Comparison of verticality perception and postural sway induced by double temple-mastoidal and bipolar binaural 20 Hz sinusoidal galvanic vestibular stimulation

Abstract

BACKGROUND:

Galvanic vestibular stimulation (GVS) is believed to be one of the most valuable tools for studying the vestibular system. In our opinion, its combined effect on posture and perception needs to be examined more.

OBJECTIVE:

The present study was conducted to investigate the effect of a 20 Hz sinusoidal Galvanic Vestibular Stimulation (sGVS) on the body sway and subjective visual vertical (SVV) deviation through two sets of electrode montages (bipolar binaural and double temple-mastoidal stimulation) during a three-stage experiment (baseline, threshold, and supra-threshold levels).

METHODS:

While the individuals (32 normal individuals, 10 males, the mean age of 25.37±3.00 years) were standing on a posturography device and SVV goggles were put on, the parameters of the body sway and SVV deviation were measured simultaneously. Following the baseline stage (measuring without stimulation), the parameters were investigated during the threshold and supra-threshold stages (1 mA above the threshold) for 20 seconds. This was done separately for each electrode montage. Then, the results were compared between the three experimental stages and the two electrode montages.

RESULTS:

In both electrode montages, “the maximum amplitude” of the mediolateral (ML) and anteroposterior (AP) body sway decreased and increased in the threshold and supra-threshold stages, respectively, compared to the baseline stage. Comparison of the amount of “amplitude change” caused by each electrode montages showed that the double temple-mastoidal stimulation induced a significantly greater amplitude change in body sway during both threshold and supra-threshold stages (relative to the baseline stage).

The absolute mean values of the SVV deviation were significantly different between the baseline and supra-threshold levels in both electrode montages. The SVV deviation in double temple-mastoidal stimulation was a bit greater than that in the bipolar binaural stimulation.

CONCLUSION:

Double temple-mastoidal stimulation has induced greater amount of change in the body sway and SVV deviation. This may be due to the more effective stimulation of the otoliths than semicircular canals.

Keywords

Galvanic vestibular stimulation subjective visual vertical postural sway otolith

1 Introduction

The ability to stand upright depends on the sensory information received from the vestibular, visual, and proprioceptive systems [38]. The peripheral vestibular system consists of three parts, namely the semicircular canals, the otoliths, and the peripheral portion of the vestibular afferents. Functionally, it maintains postural balance through three primary mechanisms: vestibulo-ocular reflex, vestibulo-spinal reflex, and internal representation of gravity [2, 26]. Three semicircular canals sense angular head motion, which provides VOR to maintain gaze during head motion. Otoliths, utricle and saccule, respond to linear acceleration and static head tilts relative to gravity [2]. They act as gravity-inertial sensors that contribute to the verticality perception [1]. Furthermore, otoliths contribute to postural control through reticulospinal pathways [13].

In order to study the role of the vestibular system in postural control, a pure vestibular stimulus is required. Galvanic vestibular stimulation has become a popular tool for stimulating and improving the vestibular system function [25 , 51]. GVS is a low-intensity direct current (DC) electrical impulse that is transmitted to the vestibular afferents through electrodes placed over the mastoid bones [21]. GVS has different parameters, such as waveform, electrode montages, and frequency of the stimulus. According to the previous investigations, each parameter can be set differently to create various stimulus-induced reactions [18]. The following illustrates how changing these parameters affect the resulting responses.

The two most commonly used waveforms in previous studies were noisy and sinusoidal stimuli. Noisy galvanic vestibular stimulation showed positive impacts on postural control in both healthy subjects [31] and individuals with vestibular deficits [49]. These studies suggest that nGVS can be utilized as a rehabilitation tool for vestibular disorders. Non-noise waveforms, including sinusoidal, are employed in studies that aim to examine the function of the vestibular system [33]. Generally, these kinds of waveforms simulate vestibular deficits by worsening the postural control [31]. Sinusoidal GVS (sGVS) has some special properties that make it unique. In studies that require stimulation at a specific frequency, sinusoidal stimulation has been commonly used [37]. Furthermore, the dynamic properties of sinusoidal stimuli provide further efficiency in simulating natural head motion [28, 48].

Another critical parameter of GVS is different electrode montages which can trigger different behavioral responses [3]. The most commonly used electrode montage is a bipolar binaural stimulation; the cathode electrode is placed over one mastoid, and the anode electrode is placed over the other mastoid [21]. According to the Fitzpatrick model, behavioral responses evoked by bipolar binaural stimulation are often due to the semicircular canals rather than the otoliths [21]. But other studies revealed that delivering co-polar GVS (same polarity in the vestibules of both sides) can lead to better stimulation of otoliths [3, 10].

Frequency specificity is another imperative parameter of GVS stimulation. According to previous studies, this parameter might play a critical role in selective stimulation of otoliths and/ or semicircular canals. Previous studies have shown that 11–20 Hz GVS mainly stimulates the otolithic afferents and the reticulospinal pathways (reticular formations), which are essential inputs for postural control. However, low-frequency GVS (2–10 Hz) mainly activates semicircular canals and VOR pathways [14]. There are contradictory opinions regarding the selective stimulation of either the otoliths or the semicircular canals through GVS delivery. Various human and animal studies have demonstrated that both semicircular canals and otoliths are stimulated by GVS in the same way [12 , 17]. According to Annie Kwan et al., the transmastoid GVS elicits similar high pass tuning and neuronal detection thresholds in semicircular canals and otoliths afferents. Additionally, they demonstrated that afferents’ sensitivity to GVS depends more on their firing rate regularity than on which end organ they originate from [34]. Conversely, other studies postulated that although delivering GVS stimulates a large portion of the vestibular nerve, only the part related to the otoliths responds, and the part associated with the semicircular canals is inhibited [9].

Previous studies have shown that GVS can affect the subject’s verticality perception [23]. One way to assess the subjective verticality perception is to ask the subject to adjust a line in the vertical direction that they perceive (Subjective visual vertical or SVV test). SVV test is a psychophysical measurement of the angle between the line (subjective vertical perception) and the actual vertical direction, which examines the function of otoliths [5]. Based on the GVS waveform, bipolar binaural GVS could increase and/or decrease the SVV deviation, and it will be adjusted toward the anode side [42, 56]. Katharina Volkening et al. examined the effect of bipolar binaural GVS on SVV test results in normal individuals. They noticed that SVV deviated toward the anode side with a greater angle than the baseline (without GVS) [56].

GVS also affects the postural responses in terms of amplitude and direction of the body sway [56]. The Computerized Dynamic Posturography (CDP) is the most frequent way to examine the postural responses in terms of the amplitude and the direction of the body sway [36]. Previous studies have suggested that bipolar binaural GVS increases and/or decreases the amplitude of body sway (again based on the stimulus waveform) and induces body sway toward the anode side [36, 49]. MacDougall et al. (2012) investigated the effect of bipolar binaural GVS on the mediolateral and anteroposterior sway during the Sensory Organization Test (SOT). This experiment showed that body sway occurs in all the stages of SOT in both ML and AP directions, but AP sway becomes larger in stages with vision denied (stages 2 and 5) or distorted (steps 3 and 6) [36]. sGVS appears to have a remarkable effect on the amplitude of body sway based on its frequency and intensity. Previous studies have shown that during a high frequency and low-intensity sGVS, the body features acts as a low pass filter and limits the body sway [8, 13]. This effect is considered as the predominance of structural [52], neural [22 , 55], and biomechanical [8, 35] low pass filters over the stimulation effects.

To the best of our knowledge, the effect of a 20 Hz sGVS on SVV deviation and standing posture has not been investigated simultaneously. In addition, it seems that the impact of co-polar sGVS on SVV adjustment has not been studied so far. Therefore, we delivered a 20 Hz sGVS stimulation through two sets of electrode montages, including bipolar binaural and double temple-mastoidal stimulation. In this study, we investigated how our selected stimulus properties affected the body sway and SVV deviation in healthy individuals with standing posture. Previous studies have shown that more effective stimulation of otoliths may lead to different perceptual and motor responses [2 , 30]. We examined this hypothesis in healthy individuals by using SVV test, posturography device, and sGVS with two electrode montages.

