Visual feedback-dependent modulation of arousal, postural control, and muscle stretch reflexes assessed in real and virtual environments

Abstract

Introduction: The mechanisms regulating neuromuscular control of standing balance can be influenced by visual sensory feedback and arousal. Virtual reality (VR) is a cutting-edge tool for probing the neural control of balance and its dependence on visual feedback, but whether VR induces neuromodulation akin to that seen in real environments (eyes open vs. closed or ground level vs. height platform) remains unclear.

Methods: Here we monitored 20 healthy young adults (mean age 23.3 ± 3.2 years; 10 females) during four conditions of quiet standing. Two real world conditions (eyes open and eyes closed; REO and REC) preceded two eyes-open virtual 'low' (ground level; VRL) and 'high' (14 m height platform; VRH) conditions. We measured arousal via electrodermal activity and psychosocial questionnaires rating perceived fear and anxiety. We recorded surface electromyography over the right soleus, medial gastrocnemius, and tibialis anterior, and performed force plate posturography. As a proxy for modulations in neural control, we assessed lower limb reflexive muscle responses evoked by tendon vibration and electrical stimulation.

Results: Physiological and perceptual indicators of fear and anxiety increased in the VRH condition. Background soleus muscle activation was not different across conditions; however, significant increases in muscle activity were observed for medial gastrocnemius and tibialis anterior in VRH relative to REO. The mean power frequency of postural sway also increased in the VRH condition relative to REO. Finally, with a fixed stimulus level across conditions, mechanically evoked reflexes remained constant, while H-reflex amplitudes decreased in strength within virtual reality.

Discussion: Notably, H-reflexes were lower in the VRL condition than REO, suggesting that these ostensibly similar visual environments produce different states of reflexive balance control. In summary, we provide novel evidence that VR can be used to modulate upright postural control, but caution that standing balance in analogous real and virtual environments may involve different neural control states.

Keywords: H-reflexes; electrodermal activity; electromyography; muscle stretch reflexes; postural sway; tendon vibration; virtual reality.

FIGURE 1

Experimental setup and raw data. (A) Photo of a participant standing on the force plate, instrumented with a VR headset, lower-limb surface electromyography (EMG) electrodes, tibial nerve electrical stimulation electrodes over the popliteal fossa, a custom-made wearable tendon vibrator over the right Achilles tendon, and electrodermal activity (EDA) electrodes on the left palm. (B) Sample visual environment displayed to participants within the ‘Low’ and ‘High’ (looking down) VR conditions. (C) Raw data traces of EDA, force plate (centre of pressure, COP, in the anterolateral, AP, or mediolateral, ML, planes), rectified EMG (right soleus, RSOL; right medial gastrocnemius, RMG; right tibialis anterior, RTA), and vibrator acceleration (ACC), during an exemplary trial.

FIGURE 2

Physiological and perceptual indicators of postural threat-induced fear and anxiety. (A,B) Individual participant (gray circles) and box plots showing median and interquartile range for EDA under real (A) and virtual environments. (B) Individual and average participant questionnaire anxiety rating scores (out of a total of 0-144 points) under the different VR conditions. Asterisks denote significance at the p < 0.01** level.

FIGURE 3

Postural sway and soleus muscle activity across real and virtual conditions. (A) Root-mean-squared (RMS) amplitude and (B) mean power frequency of anteroposterior (AP) centre of pressure (COP) excursions, as well as (C) RMS soleus muscle activity, across each visual condition. Asterisks denote significance at the p < 0.01** level.

FIGURE 4

H-reflex responses across real and virtual visual conditions. (A) Anteroposterior (AP) centre of pressure (COP) during 5 consecutive electrical stimulation pulses (bars overtop), delivered to the tibial nerve at the popliteal fossa (S29, REO). A notable, forward-directed transient in AP COP postural sway resulting from rapid plantar flexion is observed in response to each electrical pulse. (B) Exemplary postural (AP COP) and muscular (RSOL EMG) response to electrical stimulation from a single pulse (initial pulse in panel A). (C) Stimulus-triggered average of the 5 consecutive pulses from this trial, with the RSOL stimulus artifact, M-Wave, and H-Reflex labeled. (D) Individual participant (gray dashed lines/diamonds) and average (black bars) peak-to-peak (P2P) H-reflex amplitudes. Asterisks denote significance at the p < 0.001*** level.

FIGURE 5

Noisy tendon vibration evoked responses across real and virtual conditions. (A) Pooled coherence spectral estimates (N = 20) under real (REO vs. REC) and virtual (VRL vs. VRH) visual environments. (B) Pairwise difference of coherence (DoC) tests comparing the pooled coherence spectra under real (REC-REO) and virtual (VRH-VRL) conditions. Comparisons between baseline (REO) and virtual (VRL-REO, VHR-REO) visual conditions are also shown. Solid horizontal black lines denote the 95% CI for each pairwise DoC test and values exceeding these lines (asterisks) indicate that, in all cases, NTV-muscular coherence was higher under conditions of greater postural threat. (C) Pooled cumulant density estimates for the relationship between vibration magnitude and rectified soleus activity under real and virtual visual environments. (D) Individual participant (gray dashed lines/diamonds) and average (black bars) peak-to-peak (P2P) cumulant density amplitudes.