Directional effects of whole-body spinning and visual flow in virtual reality on vagal neuromodulation

1.1 Vestibulo-vagal links

Apart from investigations into vestibulo-sympathetic responses [78], there has been a lack of investigation into vestibulo-vagal interactions. The horizontal semi-circular canals (SCC) play a key role in these interactions, given their neural connections to the medial vestibular nuclei (MVN) and the inferior vestibular nuclei (IVN) in primates [13]. From the MVN and IVN there are bilateral projections to the nucleus tractus solitarius (NTS), an ANS relay centre that mediates afferent and efferent vagal activity [79], the dorsal motor nucleus of the vagus nerve, and nucleus ambiguous [8, 62]. Otolith organs also contribute inputs to the NTS, as primary afferents from the otoliths innervate the inferior vestibular nuclei [13]. Otolith organs however, are the main mediators for the vestibulo-sympathetic reflex, while the horizontal SCCs given their neural connections, are a structure more specific to vagal interactions [78]. In an animal study, single shock electrical stimulation of the vestibular nerve excites the vagus nerve further down in the ipsilateral, but not contralateral pathway [2]. Although it is not clear which part, or if all the vestibular labyrinth including the SCCs and otolith organs were stimulated, this study provides physiological evidence for vestibulo-vagal links [2]. Vagal fibres project out from the NTS to the associated dorsal motor nucleus of the vagus nerve and the nucleus ambiguus, sending right and left parasympathetic vagal nerve fibres to the sinoatrial (SA) and atrioventricular nodes (AV) of the heart respectively [30]. Additionally, activation of the NTS supports increase of parasympathetic vagal activity by acting on the caudal ventrolateral medulla (CVLM). The CVLM inhibits the rostral ventrolateral medulla (RVLM), responsible for sympathetic activation [9, 72], thereby creating a neural pathway for the vestibular system to influence both sympathetic and parasympathetic branches of cardiac activity (Fig. 1).

OPEN IN VIEWER

Fig. 1

Neurovisceral model [8, 72] showing mirrored anatomical pathways to left or right vagus nerve, and the heart. Diagram shows top-down emotional, cognitive and physiological regulation along with vestibular autonomic control of the heart. PFC = Prefrontal cortex, mPFC = medial prefrontal cortex, OFC = orbitofrontal cortex, CC = cingulate cortex, CeA = central nucleus of the amygdala, PVN = paraventricular nucleus. LHA= lateral hypothalamic area, PAG= periaqueductal gray, PPN = parabrachial pontine nucleus, LC = locus coeruleus, LTF = medullary lateral tegmental field, hSCC = horizontal semi-circular canal, MVN= medial vestibular nuclei, IVN = inferior vestibular nuclei, NTS = nucleus tractus solitarius, DMV= dorsal motor nucleus of the vagus nerve, NAm= nucleus ambiguus, CVLM= caudal ventrolateral medulla. RVLM= rostral ventrolateral medulla. IML Column = intermediolateral column of the spinal cord, AV node = atrioventricular. SA node= sinoatrial node.