To this end, we:

investigated the effect of bipolar binaural and double temple-mastoidal stimulation on body sway parameters (amplitude and direction).

determined the effect of two different electrode montages (bipolar binaural and double temple-mastoidal) on SVV deviation (amount and direction) in standing posture;

A three-stage procedure has been utilized to examine the body sway parameters and SVV deviations. Due to the different stimulation thresholds in individuals [32], we examined the stimulation threshold for each experimental subject after recording their baseline values (without stimulation). As a next step, to create the same level of stimulation in all participants, the stimulus was then delivered with an intensity of 1 mA above the threshold (supra-threshold). Finally, we compared the body sway and SVV deviation between the three stages to find out how different electrode montages affect the body sway and SVV deviation.

2 Materials and methods

2.1 Individuals

This study was performed on 32 normal individuals (10 males and 22 females; age range of 18 to 35 years, the mean age of 25.37±3.00 years) at Rofeideh Rehabilitation Hospital in Tehran, Iran. Their results of the video head impulse test (vHIT) and SVV test were in normal limits (i.e., VOR gains were more than 0.8 in all six planes and SVV deviations were less than±2 degrees). They had no negative history of neurological diseases, vestibular deficits, and/or hearing problems. All of the individuals gave written informed consent. The procedure was in accordance with the Helsinki Declaration and the Ethics Committee of the University of Social Welfare and Rehabilitation Sciences approved the experiment (IR.USWR.REC.1399.143).

2.2 Apparatus

2.2.1 Posturography device

The maximum amplitude of the body sway (the maximum deviations from the individual’s center of pressure in the AP and ML directions.) was examined during quiet stance on the SPS posturography force plate (Synapsys posturography system, Inventis Co., Italy) at a sampling rate of 100 Hz with 12-bit resolution. The force plate encompasses three sensors that continuously monitor the body sway as the center of pressure (COP). The amplitude and direction of the body sway were displayed on a COP stabilogram (kinetizieogram) separately in both AP and ML directions.

2.2.2 Subjective visual vertical

The SVV test was utilized to assess the subjective verticality perception. The instrument of this test included the SVV goggles and SVV-VII software (Synapsys, Inventis Co, Italy). These special goggles limit the visual field to 2 mm diameter holes to avoid environmental visual landmarks.

Through the SVV-VII software, the SVV luminous line was projected onto the front wall in a dark background with the desired angle. The angle difference between the luminous line and the actual vertical direction was determined with a resolution of 0.1 degrees. The rotation rate of the luminous line was 1 °/s. Moreover, the SVV test was performed in a standing posture on the posturography device while the body sway was also recorded (Fig. 1A).

Fig. 1

A Experimental setup. A 20 Hz sGVS was delivered through two electrode montages and its effects on body sway and SVV deviation was investigated. B. Electrode montages and their polarities in this study; bipolar binaural: 1. ARCL: anode right, cathode left. 2. ALCR: anode left, cathode right. Double temple-mastoidal: 3. AMCT: anode mastoids, cathode temples. 4. ATCM: anode temples, cathode mastoids.

2.2.3 Galvanic vestibular stimulation

GVS was delivered through the AES device ETT01 model, IEC60601-1-2, ISO13485 (Tanin Pardaz-e-Pasargad, Iran). Different types of stimulation waveforms can be delivered through this two-channel device within the frequency and intensity range up to 20 KHz and 5000μA with a resolution of 1 Hz and 1μA.

In this experiment, four areas of the skull (mastoids and temples on both sides) were selected, and their skin was cleaned with the Nuprep cleansing gel (Nuprep, Weaver and company the U.S.A). Those spots were dried and F-55 EKG electrodes (SKINTACT ECG electrodes, Austria) coated with electrode gel (Aqusuu ultrasound transmission gel, Royan Alchemists of Science and Tech, Iran) were attached to achieve a uniform current density and make sure of minimization of any irritation to the skin during the GVS presentation. GVS was delivered through two sets of electrode montages:

bipolar binaural stimulation;

double temple-mastoidal stimulation.

2.3 Grouping the individuals

The SVV deviation and direction of the body sway can be influenced by the two factors of stimulus polarity and the initial angle of the SVV target. Therefore, the experimental subjects were randomly assigned into different groups based on the GVS polarity and initial angle of SVV target.

2.3.1 Grouping based on the GVS polarity

Using each electrode montages mentioned above, GVS can be delivered through different polarities, which causes various effects on the body sway and SVV test. Therefore, to investigate the polarity effect, the individuals were grouped as follows:

Bipolar binaural stimulation consisted of two groups of sixteen individuals with a cathode or anode electrode on each mastoid:

ARCL: Anode Right, Cathode Left;

ALCR: Anode Left, Cathode Right.

Moreover, double temple-mastoidal stimulation comprised two groups of sixteen individuals with cathode or anode electrode on the mastoids and the opposite polarity on the temples:

AMCT: Anode Mastoids, Cathode temples;

ATCM: Anode temples, Cathode Mastoids.

2.3.2 Grouping based on the initial angle of SVV test target

The direction of the SVV deviation is affected not only by the stimulation polarity but also by the initial angle of the SVV test target. Therefore, we needed another grouping based on both the stimulation polarity and initial angle. The initial angle for different groups was considered as follows (Fig. 1B):

bipolar binaural stimulation with each group having sixteen individuals:

ARCL: –12 degrees

ALCR: +12 degrees

double temple-mastoidal stimulation with each group comprising eight individuals:

AMCT: –12 degrees

AMCT: +12 degrees

ATCM: –12 degrees

ATCM: +12 degrees

2.4 Task and procedure

Firstly, a preliminary group of five individuals was used to establish the stimulus parameters. In this part of the study, SVV was presented with three different initial angles of ±6, ±12, and ±18 degrees off the center. But because tolerating the supra-threshold intensity to examine several degrees of SVV was unpleasant to the individuals, only±12 degrees was considered as the SVV initial angle, which is greater than 2 standard deviations from the average value. Furthermore, with this angel, the SVV target’s arrival time to the center corresponds to the examination time window.

In this experiment, our semi-anechoic room was relatively dark. The individuals stood on the force plate of the posturography device with hands at their sides, facing forward, and feet 30 degrees apart from each other. SVV goggles were put on, and the luminous line was projected to the front wall (Fig. 1A). We asked the individuals to be completely comfortable and avoid causing contraction in their body or resistance against the body sways. For the SVV test, the luminous line was adjusted by the researcher through the SVV software, and the individuals were supposed to report verbally when the luminous line was in the vertical direction as they perceived.

There were three stages involved in this process:

The baseline: First, The body sway and SVV deviation were recorded for 20 seconds without stimulation as the baseline. The variables were recorded twice, and the statistics were calculated on the mean of two trials.

For the following two stages, if the impedance between the electrodes was confirmed to be less than 1 kΩ, sGVS at a frequency of 20 Hz was delivered to investigate the body sway and SVV deviation parameters through both electrode montages:

Threshold detection: We considered the threshold criterion in this study similar to that in previous studies [19, 50]. No stimulation was delivered in about the first 5 seconds to ensure that the body sway was established. Moreover, it is used as a reference for detecting any shift in stabilogram during sGVS presentation. sGVS was then delivered from an initial intensity of 200μA for 7 seconds. By delivering different intensities (50μA ascending step, 20μA descending step) and monitoring stabilogram, we attempted to find the lowest intensity level that would result in a constant shift from the first 5 seconds, which was initiated simultaneously with the stimulus and was constant throughout the entire time of the stimulus presentation. The AP and ML maximum amplitudes of body sway were altered by at least 1 mm relative to their baseline values at this stimulus intensity. After 30 seconds of rest, the intensity found in the first trial was repeated again for the second time. If it produced a shift of stabilogram with the same characteristics as the first trial, that intensity was considered to be the confirmed threshold (Fig. 2). Then for data analysis, the average amounts of the two trials were considered. This was done for each electrode montage separately.

