Impaired tilt perception in Parkinson's disease: a central vestibular integration failure

Abstract

Introduction: Impaired balance control is a hallmark symptom in Parkinson's disease (PD). Altered sensory-motor integration contributes to the deficiency. We aimed to determine whether impaired vestibular signal processing added to the disorder. We exposed patients (N = 11; 68±6y) and age-matched healthy subjects (hS: N = 19; 65±11y) on a motion platform in complete darkness to two consecutive forward tilt movements (12 series; N = 24; overall 288 trials) and asked them to indicate which tilt was perceived larger. By combing tilt movements with translations we manipulated vestibular sensory input in order to investigate whether putative impairment resulted from a deficiency of the sensory organs (semicircular canals in 'single-SCC-cue-condition', otoliths in 'single-OT-cue-condition') themselves or to a sensory integration failure ('multi-cue-condition').

Results: Tilt discrimination in the multi-cue-condition was inferior in patients compared to hS (p = 0.02). No significant differences between the two groups were found for both single-cue-conditions. Comparison of multi-cue-condition with a prediction resulting from the combination of both single-cue-conditions by optimal observer theory revealed that patients (p = 0.04), in contrast to hS, failed to efficiently combine SCC and OT information to improve tilt perception.

Conclusion: We found that PD patients distinguished forward tilts less precise than hS, suggesting impaired vestibular perception. Tilt discrimination in patients, moreover, did not improve as much as in hS in conditions where both SCC and OT information was available compared to conditions where only SCC or OT cues were activated. The latter provides evidence that tilt misperception in PD most likely results from an integration failure of vestibular signals.

Fig 1. Schematic illustration of tilt movements in the reference and comparison conditions with corresponding vestibular organ activation.

Upper row: initial head position; lower row: final head position. GIA = gravito-inertial acceleration; αTILT: head rotation angle in degrees; αGIA: rotation angle of the GIA vector. Comparison stimulus ‘multi-cue’: ‘Pure’ forward tilt movement. The tilt movement activates both SCC and OT as in the reference stimulus. Differences between both conditions result from the tilt angle, which, in this example, is smaller than the reference tilt angle (i.e. 3° vs. 5° tilt). Comparison stimulus ‘single-SCC-cue’: Combined tilt-translational movement. The backward translational movement modifies the αGIA. Movement parameters are thereby chosen to provide the same OT activation as in the reference stimulus (i.e. αGIA = 5°). Perceptual differences between this and the reference tilt stimulus thus result from SCC activity (i.e. αTILT = 3°). Comparison stimulus ‘single-OT-cue’: Combined tilt-translational movement. The forward translational movement modifies the αGIA providing different OT activation (i.e. αGIA = 3°). SCC activation (i.e. αGIA = 5°) is instead equal to the reference stimulus as the tilt angles are equal in both conditions. Difference: values represent the difference of vestibular, i.e. SCC and OT, activation between the reference and the stimulus.

Fig 2. Discrete approximation of the psychometric curves (dots connected by gray lines) in one healthy subject (upper panel) and one PD patient (patient index ID5; lower panel) plotted for the three conditions in separate three separate graphs (from left to right: ‘multi-cue’, ‘single-SCC’ and single-OT condition.

Abscissa: difference of SCC and/or OT activation between the comparison and reference stimulus. Ordinate: percentage of times the comparison stimulus was rated as larger tilt movement than the reference stimulus. Each of the four data points in each graph correspond to percentage recorded in all the trials of one the four comparison stimuli tested in that condition.

Fig 3. Discrete approximation of the psychometric curves (dots connected by gray lines) and cumulative Gaussian fit of the data representing the psychometric curves (dashed curves) in all healthy subjects (upper panel) and PD patients (lower panel).

Abscissa and ordinate as in Fig 2. Arrow: indicates the increased uncertainty during the backward translational—tilt movement compared to the forward translational—tilt movement, although the difference of vestibular activation between the comparison and reference stimulus was equal (see text for further explanation).