Artificial methods of vestibular stimulation widely used in literature include galvanic vestibular stimulation (GVS) [21] and caloric vestibular stimulation (CVS) [54]. In GVS, electricity is commonly applied between a cathode on one side of the posterior neck’s mastoid process, and an anode on the other, which can be used to stimulate and inhibit the vestibular nerve respectively [21]. GVS is thought to primarily target the otoliths with weak SCC activation [16], although this has been heavily contested [43]. With CVS, cold or hot water is applied to the inner ear canal to elicit SCC fluid movement [54]. GVS and CVS are non-specific to the horizontal SCCs, which are connected to the vagal nuclei through the medial and inferior vestibular nuclei [13, 25]. Due to electrode placement on the mastoid processes, GVS likely not only stimulates the vestibular nerve but branches of spinal nerves in the skin [15]. Some examples include the great auricular nerve, which communicates with the lesser occipital, auricular branch of the vagus, and posterior auricular branch of the facial nerve [15]. Additionally, any number of structures may be stimulated by the electrical potential gradient generated between the anode and cathode on each side of the posterior neck [24]. For CVS, the liquid may activate nearby thermoreceptive and nociceptive receptors in the skin that would otherwise not be activated during movement [25]. Along with innervations of the external ear, the stimulus may also reach the tympanic membrane, activating a web of at least 3 other different cranial nerves, the trigeminal (CN V), vagus (CN X) and glossopharyngeal (CN IX) nerves [25]. Such a stimulus would introduce inputs not normally included in a physiological vestibular response [25]. Additionally, CVS does not take into account the ipsilateral activation along with contralateral inhibition of the vestibular nerve during head or whole-body movements [41]. Considering GVS and CVS limitations, yaw (horizontal plane) rotation, which stimulate the horizontal SCCs and the vagus nerve, provide a more prudent method for investigating vestibulo-vagal interactions. A natural vestibular stimulation of the horizontal SCCs through whole-body yaw rotation is attainable by passively spinning in a Barany chair [60], providing a stimulus closer to real world physiological conditions. An overview of the types of vestibular stimulation mentioned and their utility for investigating vestibulo-vagal interactions is summarized in Table 1.

It is possible that laterality of vestibular nerve activation brought about through direction of rotation has implications for vestibular effects on the heart. When rotating the head or whole body in the yaw plane, the horizontal SCCs are activated and the vestibular nerve on the same side as the direction of rotation is excited, while the vestibular nerve on the contralateral side is inhibited [41]. With vestibulo-vagal interactions in mind, it is unknown in the literature how direction of rotation, and therefore, lateralisation of stimulus during clockwise (CW) and anticlockwise (ACW) rotation affects vagus nerve activity further down in the neural pathway [8, 62]. Studies so far have missed this significant lateralisation, and looked at rotation either constantly in one direction [73], or mixed directions by periodically switching between clockwise and anticlockwise rotation in a sinusoidal manner [55, 59]. Lateralisation of stimulus has clinical importance as the left vagus nerve has been frequently used over the right vagus nerve in vagal nerve stimulation (VNS) treatments for epilepsy and depression [53]. The reasoning is that it innervates the AV node and is thought to not alter cardiac activity to the same extent as the right vagus nerve, which innervates the SA node, thereby reducing cardiac complications [53]. Adding to this reasoning is evidence for dominant effects of the SA node on cardiac activity and the pacemaker cells, over that of the AV node [31, 67]. The relative contributions of left and right VNS to altering cardiac activity is, however, unclear as other studies have contradicting views, suggesting no unilateral dominance on cardiac activity [77].

1.4 Visual input on vagal response

Another aspect to consider is the unknown effects of not just vestibular but also visual input on vagus nerve responses. To our knowledge there has been no study on vagal responses that unites conditions of eyes open and closed while undergoing a constant natural vestibular stimulation in one direction, along with a purely visual stimuli mimicking the same environment in a within-subjects design [17, 73].

1.5 Measures of autonomic activity

Current technology presents two reliable non-invasive measures for both sympathetic and parasympathetic branches of autonomic activity: heart rate variability (HRV) [68] and pupil diameter [52]. HRV is derived from electrocardiogram (ECG) recordings and can be used to measure variations in cardiac activity rather than a mean level such as with heart rate (HR) [68]. Through these variations, HRV gives information that HR alone cannot by reflecting sympathovagal balance, an interplay of sympathetic and parasympathetic vagal activity on the heart [68]. HRV relies on beat-to-beat variations in heart rate, measured by the intervals between peaks of an ECG trace (R-R peak interval variation) [68]. A common parasympathetic HRV parameter is the root mean square of successive RR interval differences (RMSSD), which provides an indicator of vagal activity robust to changes in respiration [68]. Others include the parasympathetic nervous system (PNS) index and high frequency (HF) power that also reflect vagal activity, but unlike RMSSD, HF is influenced by respiration [68]. These increase with higher parasympathetic vagal activity on the heart, and vice versa [71]. HRV parameters also include sympathetic indicators such as stress index (SI), which is the square root of Baevsky’s stress index [5], sympathetic nervous system (SNS) index and low frequency (LF) power [71]. These increase with higher sympathetic activity on the heart, and vice versa [71], with the exception of LF, which is not solely a reflection of sympathetic drive, and has variations dependent on parasympathetic vagal activity as well [68].