Supra-threshold stimulation: At this stage, 20 Hz sGVS was presented for 20 seconds with an intensity of 1000μA above the threshold level for each electrode montages separately. During stimulus presentation, SVV deviation and body sway were measured simultaneously. The same procedure was repeated after 60 seconds of rest as the second trial, and the average values of the two trials were used for data analysis.

Fig. 2

AP- and ML-stabilograms during threshold detection. 1. Spontaneous body sway (before). 2. Body sway caused by sGVS (During sGVS). 3. sGVS after effects (after).

The three steps were performed according to the stated order for both electrode montages. At the supra-threshold level we were supposed to deliver stimulation with an intensity of 1 mA higher than the threshold. Therefore, the threshold had to be established first. The electrode montages, however, were arranged randomly. Some individuals have been tested first by bipolar binaural stimulation but others were investigated by double temple-mastoidal stimulation initially. A random list prepared with R software was used for randomization.

2.5 Measured parameters

2.5.1 The amplitude of the body sway

The maximum amount of body sway was measured in both AP and ML directions via the parameters of “AP maximum amplitude” and “ML maximum amplitude” (in millimeters). These two parameters contain positive numbers indicating the amount of body sway.

2.5.2 The direction of the body sway

“AP Mean” and “ML mean” parameters, which are positive or negative numbers, were utilized to determine the effect of sGVS on the direction of body sway. Positive numbers for the ML and AP mean indicate the body sway to the right and forward, respectively.

The body sways in both ML and AP directions whether delivering stimulation or not. We examined the correlation of the body sway between baseline and stimulation stages, including threshold and supra-threshold stages. It is explained in more details in the statistical analysis section.

2.5.3 The amount and direction of the SVV deviation

SVV angle showed a luminous line angle from its actual vertical direction with a positive or negative number. The number indicated the amount of luminous line deviation and the positive or negative sign stated the direction of its deviation. A positive number presented a deviation to the right side.

3 Statistical analysis

SPSS statistics version 22 (IBM, Armonk, NY, USA) was used for data analysis. P values less than 0.05 were considered significant. Initially, The Shapiro-Wilk test was used to check for the normality of the data.

The body sway data did not have a normal distribution (the Shapiro-Wilk test, p < 0.05). Therefore, the Friedman and Wilcoxon test were utilized. According to the normal distribution of the SVV data, statistical analysis was done using the repeated-measures ANOVA test.

3.1 Effects of sGVS on the amplitude of the body sway

The data are reported as median±IQR as a result of the Friedman test. To determine the effect of sGVS on the amplitude of body sway, we measured the AP and ML maximum amplitude of body sway in both stimulus intensity levels including threshold and supra-threshold, caused by each electrode montages. Then the results were compared with the corresponding values in the baseline level.

3.2 Effects of sGVS on the direction of the body sway

The direction of body sway is reported qualitatively based on AP and ML mean parameters in frequency and percentage of the individuals who had backward/forward sway.

3.3 Comparison of the ML and AP body sway caused by sGVS

To compare the effects of two electrode montages on the amount of body sway, the maximum amplitude of AP sway in the double temple-mastoidal stimulation was compared with the maximum amplitude of ML sway in bipolar binaural stimulation using the Wilcoxon test.

3.4 Correlation of the body sway

The data are reported as the correlation coefficient as a result of Spearman’s rho. This section would allow us to determine whether the effects of each electrode montage should be measured based on the body sway in ML or AP directions. To this purpose, a correlation was investigated between the body sway in baseline and stimulation stages (threshold and supra-threshold stages). For each electrode montage, we investigated the correlation of the ML body sway in every stimulation stage (including threshold and supra-threshold stages) to the ML sway in the baseline stage. A similar comparison was made for the AP sway (evaluating the correlation between the AP sway during the baseline and the stimulation stages, separately for each electrode montage).

The null hypothesis of a zero correlation is rejected with a significant level less than 0.05. The strength of the body sway relationship between the two stages is indicated with a correlation coefficient.

3.5 Effect of sGVS on the amount of SVV deviation

These data are reported as mean±SD (minimum-maximum). With within-subjects (condition), a repeated-measures ANOVA was performed to determine the effect of sGVS on the amount of SVV deviation. According to Funabashi et al., It was calculated as absolute differences between baseline and the two other stages [24]. Hence, the effect of polarity is considered the same among both electrode montages. A Greenhouse-Geisser correction was applied when sphericity was rejected. Multiple comparisons were performed using the Bonferroni method when significant results were found.

3.6 Effect of sGVS on the direction of SVV deviation

The direction of SVV deviation is reported qualitatively in the percent and frequency of individuals who adjusted the luminous line to right/left (based on the anode side).

4 Results

In the threshold stage, none of the individuals had any negative sensation of the sGVS but in the supra-threshold stage individuals had tingling, itching, knocking sensation under the electrode surface, illusion of movement, fear of falling, and Blinking light sensation. Just one of the individuals wanted to stop the stimulation. This individual was excluded from the experiment and only participants who completed the study are included in statistical analysis.

4.1 sGVS threshold

The averaged estimated threshold for both electrode montages is reported as Mean±SD (minimum-maximum).the sGVS threshold for bipolar binaural and double temple-mastoidal stimulation was 378.43μA±83.9 (230–550μA) and 282.5μA±71.52 (110–450μA), respectively.

4.2 Effects of sGVS on the amplitude of ML body sway

The median values (Q1, Q3) of the maximum amplitude of ML sway decreased at the threshold (bipolar binaural stimulation: 14.95 mm (11.28, 19.67), double temple-mastoidal stimulation: 15.95 mm (12.5, 19.23)) and increased at the supra-threshold level of both electrode montages (bipolar binaural stimulation: 18.42 mm (15.97, 23.15), double temple-mastoidal stimulation: 18.37 mm (15.05, 23.71)) in compare with its corresponding values of baseline stage (16.30 mm (13.01, 20.08)).

The maximum amplitude of ML sway was compared in pairs between different stages (Friedman test, p < 0.001, effect size: 0.31). It was found that there was a significant difference in the maximum amplitude of the ML sway between stages: bipolar binaural threshold vs. bipolar binaural supra-threshold (P < 0.001), bipolar binaural supra-threshold vs. baseline (P = 0.009), bipolar binauraul threshold vs. double temple-mastoidal supra-threshold (P = 0.002), and double temple-mastoidal threshold vs. bipolar binaural supra-threshold (P = 0.005). Star signs in Table 1 illustrate P-values of maximum amplitude of ML sway with a significant difference (Fig. 3).

Table 1
Descriptive results and pairwise comparison of Maximum amplitude of AP and ML body sway between different stages of both electrode montages (median±IQR, Friedman test, p < 0.05, effect size: 0.31)

Maximum amplitude of ML sway (mm) Maximum amplitude of AP sway (mm)

Electrode montage stages median IQR (q1,q3) median IQR (q1,q3)

Bipolar binaural stimulation (1) Baseline 16.30 13.01,20.08 22.67 16.71,27.1

threshold 14.95 11.28,19.67 18.7 13.97,23.75

pairwise comparison (baseline-threshold) P = 0.69 P = 1.000

baseline 16.30 13.01,20.08 22.67 16.71,27.1

supra-threshold 18.42 15.97,23.15 26.32 22.73,38.93

Pairwise comparison (baseline-supra-threshold) P = 0.009^* P = 0.02^*

Threshold 14.95 13.01,20.08 18.7 13.97,23.75

Supra-threshold 18.42 15.97,23.15 26.32 22.73,38.93

Pairwise comparison (threshold-Supra-threshold) P < 0.001^* P < 0.001^*

Double temple-mastoidal stimulation (2) Baseline 16.54 13.01,20.08 22.67 16.71,27.1

