Central crosstalk for somatic tinnitus: abnormal vergence eye movements

Abstract

Background: Frequent oulomotricity problems with orthoptic testing were reported in patients with tinnitus. This study examines with objective recordings vergence eye movements in patients with somatic tinnitus patients with ability to modify their subjective tinnitus percept by various movements, such as jaw, neck, eye movements or skin pressure.

Methods: Vergence eye movements were recorded with the Eyelink II video system in 15 (23-63 years) control adults and 19 (36-62 years) subjects with somatic tinnitus.

Findings: 1) Accuracy of divergence but not of convergence was lower in subjects with somatic tinnitus than in control subjects. 2) Vergence duration was longer and peak velocity was lower in subjects with somatic tinnitus than in control subjects. 3) The number of embedded saccades and the amplitude of saccades coinciding with the peak velocity of vergence were higher for tinnitus subjects. Yet, saccades did not increase peak velocity of vergence for tinnitus subjects, but they did so for controls. 4) In contrast, there was no significant difference of vergence latency between these two groups.

Interpretation: The results suggest dysfunction of vergence areas involving cortical-brainstem-cerebellar circuits. We hypothesize that central auditory dysfunction related to tinnitus percept could trigger mild cerebellar-brainstem dysfunction or that tinnitus and vergence dysfunction could both be manifestations of mild cortical-brainstem-cerebellar syndrome reflecting abnormal cross-modality interactions between vergence eye movements and auditory signals.

Figure 1. The Vergence eye movement task.

(A) Spatial arrangement for vergence: two diodes on an horizontal plane, one at 20 cm (required convergence 17°) and another 150 cm (required vergence 2.3°) from the subject's eyes. (B) Paradigm used for the stimulation: the fixation LED stayed on for a random period between 1.5 and 2 sec; the target LED was kept on for 1.5 sec; a black period of 500 ms was used for break.

Figure 2. Typical recordings of vergence eye movements.

Convergence and divergence with their corresponding velocity traces plotted at different scales are obtained by difference of the position signal between the two eyes (LE-RE); the arrows ‘i’ and ‘p’ indicate the onset and the offset of convergence, respectively; the dashed line indicates the required convergence change.

Figure 3. The latency of vergence eye movements.

Individual mean latency with standard deviation for convergence and divergence in control and tinnitus subjects.

Figure 4. The accuracy of vergence eye movements.

Individual mean of accuracy with standard deviation for convergence and divergence in control and tinnitus subjects.

Figure 5. The peak velocity of vergence eye movements.

Individual mean peak velocity with standard deviation for convergence and divergence in control and tinnitus subjects.

Figure 6. The duration of vergence eye movements.

The Individual mean duration with standard deviation for convergence and divergence in control and tinnitus subjects.

Figure 7. Group mean values of each parameter.

Latency (A), Gain (B), Peak velocity (C) and Duration (D) for convergence and divergence in control and tinnitus subjects. Asterisks show statistically significant difference between controls and tinnitus.

Figure 8. Embedded saccades.

(A) Group mean number of embedded saccades during complete execution of the vergence–total trajectory seen example in Fig. 2 from ‘i’ to ‘e’. (B) the group mean amplitude of the subgroup of saccades coinciding with the peak velocity of vergence. Tinnitus subjects show significantly more embedded saccades, and higher amplitude of coinciding saccades than controls (asterisks).

Figure 9. The correlation between the vergence and the coinciding saccades.

(A) The correlation between the peak velocity of vergence and the amplitude of saccades coinciding with the peak velocity of vergence; (B) the correlation between the peak velocity of vergence and the peak velocity of saccades coinciding with the peak velocity of vergence. Correlation coefficients ‘r’ and levels of significance are shown next to each cluster. Positive significant correlations for both divergence and convergence occur for controls but for tinnitus patients. Values in boxes indicate correlations for the subgroup of saccades with amplitudes equal or less than 4°; no significant correlation exists for controls or tinnitus.

Figure 10. The main sequence of vergence.

The correlation between the peak velocity and the amplitude of convergence (A) and of divergence (B) without coinciding saccades. All correlation coefficients and levels of significance are shown next to each cluster. Correlations are positive and significant, i.e. the peak velocity of vergence increases as the vergence amplitude increases for both controls and tinnitus subjects.

Figure 11. Group mean peak velocity of vergence.

With saccades coinciding with the peak velocity of vergence (A) and without such saccades (B). Tinnitus subjects show lower peak velocity of vergence than controls for all cases (asterisks). Also, for controls only, vergence velocity is higher with coinciding saccades than without such saccades.