Pupil diameter changes reflect autonomic changes, but from entirely different neural circuits in the body compared to HRV changes [30, 52] Pupil constriction arises from dominant parasympathetic activity routed through the Edinger-Westphal (EW) nucleus and controlled by the ciliary ganglion, while pupil dilation in response to dominant sympathetic activity is controlled by information that travels from the intermediolateral column (IML) in C8-T2 segments of the spinal cord, to the superior cervical ganglion [52]. The vestibular nuclei, including the MVN, have direct projections to putative preganglionic parasympathetic neurons in the Edinger-Westphal, anteromedian and parabrachial nuclei in rabbits, which would likely mediate vestibular effects on pupil diameter changes [7]. Despite known connections, the response of pupil diameter to natural vestibular stimulation has not been investigated thoroughly. One study used artificial mechanical air puff stimulation to the utricle of the cat and monkey, and reported conjugate nystagmus, with pupillary constriction during fast phases and dilation during slow phases of nystagmus [19]. These responses were unaffected by anatomical sectioning of sympathetic input to the eye, concluding that pupil dilation could be achieved by pure inhibition of parasympathetic tone [19]. To the best of our knowledge no study to date has measured pupil diameter while stimulating the horizontal SCC using whole-body rotation in humans, nor compared results to that of cardiac autonomic activity in the same scenario.

1.6 Aim

The present study aimed to investigate potential direction specific effects of constant passive whole-body yaw rotation in real environments with eyes open and closed, and purely visual flow without body movement in virtual environments on the modulation of ANS functioning. In this context, the study focused on both branches of autonomic activity with the aid of HRV and pupil response monitorisation. A novel approach was incorporated by combining and separating visual and vestibular inputs. This was achieved through the usage of an electronically powered spinning chair in real-life, and ‘VR spinning’ where Virtual Reality (VR) videos mimic the same real-world visual spinning environment in both CW and ACW directions.

1.7 Contributions

In summary, we contribute the following:

Vagal modulation through spinning of the whole-body in the real world or through spinning of the visual environment is direction dependent, meaning only one direction of spinning has an observable effect on vagal activity.

2.1 Subjects

Twenty seven healthy, adult male subjects, age range of 20– 30 years (mean age 22.4, standard deviation±2.6) with a pass score on the motion sickness susceptibility questionnaire [29] (MSSQ-Short) (20 and below), which includes people from a general population below the 75th percentile for motion sickness susceptibility [45]. Subjects were recruited from the student and working population. Exclusion criteria were previous diagnosis of neurological disorder, cardiovascular disease, diabetes, gastrointestinal disorder, on medication or smoking. All subjects had either normal or corrected visual acuity with contact lenses. HRV was obtained from n = 27 participants. Pupil diameter data was only obtained from n = 23 participants out of n = 27, with four excluded due to data loss.

This study was approved by the University of Otago Ethics Committee (H19/085) and performed in accordance with relevant guidelines and regulations. All participants provided signed consent.

2.2 Experimental equipment

2.3 Protocol

Twenty-seven subjects each participated in four sessions of whole-body yaw rotation, and two sessions of VR spinning in a randomized order, giving six stimulation types (Fig. 2). Both vestibular and visual stimulation were either provided together, or separately in different directions of rotation. A minimum of twenty-four hours wash out period was used between each session.