Threshold 15.95 12.5,19.23 21.12 14.36,28.33

pairwise comparison (baseline-threshold) P = 1.000 P = 1.000

baseline 16.54 13.01,20.08 22.67 16.71,27.1

supra-threshold 18.37 15.05,23.71 25.97 20.15,36.57

Pairwise comparison (baseline-supra-threshold) P = 0.32 P = 0.006^*

Threshold 15.95 12.5,19.23 21.12 14.36,28.33

Supra-threshold 18.37 15.05,23.71 25.97 20.15,36.57

Pairwise comparison (threshold-Supra-threshold) P = 0.21 P < 0.001^*

Electrode montage 1 Vs. electrode montage 2 Threshold 1 14.95 11.28,19.67 18.7 13.97,23.75

Threshold 2 15.95 12.5,19.23 21.12 14.36,28.33

Pairwise comparison (Threshold 1- Threshold 2) P = 0.96 P = 1.000

Supra-threshold 1 18.42 15.97,23.15 26.32 22.73,38.93

Supra-threshold 2 18.37 15.05,23.71 25.97 20.15,36.57

Pairwise comparison (Supra-threshold 1- Supra-threshold 2) P = 0.01 P = 1.000

Threshold 1 14.95 11.28,19.67 18.7 13.97,23.75

Supra-threshold 2 18.37 15.05,23.71 25.97 20.15,36.57

Pairwise comparison (Threshold 1- Supra-threshold 2) P < 0.002^* P < 0.001^*

Threshold 2 15.95 12.5,19.23 21.12 14.36,28.33

Supra-threshold 1 18.42 15.97,23.15 26.32 22.73,38.93

Pairwise comparison (Threshold 2- Supra-threshold 1) P < 0.005^* P = 0.001^*

		Maximum amplitude of ML sway (mm)	Maximum amplitude of AP sway (mm)
Bipolar binaural stimulation (1)	Baseline	16.30	13.01,20.08	22.67	16.71,27.1
	threshold	14.95	11.28,19.67	18.7	13.97,23.75
	pairwise comparison (baseline-threshold)	P = 0.69	P = 1.000
	baseline	16.30	13.01,20.08	22.67	16.71,27.1
	supra-threshold	18.42	15.97,23.15	26.32	22.73,38.93
	Pairwise comparison (baseline-supra-threshold)	P = 0.009^*	P = 0.02^*
	Threshold	14.95	13.01,20.08	18.7	13.97,23.75
	Supra-threshold	18.42	15.97,23.15	26.32	22.73,38.93
	Pairwise comparison (threshold-Supra-threshold)	P < 0.001^*	P < 0.001^*
Double temple-mastoidal stimulation (2)	Baseline	16.54	13.01,20.08	22.67	16.71,27.1
	Threshold	15.95	12.5,19.23	21.12	14.36,28.33
	pairwise comparison (baseline-threshold)	P = 1.000	P = 1.000
	baseline	16.54	13.01,20.08	22.67	16.71,27.1
	supra-threshold	18.37	15.05,23.71	25.97	20.15,36.57
	Pairwise comparison (baseline-supra-threshold)	P = 0.32	P = 0.006^*
	Threshold	15.95	12.5,19.23	21.12	14.36,28.33
	Supra-threshold	18.37	15.05,23.71	25.97	20.15,36.57
	Pairwise comparison (threshold-Supra-threshold)	P = 0.21	P < 0.001^*
Electrode montage 1 Vs. electrode montage 2	Threshold 1	14.95	11.28,19.67	18.7	13.97,23.75
	Threshold 2	15.95	12.5,19.23	21.12	14.36,28.33
	Pairwise comparison (Threshold 1- Threshold 2)	P = 0.96	P = 1.000
	Supra-threshold 1	18.42	15.97,23.15	26.32	22.73,38.93
	Supra-threshold 2	18.37	15.05,23.71	25.97	20.15,36.57
	Pairwise comparison (Supra-threshold 1- Supra-threshold 2)	P = 0.01	P = 1.000
	Threshold 1	14.95	11.28,19.67	18.7	13.97,23.75
	Supra-threshold 2	18.37	15.05,23.71	25.97	20.15,36.57
	Pairwise comparison (Threshold 1- Supra-threshold 2)	P < 0.002^*	P < 0.001^*
	Threshold 2	15.95	12.5,19.23	21.12	14.36,28.33
	Supra-threshold 1	18.42	15.97,23.15	26.32	22.73,38.93
	Pairwise comparison (Threshold 2- Supra-threshold 1)	P < 0.005^*	P = 0.001^*

Fig. 3

The individual values of ML maximum amplitude of body sway in all tested subjects in different stages of baseline, threshold and supra-threshold with each electrode montages (median±IQR). The box plots indicate that the median value decreases at the threshold stage of both electrode montages relative to its value in the baseline and increases at the supra-threshold stages. T: threshold. S.T: supra-threshold. Star sign illustrate a significant difference.

4.3 Effects of sGVS on direction of ML body sway

In bipolar binaural stimulation, the ML sway was towards the anode in most individuals of both ARCL and ALCR groups. Within the ARCL group, most of the individuals (P: 68%, F: 11) exhibited ML sway to the right (anode side) and other individuals (P: 32%, F: 5) to the left (cathode side). There was no difference in the percentage and frequency of ML sway in ALCR group except that because the anode electrode was on the right, most individuals exhibited sway to the right side (P: 68%, F: 5) (Fig. 5. a, b).

Fig. 5

The individual values of SVV deviation in all tested subjects in different stages of baseline and supra-threshold level with each electrode montages. The box plots indicate that the median value has increased at the supra-threshold stage of both electrode montages.

4.4 Effects of sGVS on amplitude of AP body sway

In contrast to its baseline value (22.67 mm (16.71, 27.1)), the AP sway decreased at the threshold level of both electrode montages (bipolar binaural stimulation: 18.7 mm (13.97, 23.75), double-temple mastoidal stimulation: 21.12 mm (14.36, 28.33)), and increased at the supra-threshold level (bipolar binaural stimulation: 26.32 mm (22.73, 38.93), double temple-mastoidal stimulation: 25.97 mm (20.15, 36.57)).

Pairwise comparison between the stages revealed a significant difference (Friedman test, p < 0.001, effect size: 0.31) between baseline vs. bipolar binaural supra-threshold (P = 0.02), bipolar binaural threshold vs. bipolar binaural supra-threshold (P < 0.001), baseline vs. double temple-mastoidal supra-threshold (P = 0.006), double temple-mastoidal threshold vs. double temple-mastoidal supra-threshold (P < 0.001), bipolar binaural threshold vs. double temple-mastoidal supra-threshold (P < 0.001) and double temple-mastoidal threshold vs. bipolar binaural supra-threshold (P = 0.001) and). Star signs in Table 1 illustrate P-values of maximum amplitude of AP sway with a significant difference (Fig. 4).

Fig. 4

The individual values of AP maximum amplitude in all tested subjects in different stages of baseline, threshold and supra-threshold with both electrode montages (median±IQR). The box plots indicate that the median value decreases at the threshold stage of both electrode montages relative to its value in the baseline and increases at the supra-threshold stages. T: threshold. S.T: supra-threshold. Star sign illustrate a significant difference.

4.5 Effects of sGVS on direction of AP body sway

Most individuals of both groups of double temple-mastoidal stimulation (AMCT and ATCM), showed predominantly backward AP sway. most individuals (P: 93%, F: 15) in both groups experienced backward sway, while 1 individual (7%) had forward sway; Thus AP sway in most individuals of AMCT group was to the anode side and in ATCM group was to the cathode side.

4.6 Comparison of the ML and AP body sway caused by sGVS

In all stages including baseline, a significant difference was observed (Wilcoxon test, p < 0.05, effect size: 0.36) between the ML sway in bipolar binaural stimulation and the AP sway in double temple-mastoidal stimulation.

