Frequency analyses of postural sway demonstrate the use of sounds for balance given vestibular loss

Purpose: To investigate how adults with unilateral vestibular hypofunction and healthy controls incorporate visual and auditory cues for postural control in an abstract visual environment.

Methods: Participants stood on foam wearing the HTC Vive, observing an immersive 3-wall display of 'stars' that were either static or dynamic (moving front to back at 32 mm, 0.2 Hz) with no sound, static white noise, or moving white noise played via headphones. Each 60-second condition repeated twice. We recorded the center-of-pressure variance, and its power spectral density [PSD, cm²] components in low [0, 0.25 Hz], mid [0.25, 0.5 Hz] and high [0.5, 1 Hz] frequencies in the anterior-posterior direction. We used linear mixed-effects models to compares healthy controls (n = 41, mean age 52 years, range 22-78) to participants with unilateral peripheral vestibular hypofunction (n = 28, 61.5, 27-82), adjusting for age.

Results: Variance and low PSD: we observed a significant vestibular by visual load interaction in the presence of sounds, such that the vestibular group had significantly higher sway than controls only on dynamic visuals in the presence of sounds. Mid PSD: the vestibular group had significantly higher sway than controls regardless of condition. High PSD: the vestibular group had significantly higher sway than controls, except for the presence of sounds on static visuals.

Conclusions: Patients with vestibular hypofunction used sounds to reduce sway in a static abstract environment and were somewhat destabilized by it in a dynamic environment. This suggests that sounds, when played from headphones, may function as an auditory anchor under certain level of challenge and specific tasks regardless of whether it's stationary or moving. Our results support that increased sway in middle frequencies reflects vestibular dysfunction.