central vestibular disorders, Brain 119 (Pt 5) (1996) 1523–1534. [4] I.V. Bonan, A. Marquer, S. Eskiizmirliler, A.P. Yelnik, P.P. Vidal, Sensory reweighting in controls and stroke patients, Clin. Neurophysiol. 124 (2012) 713–722. [5] A.M. Bronstein, M. Guerraz, Visual-vestibular control of posture and gait: physiological mechanisms and disorders, Curr. Opin. Neurol. 12 (1999) 5–11. [6] A.M. Bronstein, D.A. Perennou, M. Guerraz, D. Playford, P. Rudge, Dissociation of visual and haptic vertical in two patients with vestibular nuclear lesions, Neurology 61 (2003) 1260–1262.
79
[7] M.O. Ernst, H.H. Bulthoff, Merging the senses into a robust percept, Trends Cogn. Sci. 8 (2004) 162–169. [8] R.C. Fitzpatrick, B.L. Day, Probing the human vestibular system with galvanic stimulation, J. Appl. Physiol. 96 (2004) 2301–2316. [9] B. Fourre, B. Isableu, D. Bernardin, M. Gueguen, G. Giraudet, N. Vuillerme, C. Pagano, M.A. Amorim, The role of body centre of mass on haptic subjective vertical, Neurosci. Lett. 465 (2009) 230–234. [10] J. Funk, K. Finke, H.J. Muller, R. Preger, G. Kerkhoff, Systematic biases in the tactile perception of the subjective vertical in patients with unilateral neglect and the influence of upright vs. supine posture, Neuropsychologia 48 (2010) 298–308. [11] J. Funk, K. Finke, H.J. Muller, K.S. Utz, G. Kerkhoff, Effects of lateral head inclination on multimodal spatial orientation judgments in neglect: evidence for impaired spatial orientation constancy, Neuropsychologia 48 (2010) 1616–1627. [12] M. Guerraz, M. Luyat, D. Poquin, T. Ohlmann, The role of neck afferents in subjective orientation in the visual and tactile sensory modalities, Acta Otolaryngol. 120 (2000) 735–738. [13] K. Jahn, A. Naessl, M. Strupp, E. Schneider, T. Brandt, M. Dieterich, Torsional eye movement responses to monaural and binaural galvanic vestibular stimulation: side-to-side asymmetries, Ann. N. Y. Acad. Sci. 1004 (2003) 485–489. [14] R. Joassin, V. Bonniaud, J. Barra, A. Marquer, D. Perennou, Somaesthetic perception of the vertical in spinal cord injured patients: a clinical study, Ann. Phys. Rehabil. Med. 53 (2010) 568–574. [15] H.O. Karnath, S. Ferber, J. Dichgans, The origin of contraversive pushing: evidence for a second graviceptive system in humans, Neurology 55 (2000) 1298–1304. [16] G. Kerkhoff, Multimodal spatial orientation deficits in left-sided visual neglect, Neuropsychologia 37 (1999) 1387–1405. [17] C. Krewer, K. Riess, J. Bergmann, F. Muller, K. Jahn, E. Koenig, Immediate effectiveness of single-session therapeutic interventions in pusher behaviour, Gait Posture 37 (2013) 246–250. [18] C. Lopez, O. Blanke, The thalamocortical vestibular system in animals and humans, Brain Res. Rev. 67 (2011) 119–146. [19] H.G. MacDougall, A.E. Brizuela, A.M. Burgess, I.S. Curthoys, Between-subject variability and within-subject reliability of the human eye-movement response to bilateral galvanic (DC) vestibular stimulation, Exp. Brain Res. 144 (2002) 69–78. [20] H.G. MacDougall, A.E. Brizuela, I.S. Curthoys, Linearity, symmetry and additivity of the human eye-movement response to maintained unilateral and bilateral surface galvanic (DC) vestibular stimulation, Exp. Brain Res. 148 (2003) 166–175. [21] F. Mars, K. Popov, J.L. Vercher, Supramodal effects of galvanic vestibular stimulation on the subjective vertical, Neuroreport 12 (2001) 2991–2994. [22] F. Mars, J.L. Vercher, K. Popov, Dissociation between subjective vertical and subjective body orientation elicited by galvanic vestibular stimulation, Brain Res. Bull. 65 (2005) 77–86. [23] D. Perennou, C. Piscicelli, G. Barbieri, M. Jaeger, A. Marquer, J. Barra, Measuring verticality perception after stroke: why and how? Neurophysiol. Clin. 44 (2014) 25–32. [24] A. Saj, J. Honore, M. Rousseaux, Perception of the vertical in patients with right hemispheric lesion: effect of galvanic vestibular stimulation, Neuropsychologia 44 (2006) 1509–1512. [25] J.R. Schuler, C.J. Bockisch, D. Straumann, A.A. Tarnutzer, Precision and accuracy of the subjective haptic vertical in the roll plane, BMC Neurosci. 11 (2010) 83. [26] A. Severac Cauquil, M. Faldon, K. Popov, B.L. Day, A.M. Bronstein, Short-latency eye movements evoked by near-threshold galvanic vestibular stimulation, Exp. Brain Res. 148 (2003) 414–418. [27] R.J. St George, B.L. Day, R.C. Fitzpatrick, Adaptation of vestibular signals for self-motion perception, J. Physiol. 589 (2011) 843–853. [28] A.A. Tarnutzer, C.J. Bockisch, D. Straumann, Roll-dependent modulation of the subjective visual vertical: contributions of head- and trunk-based signals, J. Neurophysiol. 103 (2010) 934–941. [29] C.M. Tax, A.P. Bom, R.L. Taylor, N. Todd, K.K. Cho, R.C. Fitzpatrick, M.S. Welgampola, The galvanic whole-body sway response in health and disease, Clin. Neurophysiol. 124 (2013) 2036–2045. [30] S.W. Wade, I.S. Curthoys, The effect of ocular torsional position on perception of the roll-tilt of visual stimuli, Vision Res. 37 (1997) 1071–1078. [31] D.L. Wardman, J.L. Taylor, R.C. Fitzpatrick, Effects of galvanic vestibular stimulation on human posture and perception while standing, J. Physiol. 551 (2003) 1033–1042. [32] D. Wilkinson, O. Zubko, M. Sakel, S. Coulton, T. Higgins, P. Pullicino, Galvanic vestibular stimulation in hemi-spatial neglect, Front. Integr. Neurosci. 8 (2014) 4. [33] A.P. Yelnik, F.O. Lebreton, I.V. Bonan, F.M. Colle, F.A. Meurin, J.P. Guichard, E. Vicaut, Perception of verticality after recent cerebral hemispheric stroke, Stroke 33 (2002) 2247–2253. [34] R. Zink, S.F. Bucher, A. Weiss, T. Brandt, M. Dieterich, Effects of galvanic vestibular stimulation on otolithic and semicircular canal eye movements and perceived vertical, Electroencephalogr. Clin. Neurophysiol. 107 (1998) 200–205. [35] A. Zwergal, N. Rettinger, C. Frenzel, M. Dieterich, T. Brandt, M. Strupp, A bucket of static vestibular function, Neurology 72 (2009) 1689–1692.