Table 3
Correlation of the ML and AP body sway between three different stages of the experiment (Spearman’s rho, p < 0.05, correlation coefficient > 60)

threshold Supra-threshold

Base line Double temple-mastoidal Bipolar binaural Double temple-mastoidal Bipolar binaural

AP maximum amplitude correlation coefficient = 63 P < 0.01^* correlation coefficient = 43 P = 0.01 correlation coefficient = 70 P < 0.01^* correlation coefficient = 35 P = 0.04

ML maximum amplitude correlation coefficient = 15 P = 0.4 correlation coefficient = 84 P < 0.01^* correlation coefficient = 32 P = 0.07 correlation coefficient = 80 P < 0.01^*

	threshold	Supra-threshold
AP maximum amplitude	correlation coefficient = 63 P < 0.01^*	correlation coefficient = 43 P = 0.01	correlation coefficient = 70 P < 0.01^*	correlation coefficient = 35 P = 0.04
ML maximum amplitude	correlation coefficient = 15 P = 0.4	correlation coefficient = 84 P < 0.01^*	correlation coefficient = 32 P = 0.07	correlation coefficient = 80 P < 0.01^*

In order to eliminate the baseline impact, we calculated the amount of amplitude change for individuals by subtracting the maximum amplitude of body sway at threshold and supra-threshold stages from their value in the baseline. Then the median and quartiles were calculated which description is as follows (Table 2).

Table 2

Descriptive results (median±IQR) and pairwise comparison of the amplitude change of body sway due to GVS delivery at the threshold and supra-threshold intensity levels. Star sign illustrate a significant difference (Wilcoxson test, p < 0.05)

Condition	Electrode montages	Amplitude change
Threshold	Double temple-mastoidal	3.07 (1.71,12.34)
	Bipolar binaural	1.6 (0.82,3.88)
	pairwise comparison	P = 0.01^*
Supra-threshold	Double temple-mastoidal	5.6 (3.36,10.60)
	Bipolar binaural	3.35 (2.40,4.70)
	pairwise comparison	P < 0.01^*

There was significant difference between the amplitude change of ML body sway caused by bipolar binaural stimulation and the amplitude change of AP body sway caused by double temple-mastoidal stimulation in both threshold (P = 0.01) and supra threshod satges (P < 0.01). Double temple-mastoidal stimulation caused greater amount of amplitude change than bipolar binaural stimulation in both threshold (3.07(1.71, 12.34)) and supra-threshold (5.6(3.36, 10.60)) stages.

4.7 Correlation of body sway

The results showed that the ML sway in the baseline stage was significantly correlated with the ML sway in the threshold and supra-threshold stages caused by the bipolar binaural stimulation (P < 0.01) with correlation coefficient of 84% and 80%, respectively (correlation coefficient > 60%). The AP sway in the baseline stage has a significant correlation with the AP sway in the threshold and supra-threshold stages in the double temple-mastoidal stimulation (P < 0.01) with correlation coefficient of 69% and 70%, respectively (correlation coefficient > 60%).

Based on these results, the ML sway was used to discuss the effect of the bipolar binaural stimulation on the body sway while the AP sway was evaluated as a measure of the effects of the double temple-mastoidal stimulation.

4.8 Effect of sGVS on SVV deviation

With the 95% confidence intervals, the absolute average values of SVV deviation at base line was 1.58±0.99° (0.15–4.25). Due to the supra-threshold stimulation of both electrode montages, the SVV deviation increased to 2.42±1.11° (0.7–4.45) by double temple mastoidal stimulation and 2.30±1.5° (0.25–6.4) by bipolar binaural stimulation.

The data had a normal distribution (Shapiro-Wilk, p > 0.05). In repeated-measure ANOVA, Mauchly’s test of Sphericity was not significant (df: 2, p > 0.05). There was a significant difference in within-subject effects (Greenhouse-Geisser, p < 0.05). Post hoc tests revealed significant differences between absolute average value of SVV deviations in the baseline and supra-threshold stages of both electrode montages (Bonferroni, p < 0.17). However, the absolute average values of SVV deviation were not significantly different between two electrode montages (Table 4, Fig. 5).

Table 4
Descriptive results and pairwise comparison of SVV deviation during three different stages of baseline, threshold and supra-threshold intensity levels of both electrode montages (Repeated-measure ANOVA)

SVV deviation (degrees)

Electrode montages stages Mean SD

Bipolar binaural stimulation (1) Baseline 1.58 0.99

supra-threshold 2.30 1.53

Pairwise comparison(Baseline- supra-threshold) P < 0.01^*

Double temple-mastoidal stimulation (2) Baseline 1.58 0.99

supra-threshold 2.42 1.11

Pairwise comparison (Baseline- supra-threshold) P < 0.01^*

Electrode montage 1 Vs. electrode montage 2 supra-threshold 1 2.30 1.53

supra-threshold 2 2.42 1.11

Pairwise comparison P = 0.244

		SVV deviation (degrees)
Bipolar binaural stimulation (1)	Baseline	1.58	0.99
	supra-threshold	2.30	1.53
	Pairwise comparison(Baseline- supra-threshold)	P < 0.01^*
Double temple-mastoidal stimulation (2)	Baseline	1.58	0.99
	supra-threshold	2.42	1.11
	Pairwise comparison (Baseline- supra-threshold)	P < 0.01^*
Electrode montage 1 Vs. electrode montage 2	supra-threshold 1	2.30	1.53
	supra-threshold 2	2.42	1.11
	Pairwise comparison	P = 0.244

4.9 Effect of sGVS on direction of SVV deviation

With bipolar binaural stimulation, SVV was deviated toward the anode side in most individuals (P:68%, F:11) of both groups; Thus in ARCL group (anode right, cathode left), SVV was deviated to the right and in ALCR group (anode left, cathode right) it was deviated to the left. In all test groups receiving double temple-mastoidal stimulationt, SVV was adjusted to the side of the initial angle; most of the individuals in the ATCM group deviated the SVV to the left when the initial angle was –12 degrees (P: 75%, F: 6), while all of them deviated the SVV to the right when the initial angle was +12 degrees (P: 100%, F: 8). In the AMCT group, most of the individuals adjusted the SVV to the left when the initial angle was –12 degrees (P: 88%, F: 7), while most of them adjusted it to the right when the initial angle was 12 degrees (P: 75%, F: 6).

5 Discussion

The current study compared the body sway and SVV deviation caused by a 20 Hz sGVS between two electrode montages (bipolar binaural and double temple-mastoidal stimulation) over three experimental stages (the baseline, threshold, and supra-threshold intensity levels) in normal individuals. The results revealed that in both electrode montages, the body sway and SVV deviation were affected in terms of amplitude and direction. The double temple-mastoidal stimulation was found to emerge a more significant amount of body sway compared with the bipolar binaural stimulation in both threshold and supra-threshold stages. The findings will be discussed in more detail below.

5.1 Effect of sGVS on body sway

Previous investigations showed that the body sway occurs in both ML and AP directions, both spontaneously and in response to GVS stimulation [29, 31]. Various methods have been employed to investigate the relationship between the spontaneous and stimulus-induced body sway. Day et al. considered the vector summation of the ML and AP body sway. They found that for both bipolar binaural and co-polar stimulation the vector summation of body sway happens in the same direction of the stimulation plane [14]. Scinicariello et al. demonstrated that co-polar GVS results in coherent AP body sway [45]. Furthermore, Pavlik et al. showed that bipolar binaural stimulation causes a coherent ML body sway [37]. The present study also demonstrated that sGVS caused body sway in both directions. Based on the previous studies, a correlation has been investigated to examine the relationship between the body sway in the baseline and stimulation stages. As the result, the ML sway at the baseline stage correlated with ML sway at the threshold and supra-threshold stages during bipolar binaural stimulation. The AP sway in the baseline stage was correlated with the AP sway caused by the threshold and supra-threshold stages of double temple-mastoidal stimulation. Therefore, the ML and AP sways were utilized to discuss the effect of sGVS during the bipolar binaural and double temple-mastoidal stimulation, respectively.

5.1.1 Effect of sGVS on the amplitude of body sway

The current study showed that the body sway decreased during the threshold stage of both electrode montages, though not significantly, but increased significantly at the supra-threshold level compared to the baseline values. Previous studies have shown that structural [52], biomechanical [8, 35], and neural [22 , 55] filters can limit the body sways, although we did not examine it directly in our study.