OPEN IN VIEWER

Fig. 2

Experiment flowchart showing stimulation types given at random and relevant recordings used for each stimulation type.

Total time for whole-body yaw rotation experiments in the Barany chair and recordings was six minutes. ECG and pupil diameter were recorded for eyes open conditions, whereas only ECG was recorded for eyes closed conditions. Subjects started with two minutes of stationary sitting. Then, stimulation commenced with either CW open, CW closed, ACW open, or ACW closed passive whole-body yaw rotation, at 90°/s constant velocity for two minutes. After rotation, subjects sat stationary for two minutes. For VR conditions, each subject viewed the same spinning environments as experienced during whole-body yaw rotation on video in VR, while their ECG was recorded. Total time for VR immersion and ECG recording was six minutes. The VR environment faced a bullseye for the baseline period of 2 minutes, then visually rotated in the CW or ACW direction, at 90°/s constant velocity for two minutes. After this, subjects remained watching the stationary VR environment for two minutes post-stimulation.

Prism version 8.2.1 for Windows (GraphPad Software, San Diego, California USA) was used for statistical analysis and graphing. Prism version 8.2.1 uses Barlett’s and Brown-Forsythe tests for one-way ANOVAs, and methods outlined by Glantz and Slinker [28] for two-way ANOVAs. A repeated measures one-way ANOVA was used to determine overall significance in changes of pupil diameter, RMSSD, PNS, SI and SNS in baseline, stimulation and post stimulation stages of every stimulus type. Post-hoc Tukey’s multiple comparison test was used to compare between stages [49]. A repeated measures two-way ANOVA [28] was used to compare the extent of statistically significant changes between stimulation types. Geisser-Greenhouse correction [49] was applied for non-sphericity of data. p < 0.05 was set as the limit for significance. For exact values of one-way ANOVA descriptive statistics refer to Table 2.

3.1 RMSSD

According to one-way ANOVA a statistically significant decrease was found in RMSSD between stages for CW open, CW closed and CW VR, (p < 0.05). Statistically significant decreases were found in post-hoc tests for RMSSD from baseline to stimulation in the CW open (p = 0.0020), CW closed (p = 0.0460) and CW VR (p = 0.0277) conditions, and from baseline to post-stimulation in the CW closed (p = 0.0245) and CW VR (p = 0.0056) conditions (Fig. 3a). No statistical difference in RMSSD was observed between stages using one-way ANOVA for any of the ACW conditions (ACW open, ACW closed, ACW VR) between baseline, stimulation and post stimulation, (p > 0.05) (Fig. 3b).

OPEN IN VIEWER

Fig. 3

HRV indicators related to parasympathetic activity for different stimuli. Root Mean Square of Successive Differences (RMSSD) and during baseline, stimulation and post stimulation for a) CW open, CW closed, CW VR, b) ACW open, ACW closed, ACW VR conditions. PNS index during baseline, stimulation and post stimulation for a) CW open, CW closed, CW VR, b) ACW open, ACW closed, ACW VR conditions. Error bars represent mean with standard deviation. ^*indicates significant difference between conditions. ^*p < 0.05, ^**p < 0.01.

No statistically significant (p = 0.4241) difference between stimulation types was found in RMSSD when comparing the baseline, stimulation and post-stimulation of CW open, CW closed and CW VR conditions as determined by two-way ANOVA F(1.544, 40.15) = 0.8078.

According to One-way ANOVA, CW VR is the only condition to show a statistically significant decrease between stages (p = 0.0027), with post hoc tests finding a decrease in PNS index between baseline and post stimulation (p = 0.0053) (Fig. 3c). The rest of the conditions, CW open, CW closed, ACW open, ACW closed, ACW VR, did not show any change in PNS index from baseline p > 0.05 (Fig. 3c-d).

