Using Low Levels of Stochastic Vestibular Stimulation to Improve Balance Function

Abstract

Low-level stochastic vestibular stimulation (SVS) has been associated with improved postural responses in the medio-lateral (ML) direction, but its effect in improving balance function in both the ML and anterior-posterior (AP) directions has not been studied. In this series of studies, the efficacy of applying low amplitude SVS in 0-30 Hz range between the mastoids in the ML direction on improving cross-planar balance function was investigated. Forty-five (45) subjects stood on a compliant surface with their eyes closed and were instructed to maintain a stable upright stance. Measures of stability of the head, trunk, and whole body were quantified in ML, AP and combined APML directions. Results show that binaural bipolar SVS given in the ML direction significantly improved balance performance with the peak of optimal stimulus amplitude predominantly in the range of 100-500 μA for all the three directions, exhibiting stochastic resonance (SR) phenomenon. Objective perceptual and body motion thresholds as estimates of internal noise while subjects sat on a chair with their eyes closed and were given 1 Hz bipolar binaural sinusoidal electrical stimuli were also measured. In general, there was no significant difference between estimates of perceptual and body motion thresholds. The average optimal SVS amplitude that improved balance performance (peak SVS amplitude normalized to perceptual threshold) was estimated to be 46% in ML, 53% in AP, and 50% in APML directions. A miniature patch-type SVS device may be useful to improve balance function in people with disabilities due to aging, Parkinson's disease or in astronauts returning from long-duration space flight.

Fig 1. An exemplar plot of the 12 measures (ML direction: Fy, Mx, Hay, Hrv, Tay, Trv and AP direction: Fx, My, Hax, Hpv, Tax, Tpv) of interest for one subject during the balance task, for both baseline and stimulus periods for an optimal trial at the level of 350 μA (= 40% of this subjects perceptual threshold).

Numerals in the bottom right of each panel represent RMS value.

Fig 2. Ratio data of all twelve measures of RMS during the stimulus period to the RMS during the baseline period at different stimulus level ranges, for the same subject as in Fig 1.

Note that for this subject, performance improved in all directions (ML, AP, and APML) at the same stimulus amplitude range of ±350 μA. Further, this subject had to be assisted to prevent falling and did not complete trials at the two highest nominal stimulation levels tested (±2800 μA, ±3300 μA).

Fig 3. Mean (± one Standard Error of Mean) across all subjects (n = 45) showing RMS values of the six measures of interest for ML (see data in S2 Table), and AP (see data in S3 Table), and twelve measures for APML (see data in S4 Table), during the two periods (baseline and stimulus) of the control and optimal trials.

Fig 4. Exemplar plot of different measures of interest for a typical subject during the thresholding task.

Fig 5. Nonlinear regression fitted lines of logistic psychometric function on the normalized percent time data at different stimulation amplitudes for a typical subject.

(A) motion perception using Joystick data (red squares indicate normalized percent time of perceptual motion detected at each stimulation level), (B) body sway using physiological data: Fy, Mx, Hay, Hrv, Tay, Trv (individual normalized percent time data points quantified for each of the different body sway measures are not shown for clarity).

Fig 6. The threshold amplitude data estimated using the seven measures (perceptual: Joystick, and physiological body sway: Fy, Mx, Hay, Hrv, Tay, Trv) for each of the 18 subjects (see data in S5 Table).