5.1.2 Effect of sGVS on the direction of ML body sway

According to Day et al., during GVS, the body sway occurs in both AP and ML directions, and as a result of their vector summation, the body sways toward the anode electrode [16]. Our findings on the effect of bipolar binaural stimulation on ML body sway are consistent with the results of this study. In the present study, the ML body sway was towards the anode side in most individuals in both groups of bipolar binaural stimulation.

5.1.3 Effect of sGVS on the direction of AP body sway

According to Day et al., We expect the AP body sway to be towards the anode electrode during co-polar stimulation of mastoids [14]. However in the current study, during double temple-mastoidal stimulation, AP sway was always backward even if mastoids were stimulated with cathode electrodes.

Standing posture is achieved primarily through sensory inputs of the visual, vestibular, and proprioceptive systems. However, when the body is exposed to multiple environmental circumstances, maintaining a stable stance gets more complicated because these feedback mechanisms may not always be accurate or available [44]. In a standing individual, sensory inputs are delivered in two pathways during GVS stimulation; top-down and bottom-up sensory pathways. Top-down pathways transmit sensory inputs from the vestibular system to the lower limbs. In addition, sensory information received from the legs is also transmitted through the bottom-up pathway [39]. Lower limb responses appear to be influenced by non-vestibular inputs [6] and may outweigh the vestibular afferent changes caused by vestibular sensory inputs [20]. Wardman et al. showed that GVS-induced body sways in the standing posture can be influenced by the sensory information from the legs [57]. In the current study, the AP sway was in the backwards direction during the baseline stage for most individuals. In addition, during the supra-threshold stage, more individuals showed backward AP sway. This might indicate the effects of the legs’ sensory inputs on the direction of the AP sway. In other words, in the baseline stage, the legs’ sensory inputs were somehow that backward body sway predominated over forward body sway. Furthermore, in the supra-threshold stage, the visual and vertical perceptual inputs have been challenged by SVV presentation. The conflict between the sensory signals may have increased the number of individuals with backward body sway compared to the baseline stage. This statement needs to be more evaluated in future research.

Fear of falling might be another influential factor. A decrease in balance confidence may lead to fear of falling, which can indirectly result in a decline in the postural control strategy. Mark G. Carpenter et al. showed that as we stand in a high threat condition, fear of falling leads to a backward shift of the mean AP position of COP. However, no significant changes were observed in the ML direction because hip strategy and stance width regulate this change [7]. Although the current study did not intend to examine the effect of fear of falling, participants in the supra-threshold stage of double temple-mastoidal stimulation reported an illusion of movement, disequilibrium, and fear of falling. Consequently, fear of falling may be an influential factor that led to increasing backward sway during the supra-threshold stage.

5.2 Effects of sGVS on SVV deviation

5.2.1 Effect of sGVS on the amount of SVV deviation

Based on the results of our experiment, SVV deviation significantly increased in both electrode montages compared to the baseline. In consistent with previous studies, bipolar binaural stimulation increased the amount of SVV deviation [53, 56]. To the best of our knowledge, this is the first study to investigate the effect of co-polar stimulation on SVV deviation.

5.2.2 Effect of sGVS on the direction of SVV deviation

In bipolar binaural stimulation, the SVV was adjusted to the anode side and in double temple-mastoidal stimulation, the deviation was observed toward the initial angle. There are neither compatible nor incompatible studies for co-polar stimulation, but the results for bipolar binaural stimulation are consistent with those of previous studies; with the bipolar binaural stimulation, SVV deviated toward the anode side [53, 56].

Two factors were found to affect the SVV deviation during GVS stimulation: the stimulation polarity and the initial angle of luminous line. Bipolar binaural GVS simulates the push-pull mechanism [11]. This causes an asymmetrical change in the vestibular nerve function (increasing the firing rate on the cathode side and decreasing it on the anode side), similar to that obtained by tilting a person’s head either left or right. Previous studies have shown that asymmetrical stimulation of the vestibular systems, whether caused by natural head tilt or GVS stimulation, affects the verticality perceptions and the results of SVV tests [27, 33]. Therefore, it appears that during the bipolar binaural stimulation of the current study, the polarity was the influential factor on SVV deviation. Meanwhile, with co-polar stimulation, mastoids of both sides are stimulated with the similar polarity; primary vestibular afferents of both sides are stimulated or inhibited simultaneously. Hence, there is no asymmetry in the vestibular nerve function of both sides [9]. In this state, the vestibular nerve have the same polarity as when the head has been bent forward or backward [2]. Hence, with the double-temple mastoidal stimulation, the vestibular systems of both sides had symmetrical function. Thus, it seems unlikely that during double-temple mastoidal stimulation, the polarity effect was the influential factor on the SVV deviation.

Another factor affecting the SVV deviation is the initial angle of luminous line. Pagarkar et al. showed that the SVV tilt was biased in the direction of the luminous line. It is hypothesized that it results from an interaction between the central vestibular and visual systems [43].

In this study, during double-temple mastoidal stimulation, the SVV (visual stimulation) was presented on the frontal plane, while sGVS (vestibular stimulation) was presented on the sagittal plane. There seems to be a difference in the visual and vestibular stimulation plane that caused the initial angle effect to overcome the polarity effect. If the SVV was also presented on the sagittal plane, it would probably deviate in the backward and forward directions with the anodal and cathodal stimulation, respectively. This effect is consistent with the expected polarity effects. However, presenting SVV in the sagittal plane is complicated. While the line is presented on the frontal plane, the direction of the luminous line is perceived directly by processing retinal projections. Nevertheless, processing the orientation in the sagittal plane requires additional information about the depth. Since luminous line orientation is perceived through variations in its length or width, the nearer extremity appears larger optically, and the linear perspective is influenced by inclination [4, 47].

5.3 Double temple mastoidal stimulation: A more effective stimulation for otolith afferents

Despite the limitations of non-invasive GVS [46], previous studies have shown that co-polar stimulation may be more effective than other polarities in stimulating the otoliths [3]. In this study, we provided the best possible conditions for more effective stimulation of otolith afferents, including delivering sinusoidal GVS [28] with a frequency of otoliths function [14] in anterior-posterior direction between the temple and the mastoid of each side [3].

Previous studies have shown that more effective stimulation of otoliths may lead to greater perceptual and motor responses in the anteroposterior direction [2 , 30]. Aoyama et al., hypothesized that during co-polar GVS, the perception and actual head movement may be due to more effective stimulation of the otoliths [2]. Moreover, Magnussen et al. speculate that the neural responses from the semicircular canals may be assumed to neutralize each other during the co-polar GVS stimulation. In this case, the AP body sway is due to the otoliths rather than semicircular canals [30]. This consideration is in line with Curthoys et al. study [8]. In the current study, the double-temporal mastoidal stimulation emanated a greater amount of AP sway amplitude change at both the threshold and supra-threshold stages that were significantly different from the ML sway amplitude change caused by bipolar binaural stimulation. In line with the previous studies, this finding may indicate that double temple-mastoidal stimulation is probably more effective in stimulating otoliths. Furthermore, double temple-mastoidal stimulation led to a bit greater amount of SVV deviation. Taking all this together, these results indicate a possibility for more effective stimulation of otoliths in response to double temple-mastoidal stimulation.

6 Conclusion

The results showed that in both electrode montages, the amplitude and direction of the body sway and SVV deviation were affected. Based on these results, it seems that otoliths were stimulated more effectively with double temple-mastoidal stimulation. However, further studies are needed to determine a definitive conclusion regarding selective stimulation of otoliths.

Footnotes

Acknowledgments

The authors are grateful to the participants in this research and appreciate the support of the Clinical Research Development Center of Rofeideh Rehabilitation Hospital, Tehran, Iran.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.

Conflicts of interest

The authors declare that they have no conflict of interest.

Ethics approval

The procedure was in accordance with the Helsinki Declaration (2013) and the Ethics Committee of University of social welfare and rehabilitation sciences approved the experiment (IR.USWR.REC.1399.143).

Consent to participate

All of the subjects gave written informed consent.

Authors’ contributions

All the authors played a significant role in this study.

References

Akin

F.W.

and Murnane

O.D.