3.3 Stress index (SI)

One-way ANOVA found that both vestibular and visual stimulation separately and combined, significantly altered the stress index (SI) between stages in all conditions, CW open, CW closed, CW VR (Fig. 4a), ACW VR, and ACW closed (P < 0.01), except for ACW open conditions (P > 0.05) (Fig. 4b). Post-hoc tests found that there was a statistically significant increase in SI from baseline to stimulation in CW open (p = 0.0007), CW closed (p = 0.0006), CW VR (p = 0.0069) (Fig. 4a), ACW VR (p = 0.0124) and ACW closed (p = 0.0068) conditions (Fig. 4b). Statistically significant increases in SI were found from baseline to post-stimulation in CW closed (p = 0.0094) (Fig. 4a) and ACW VR (p = 0.0218) conditions (Fig. 4b). Statistically significant decreases in SI were found from stimulation to post stimulation in CW open (p = 0.0167) (Fig. 4a) and ACW closed conditions (p = 0.0430) (Fig. 4b). Upon two-way ANOVA, No statistically significant (p = 0.7523) difference between stimulation types was found in SI when comparing the baseline, stimulation and post-stimulation of CW open, CW closed, CW VR, ACW VR and ACW closed conditions (p = 0.9801), F(3.507, 91.19) = 0.08401 (Fig. 4a-b).

OPEN IN VIEWER

Fig. 4

HRV indicators related to sympathetic activity for different stimuli. Stress index (SI) during baseline, stimulation and post stimulation for a) CW open, CW closed, CW VR, b) ACW open, ACW closed, ACW VR conditions. ^*indicates significant difference between conditions. SNS index during baseline, stimulation and post stimulation for a) CW open, CW closed, CW VR, b) ACW open, ACW closed, ACW VR conditions. Error bars represent mean with standard deviation. ^*indicates significant difference between conditions. ^*p < 0.05, ^**p < 0.01. ^***p < 0.001

According to one-way ANOVA, both vestibular and visual stimulation separately and combined, significantly altered sympathetic nervous system (SNS) index between stages in all conditions, CW open, CW closed, CW VR (Fig. 4c), ACW VR and ACW closed conditions (p < 0.05) except ACW open (p > 0.05) (Fig. 4d). Post-hoc tests could not find any statistically significant difference between baseline, stimulation and post stimulation (p > 0.05) in the ACW closed condition, despite there being an indication of a difference existing between stages on the whole (Fig. 4c). Post-hoc tests found statistically significant increases in SNS index from baseline to stimulation in CW open (p = 0.0029), CW closed (p = 0.0195), CW VR (p = 0.0151) (Fig. 4c), ACW VR (p < 0.0241) (Fig. 4d). Statistically significant increases in SNS index were found from baseline to post-stimulation in CW closed (p = 0.0095) (Fig. 4c) and ACW VR (p = 0.0486) conditions (Fig. 4d). As determined by two-way ANOVA, no statistically significant (p > 0.05) difference between stimulation types in SNS index was found when comparing the baseline, stimulation and post-stimulation of CW open, CW closed, CW VR, (Fig. 4c) ACW VR and ACW closed conditions (Fig. 4d), p = 0.9359, F(3.491, 90.77) = 0.1726.

3.5 Mean HR

No significant difference was observed in mean HR between all stimulation types, CW open, CW closed, CW VR, ACW open, ACW closed, ACW VR, at baseline, stimulation and post stimulation as determined by one-way ANOVA (p > 0.05) (Fig. 5).

OPEN IN VIEWER

Fig. 5

Mean heart rate (HR) during baseline, stimulation and post stimulation for a) CW open, CW closed, CW VR, b) ACW open, ACW closed, ACW VR conditions. Error bars represent mean with standard deviation. p > 0.05.

One-way ANOVA and multiple comparisons post-hoc Tukey correction found statistically significant pupil diameter increases from baseline to stimulation (p < 0.0001) in the left and right eyes during both CW open and ACW open whole-body yaw rotation (Fig. 6).

OPEN IN VIEWER

Fig. 6

Pupil diameter during baseline, stimulation and post stimulation for left and right eyes in the ACW and CW conditions. Error bars represent mean with standard deviation. ^*indicates significant difference between conditions. ^***p < 0.001, ^****p < 0.0001.

As determined by two-way ANOVA, no statistically significant (p = 0.5120) difference was found in pupil diameters when comparing between eyes open whole-body rotation conditions of ACW left pupil, ACW right pupil, CW left pupil or CW right pupil. In other words, there was no difference between direction of rotation, nor left and right pupil diameters for each stimulus, F(1.902, 41.84) = 0.6657.

4.1 Spinning direction dependent vagal nerve neuromodulation

Direction of rotation and its influence on HRV or pupil diameter were investigated for the first time in this study. We used both CW and ACW directions of yaw rotation during spinning of the whole-body and spinning of the visual environment played through a video in VR. Demonstrated here is CW only suppression of parasympathetic vagal activity on the heart, indicated by reduced RMSSD in response to both whole-body and video in VR spinning (CW open (p = 0.0020), CW closed (p = 0.0460) and CW VR (p = 0.0277)). Consistent to our findings, a decrease in parasympathetic vagal activity in terms of RMSSD, and an increase in sympathetic activity, but using different indicators of LF power and LF/HF ratio, have been found during constant CW rotation at 90°/s in male jet pilots [73]. In addition, we observed increased sympathetic activity in almost all ACW conditions, including ACW closed and ACW VR (SI p < 0.01, SNS index p < 0.05), except for ACW open (SI p > 0.05, SNS index p > 0.05). Therefore, parasympathetic vagal suppression in response to both vestibular and visual stimulation in the present study is spinning direction dependent, and in contrast there is no spinning direction effect on sympathetic activity. Thus, even though sympathetic activity can inhibit vagal activation through inhibition of the NTS [72], sympathetic mediated vagal inhibition is likely not the sole driver for the decreased RMSSD in CW conditions. Laterality in cardiac responses to vagus nerve action is the premise for choosing the left vagus nerve for stimulation over the right vagus nerve in clinical treatments [31, 53]. In contrast, it has been suggested that there is no unilateral dominance of vagal activity on the heart [77]. Here, we present the notion that laterality of cardiac responses should be considered when using natural methods of vagal modulation such as vestibular or visual stimulation.

Vagal suppression points in part to inhibition within the vestibulo-vagal circuit specifically due to horizontal SCC stimulation. The MVN has predominantly GABA_A mediated inhibition [12], and it is suggested that neurons with GABA_A receptors are likely to provide intrinsic connections, acting as local circuit interneurons within the MVN [6]. Therefore, suppression of vagal activity may arise from local circuit inhibition of the MVN ipsilateral to the vestibular nerve stimulated. Inhibition at the MVN would then affect bilateral projections to the NTS, dorsal motor nucleus of the vagus nerve and nucleus ambiguus which control efferent vagal responses to the heart [8, 62]. Furthermore, vestibular nerve electrostimulation inhibits the contralateral MVN [39, 48] interconnected through bilateral commissural fibres [61], which could also alter signals downstream. Our results may be explained in that inhibition at the MVN ipsilateral to the clockwise direction of rotation, ultimately leads to suppression of right vagal activity on the SA node. Whereas in contrast there may be negligible contribution of the AV node innervated by the left vagus to changes in R-R intervals that constitute RMSSD and PNS index [30, 70]. For anticlockwise rotations, suppression of left vagus nerve activity on the AV node may have had negligible effects on RMSSD and PNS index, whereas subsequent changes in right vagus activity, if any, may not have been enough to produce a discernible change in RMSSD or PNS index [30, 70]. Further studies on the vestibulo-vagal circuit may be conducted for more clarification on this matter.