. Subjective visual vertical test, in: Seminars in Hearing, ©Thieme Medical Publishers, 2009, pp. 281–286.

Angelaki

D.E.

and Cullen

K.E.

, Vestibular system: the many facets of a multimodal sense, Annu Rev Neurosci 31 (2008), 125–150.

Aoyama

, Iizuka

, Ando

and Maeda

, Four-pole galvanic vestibular stimulation causes body sway about three axes, Scientific Reports 5 (2015), 1–8.

Beck

and Gibson

J.J.

, The relation of apparent shape to apparent slant in the perception of objects, Journal of Experimental Psychology 50 (1955), 125.

Böhmer

and Mast

, Assessing otolith function by the subjective visual vertical, ANNALS-NEW YORK ACADEMY OF SCIENCES 871 (1999), 221–231.

Britton

, Day

, Brown

, Rothwell

, Thompson

and Marsden

, Postural electromyographic responses in the arm and leg following galvanic vestibular stimulation in man, Experimental Brain Research 94 (1993), 143–151.

Carpenter

M.G.

, Frank

J.S.

, Silcher

C.P.

and Peysar

G.W.

, The influence of postural threat on the control of upright stance, Experimental Brain Research 138 (2001), 210–218.

Coats

, The sinusoidal galvanic body-sway response, Acta oto-laryngologica 74 (1972), 155–162.

Cohen

, Yakushin

S.B.

and Holstein

G.R.

, What does galvanic vestibular stimulation actually activate? Frontiers in Neurology 2 (2012), 90.

10.

Curthoys

, Eye movements produced by utricular and saccular stimulation,A, Aviation, Space, and Environmental Medicine 58 (1987), 192–197.

11.

Curthoys

I.S.

, Concepts and physiological aspects of the otolith organ in relation to electrical stimulation, Audiology and Neurotology 25 (2020), 25–34.

12.

Curthoys

I.S.

and MacDougall

H.G.

, What galvanic vestibular stimulation actually activates, Frontiers in Neurology 3 (2012), 117.

13.

Dakin

C.J.

, Luu

B.L.

, van den Doel

, Inglis

J.T.

and Blouin

J.-S.

, Frequency-specific modulation of vestibular-evoked sway responses in humans, Journal of Neurophysiology 103 (2010), 1048–1056.

14.

Dakin

C.J.

, Son

G.M.L.

, Inglis

J.T.

and Blouin

J.S.

, Frequency response of human vestibular reflexes characterized by stochastic stimuli, The Journal of Physiology 583, (2007) 1117–1127.

15.

Day

B.L.

and Fitzpatrick

R.C.

, Virtual head rotation reveals a process of route reconstruction from human vestibular signals, The Journal of Physiology 567 (2005), 591–597.

16.

Day

B.L.

, Marsden

J.F.

, Ramsay

, Mian

O.S.

and Fitzpatrick

R.C.

, Non-linear vector summation of left and right vestibular signals for human balance, The Journal of Physiology 588(2010), 671–682.

17.

Day

B.L.

, Ramsay

, Welgampola

M.S.

and Fitzpatrick

R.C.

, The human semicircular canal model of galvanic vestibular stimulation, Experimental Brain Research 210 (2011), 561–568.

18.

Dlugaiczyk

, Gensberger

K.D.

and Straka

, Galvanic vestibular stimulation: from basic concepts to clinical applications, Journal of Neurophysiology 121 (2019), 2237–2255.

19.

Eikema

D.J.

, Hatzitaki

, Tzovaras

and Papaxanthis

, Application of intermittent galvanic vestibular stimulation reveals age-related constraints in the multisensory reweighting of posture, Neuroscience Letters 561 (2014), 112–117.

20.

Fitzpatrick

and McCloskey

, Proprioceptive, visual and vestibular thresholds for the perception of sway during standing in humans, The Journal of Physiology 478 (1994), 173–186.

21.

Fitzpatrick

R.C.

and Day

B.L.

, Probing the human vestibular system with galvanic stimulation, Journal of Applied Physiology 96 (2004), 2301–2316.

22.

Forbes

P.A.

, Dakin

C.J.

, Vardy

A.N.

, Happee

, Siegmund

G.P.

, Schouten

A.C.

and Blouin

J.-S.

, Frequency response of vestibular reflexes in neck, back, and lower limb muscles, Journal of Neurophysiology 110 (2013), 1869–1881.

23.

Fujimoto

, Egami

, Kawahara

, Uemura

, Yamamoto

, Yamasoba

and Iwasaki

, Noisy galvanic vestibular stimulation sustainably improves posture in bilateral vestibulopathy, Frontiers in Neurology 9 (2018), 900.

24.

Funabashi

, Santos-Pontelli

T.E.G.

, Colafêmina

J.F.

, Pavan

T.Z.

, Carneiro

A.A.O.

and Takayanagui

O.M.

, A newmethod to analyze the subjective visual vertical in patients with bilateral vestibular dysfunction, Clinics 67 (2012), 1127–1131.

25.

Ghahraman

M.A.

, Zahmatkesh

, Pourbakht

, Seifi

, Jalaie

, Adeli

and Niknami

, Noisy galvanic vestibular stimulation enhances spatial memory in cognitive impairment-induced by intracerebroventricular-streptozotocin administration, Physiology & Behavior 157 (2016), 217–224.

26.

Gibson

J.J.

and Mowrer

, Determinants of the perceived vertical and horizontal, Psychological Review 45 (1938), 300.

27.

Gnanasegaram

J.J.

, Parkes

W.J.

, Cushing

S.L.

, McKnight

C.L.

, Papsin

B.C.

and Gordon

K.A.

, Stimulation from cochlear implant electrodes assists with recovery from asymmetric perceptual tilt: evidence from the subjective visual vertical test, Frontiers in Integrative Neuroscience 10 (2016), 32.

28.

Goldberg

J.M.

and Cullen

K.E.

, Vestibular control of the head: possible functions of the vestibulocollic reflex, Experimental Brain Research 210 (2011), 331–345.

29.

Grillner

, Hongo

and Lund

, The vestibulospinal tract. Effects on alpha-motoneurones in the lumbosacral spinal cord in the cat, Experimental Brain Research 10 (1970), 94–120.

30.

Iles

and Pisini

, Vestibular-evoked postural reactions in man and modulation of transmission in spinal reflex pathways, The Journal of Physiology 455 (1992), 407–424.

31.

Iwasaki

, Yamamoto

, Togo

, Kinoshita

, Yoshifuji

, Fujimoto

and Yamasoba

, Noisy vestibular stimulation improves body balance in bilateral vestibulopathy, Neurology 82 (2014), 969–975.

32.

Keywan

, Wuehr

, Pradhan

and Jahn

, Noisy galvanic stimulation improves roll-tilt vestibular perception in healthy subjects, Frontiers in Neurology 9 (2018), 83.

33.

Kim

K.-S.

, Minor

L.B.

and Della

C.C.

, Santina and D.M. Lasker, Variation in response dynamics of regular and irregular vestibular-nerve afferents during sinusoidal head rotations and currents in the chinchilla, Experimental Brain Research 210 (2011), 643–649.

34.

Kwan

, Forbes

P.A.

, Mitchell

D.E.

, Blouin

J.-S.

and Cullen

K.E.

, Neural substrates, dynamics and thresholds of galvanic vestibular stimulation in the behaving primate, Nature Communications 10 (2019), 1–15.

35.

Latt

, Sparto

, Furman

and Redfern

, The steady-state postural response to continuous sinusoidal galvanic vestibular stimulation, Gait & Posture 18 (2003), 64–72.

36.

MacDougall

H.G.

, Moore

S.T.

, Curthoys

I.S.

and Black

F.O.

, Modeling postural instability with Galvanic vestibular stimulation, Experimental Brain Research 172 (2006), 208–220.

37.

Mackenzie

S.W.

and Reynolds

R.F.

, Ocular torsion responses to sinusoidal electrical vestibular stimulation, Journal of Neuroscience Methods 294 (2018), 116–121.

38.

Magnusson

, Johansson

and Wiklund

, Galvanically induced body sway in the anterior-posterior plane, Acta Oto-laryngologica 110 (1990), 11–17.

39.

Mergner

and Rosemeier

, Interaction of vestibular, somatosensory and visual signals for postural control and motion perception under terrestrial and microgravity conditions—a conceptual model, Brain Research Reviews 28 (1998), 118–135.

40.

Nguyen

T.T.

, Nam

G.-S.

, Kang

J.-J.

, Han

G.C.

, Kim

J.-S.

, Dieterich

and Oh

S.-Y.

, Galvanic vestibular stimulation improves spatial cognition after unilateral labyrinthectomy in mice, Frontiers in Neurology (2021), 1261.

41.

Njiokiktjien

and Folkerts

, Displacement of the body’s centre of gravity at galvanic stimulation of the labyrinth, Stereotactic and Functional Neurosurgery 33 (1971), 46–54.

42.

Oppenländer

, Utz

K.S.

, Reinhart

, Keller

, Kerkhoff

and Schaadt

A.-K.

, Subliminal galvanic-vestibular stimulation recalibrates the distorted visual and tactile subjective vertical in right-sided stroke, Neuropsychologia 74 (2015), 178–183.

43.

Pagarkar

, Bamiou

D.-E.

, Ridout

and Luxon

L.M.

, Subjective visual vertical and horizontal: effect of the preset angle, Archives of Otolaryngology-Head & Neck Surgery 134 (2008), 394–401.

44.

Peterka

R.J.

, Sensorimotor integration in human postural control, Journal of Neurophysiology 88 (2002), 1097–1118.

45.

Peterson

, Fukushima

, Hirai

, Schor

and Wilson

, Responses of vestibulospinal and reticulospinal neurons to sinusoidal vestibular stimulation, Journal of Neurophysiology 43 (1980), 1236–1250.

46.

Rosengren

S.M.

, Jombik

, Halmagyi

G.M.

and Colebatch

J.G.

, Galvanic ocular vestibular evoked myogenic potentials provide new insight into vestibulo-ocular reflexes and unilateral vestibular loss, Clinical Neurophysiology 120 (2009), 569–580.

47.

Saj

, Honoré

, Bernati

, Coello

and Rousseaux

, Subjective visual vertical in pitch and roll in right hemispheric stroke, Stroke 36 (2005), 588–591.

48.

Schneider

, Bartl

and Glasauer

, Galvanic vestibular stimulation combines with Earth-horizontal rotation in roll to induce the illusion of translation, Annals of the New York Academy of Sciences 1164 (2009), 116–118.

49.

Schniepp

, Boerner

, Decker

, Jahn

, Brandt

and Wuehr

, Noisy vestibular stimulation improves vestibulospinal function in patients with bilateral vestibulopathy, Journal of Neurology 265 (2018), 57–62.

50.

Scinicariello

A.P.

, Inglis

J.T.

and Collins

, The effects of stochastic monopolar galvanic vestibular stimulation on human postural sway, Journal of Vestibular Research 12 (2002), 77–85.

51.

Shaabani

, Lotfi

, Karimian

S.M.

, Rahgozar

and Hooshmandi

, Short-term galvanic vestibular stimulation promotes functional recovery and neurogenesis in unilaterally labyrinthectomized rats, Brain Research 1648 (2016), 152–162.

52.

Srinivasan

, Tucker

D.M.

and Murias

, Estimating the spatial Nyquist of the human EEG, Behavior Research Methods, Instruments, & Computers 30 (1998), 8–19.

53.

Tardy-Gervet

M.-F.

and Séverac-Cauquil

, Effects of galvanic vestibular stimulation on perception of subjective vertical in standing humans, Perceptual and Motor Skills 86 (1998), 1155–1161.

54.

Thomas

K.E.

and Bent

L.R.

, Subthreshold vestibular reflex effects in seated humans can contribute to soleus activation when combined with cutaneous inputs, Motor Control 17 (2013), 62–74.

55.

Uno

, Idoux

, Beraneck

, Vidal

P.-P.

, Moore

L.E.

, Wilson

V.J.

and Vibert

, Static and dynamic membrane properties of lateral vestibular nucleus neurons in guinea pig brain stem slices, Journal of Neurophysiology 90 (2003), 1689–1703.

56.

Volkening

, Bergmann

, Keller

, Wuehr

, Müller

and Jahn

, Verticality perception during and after galvanic vestibular stimulation, Neuroscience Letters 581 (2014), 75–79.

57.

Wardman

D.L.

, Taylor

J.L.

and Fitzpatrick

R.C.

, Effects of galvanic vestibular stimulation on human posture and perception while standing, The Journal of Physiology 551 (2003), 1033–1042.

		Maximum amplitude of ML sway (mm)		Maximum amplitude of AP sway (mm)
Electrode montage	stages	median	IQR (q1,q3)	median	IQR (q1,q3)
Bipolar binaural stimulation (1)	Baseline	16.30	13.01,20.08	22.67	16.71,27.1
	threshold	14.95	11.28,19.67	18.7	13.97,23.75
	pairwise comparison (baseline-threshold)	P = 0.69		P = 1.000
	baseline	16.30	13.01,20.08	22.67	16.71,27.1
	supra-threshold	18.42	15.97,23.15	26.32	22.73,38.93
	Pairwise comparison (baseline-supra-threshold)	P = 0.009^*		P = 0.02^*
	Threshold	14.95	13.01,20.08	18.7	13.97,23.75
	Supra-threshold	18.42	15.97,23.15	26.32	22.73,38.93
	Pairwise comparison (threshold-Supra-threshold)	P < 0.001^*		P < 0.001^*
Double temple-mastoidal stimulation (2)	Baseline	16.54	13.01,20.08	22.67	16.71,27.1
	Threshold	15.95	12.5,19.23	21.12	14.36,28.33
	pairwise comparison (baseline-threshold)	P = 1.000		P = 1.000
	baseline	16.54	13.01,20.08	22.67	16.71,27.1
	supra-threshold	18.37	15.05,23.71	25.97	20.15,36.57
	Pairwise comparison (baseline-supra-threshold)	P = 0.32		P = 0.006^*
	Threshold	15.95	12.5,19.23	21.12	14.36,28.33
	Supra-threshold	18.37	15.05,23.71	25.97	20.15,36.57
	Pairwise comparison (threshold-Supra-threshold)	P = 0.21		P < 0.001^*
Electrode montage 1 Vs. electrode montage 2	Threshold 1	14.95	11.28,19.67	18.7	13.97,23.75
	Threshold 2	15.95	12.5,19.23	21.12	14.36,28.33
	Pairwise comparison (Threshold 1- Threshold 2)	P = 0.96		P = 1.000
	Supra-threshold 1	18.42	15.97,23.15	26.32	22.73,38.93
	Supra-threshold 2	18.37	15.05,23.71	25.97	20.15,36.57
	Pairwise comparison (Supra-threshold 1- Supra-threshold 2)	P = 0.01		P = 1.000
	Threshold 1	14.95	11.28,19.67	18.7	13.97,23.75
	Supra-threshold 2	18.37	15.05,23.71	25.97	20.15,36.57
	Pairwise comparison (Threshold 1- Supra-threshold 2)	P < 0.002^*		P < 0.001^*
	Threshold 2	15.95	12.5,19.23	21.12	14.36,28.33
	Supra-threshold 1	18.42	15.97,23.15	26.32	22.73,38.93
	Pairwise comparison (Threshold 2- Supra-threshold 1)	P < 0.005^*		P = 0.001^*

	threshold		Supra-threshold
Base line	Double temple-mastoidal	Bipolar binaural	Double temple-mastoidal	Bipolar binaural
AP maximum amplitude	correlation coefficient = 63 P < 0.01^*	correlation coefficient = 43 P = 0.01	correlation coefficient = 70 P < 0.01^*	correlation coefficient = 35 P = 0.04
ML maximum amplitude	correlation coefficient = 15 P = 0.4	correlation coefficient = 84 P < 0.01^*	correlation coefficient = 32 P = 0.07	correlation coefficient = 80 P < 0.01^*