Vestibular influences in our study are not the only contributor to the observed vagal suppression. Vagal suppression was observed in CW VR conditions where there was only constant yaw visual flow to the left. This type of visual spinning direction dependent response has not been reported yet in literature. Little is known about how direction of visual flow affects HRV. Reduced RMMSD has been observed in response to visual flow while travelling on a VR rollercoaster in the forwards but not backwards direction [50]. Our study uses a different type of visual flow to that of a VR rollercoaster, but there may be a common neural pathway mediating visually induced vagal suppression. Saccadic and pursuit eye movements involved in the vestibular ocular reflex (VOR) and optokinetic reflex may use the same orbital muscles involved in the oculocardiac reflex [23]. This reflex has been observed to cause sudden bradycardia from muscle traction of mainly the medial rectus during eye surgeries, pointing towards a possible neural pathway for visual flow to affect vagal responses [23]. Still, the explanation for differences in visual flow remains unknown. Further research of eye movement and the visual system’s connections to lateralized cardiac vagal responses could be investigated.

Regardless of stimulation type, the results of the present study demonstrated that purely visual (VR), vestibular (eyes closed conditions) and visual-vestibular combined (eyes open conditions) CW passive yaw rotation inhibits parasympathetic activity to the same extent, with no compounding effects for combined stimuli. Thus, the contribution of VOR induced by vestibular stimuli on vagus nerve activity in this study was likely small or negligible. Different speeds, time of rotation, and environments still need to be explored. Further studies may assess gain of VOR and track gaze to clarify the association of these variables with spinning direction dependent responses of vagus activity [58].

4.2 Consistency across multiple HRV measures

CW VR was the only condition where the PNS index, not used in previous studies [55, 73], significantly decreased during stimulation. Both the PNS Index and SNS Index provided by Kubios are made of valid measures, computed based on a proprietary formula [4, 71]. Known links with exercise intensity, HR and HRV are considered in the optimisation of weightages adopted in PNS and SNS computations. RMSSD is the absolute vagal activity which is robust to changes in respiration, in comparison to frequency-based parameters [44]. Moreover, a sufficient minimum of 2 minutes recording for time-domain HRV indices such as RMSSD was used [68]. Looking at the consistency in the majority of results with two HRV indices for parasympathetic activity, it seems reasonable to deduce a trend towards suppression of parasympathetic activity during clockwise stimulation only. These results add to other studies using optokinetic stimulation, or virtual reality (VR) to create visual-vestibular conflicts, that found sympathetic dominance and decreased parasympathetic activity according to HRV indicators, HF, LF and their ratios [35, 56].

4.3 Subtlety in autonomic changes

Whole-body rotation in a mechanical chair is a relatively passive ordeal compared to heart rate changes observed during exercise [74]. Despite sympathetic increases, the lack of any significant change in mean HR points towards subtle changes in cardiac response to the stimulation in all conditions, which can only be detected using HRV. Sympathetic increases in almost all vestibular and visual conditions of stimulation (ACW closed, ACW VR, CW open, CW closed and CW VR), could be due to top-down cognitive and emotional influences of anxiety, excitement and attention (Fig. 1). Other than through a neuroendocrine hypothalamic-pituitary-adrenal (HPA) axis [33], these cognitive and emotional influences can mediate sympathetic outflow through similar structures involved in autonomic changes in response to vestibular stimuli [51] such as the RVLM [14] and CVLM [57], shown in Fig. 1. These influences are known to contribute to increased SI [5], SNS index [71] and pupil diameter [52]. Vestibular influences, however, likely had little contribution to sympathetic increases, as there was no significant difference between conditions employing whole-body rotation and VR.

4.4 Pupil diameter response to vestibular stimulation

Increased pupil dilation during vestibular stimulation observed in this study is consistent with previous animal literature in cats and monkeys [19]. Given that sympathetic responses are interconnected across the heart and pupils through the IML [30, 52], an increase in pupil dilation suggests an overall dominant sympathetic increase associated with an increase in SI and SNS index. In contrast, the relative contribution of parasympathetic suppression on the pupils in this experiment cannot be observed. Fluctuations in pupil diameter, rather than an average over time, could be explored in the future as a means to figure out parasympathetic contribution [66]. Nevertheless, unlike the ability for multiple organs such as the heart and pupils to reflect sympathetic rise of activity in parallel [30, 52], our findings of parasympathetic vagal suppression cannot be assumed to also reflect on parasympathetic control of the pupils, as while there may be overlap, parasympathetic responses of the heart and pupils are controlled by different neural circuits [30, 52]. In other words, while sympathetic activation can be easily reflected in both ECG and pupil diameter data, this is not necessarily the case with parasympathetic activation. Thus, ECG and pupil diameter may not reflect the same amount of autonomic information as each other.

Nonetheless, it is conceivable that vagus modulation could interact with pupil diameter, given that both locus coeruleus – norepinephrine (LC-NE) activity and cholinergic transmission are related to pupil size and vagus activity [11, 63]. In addition, the LC is a downstream projection area of the NTS, which in turn is a projection area of the vagus nerve [26]. Furthermore, cervical subcutaneous vagal nerve stimulation has been shown to increases pupil diameter [20]. In contrast, alterations to vagal activity through transcutaneous auricular vagal nerve stimulation have resulted in no modulation of pupillary size and event-related pupil response [40]. In our study it is not known if vagus modulation had a contributing effect to increased pupil diameter during stimulation.

4.5 Neuroplasticity

Plasticity of the vagus nerve [1] was observed in sustained decrease of RMSSD from stimulation to post-stimulation compared to baseline in the CW closed and CW VR conditions. On the other hand, remaining cognitive, emotional or attentional modulation even after stimulation has ended might explain the sustained increase of sympathetic activity found between stimulation and post-stimulation in SI for CW closed, ACW VR conditions, and in SNS index for CW open and ACW VR conditions. An alternative consideration against the claim for neuroplastic responses is that the few seconds of continual endolymph flow after rotation halts, as evidenced by catch-up saccades [18, 69] may also be a contributing factor to HRV changes exhibited overall in the two minute-recordings, in which case the argument for any true plasticity could be doubted. This alternate viewpoint is doubtful as it is unknown if these few seconds of contribution would change a two-minute post stimulation recording.

Neuroplastic responses post stimulation were only detectable via ECG, but not with pupil diameter. This suggests that pupil diameter may be useful as an indicator for neuromodulation, but not neuroplasticity. Different acute neuroplastic responses of neural circuits involved in cardiac-vagal control and pupil size to the same vestibular stimulation may be an influencing factor [3, 46].

4.6 Future considerations

While adaptation to postural instability in visual environments can happen after 4-5 repeated sessions of VR [27], how quickly adaptation occurs in the autonomic nervous system to VR training is not known. The ability to use acute vestibular and visual stimulation to induce autonomic nerve plasticity would be a step towards development of training protocols to alter chronic cardiac activity in individuals, possibly through passive rotation in an automatic rotating chair, or the more accessible VR method. Our findings open an avenue into vagal suppression treatments to reduce high levels of parasympathetic activity. Such treatments might be useful in treating individuals with respiratory diseases like asthma or chronic obstructive pulmonary disease, where increased parasympathetic activity leads to greater mucus secretion and constriction of bronchial smooth muscle [22, 75]. Other conditions with vagal overactivity that may be treated include chronic hypotension [64]. These findings have implications in refining training protocols for pilots [73] and astronauts [32, 76] as well as any individuals undergoing vestibular training. Furthermore, laterality of cardiac responses to spinning provides insight into body responses to motion integrated VR gaming environments [47]. As this was an acute effects study, chronic, repeated studies are still needed to elucidate viability for clinical translation. Duration and amplitude of plasticity as well as potential meta-plasticity, which is a change of plasticity amplitude between trainings [1], could be outcomes to investigate in further research. Future studies may investigate higher speeds of rotation, which may create stronger endolymph flow and alter vestibular nerve responses [41].