Multisensory attention training for treatment of tinnitus

Abstract

Tinnitus is the conscious perception of sound with no physical sound source. Some models of tinnitus pathophysiology suggest that networks associated with attention, memory, distress and multisensory experience are involved in tinnitus perception. The aim of this study was to evaluate whether a multisensory attention training paradigm which used audio, visual, and somatosensory stimulation would reduce tinnitus. Eighteen participants with predominantly unilateral chronic tinnitus were randomized between two groups receiving 20 daily sessions of either integration (attempting to reduce salience to tinnitus by binding with multisensory stimuli) or attention diversion (multisensory stimuli opposite side to tinnitus) training. The training resulted in small but statistically significant reductions in Tinnitus Functional Index and Tinnitus Severity Numeric Scale scores and improved attentional abilities. No statistically significant improvements in tinnitus were found between the training groups. This study demonstrated that a short period of multisensory attention training reduced unilateral tinnitus, but directing attention toward or away from the tinnitus side did not differentiate this effect.

Figure 1. Multisensory attention training; an example of a participant with left-dominant tinnitus.

In this case, the change of the centre dot into a diamond always indicates a saccade to the left dot and a change into a square indicates a saccade to the right dot. Each trial started with fixation of the centre dot (a). In the integration group (b), the diamond cue is accompanied by a flash of the left dot, an auditory stimulus delivered to the left ear and vibration administered on the left temple (tinnitus side). The square cue is not accompanied with lateralized stimuli. In the attention diversion group (c), the diamond is presented alone, whereas the square is associated with flashing of the right dot and tactile and auditory stimuli presented to the right side (non-tinnitus side). Participants were also instructed to press the left or right Ctrl-key on a computer keyboard according to the direction of their saccade. Each trial terminated when the lateralized dots became hollow which cued the participant to re-fixate on the centre dot for the next trial.

Figure 2. The effect of multisensory training on the TFI.

Panel (a) shows mean TFI scores from the initial auditory examination (Baseline), before (Pre), after (Post) and three weeks after (Post 3w) the multisensory training. ** p < 0.01, * = p < 0.05. Error bars represent ± SEM. Panel (b) shows individual’s change in TFI from Pre to Post training. The dashed line represents the proposed level of clinically meaningful change in TFI.

Figure 3. The effect of multisensory treatment on secondary outcome measures.

Panel (a) shows mean TSNS scores before (Pre), after (Post) and three weeks after (Post 3w) the multisensory training. Panel (b) shows the mean saccadic latency of correctly initiated saccades towards tinnitus dominant (white circles) and non-dominant (black circles) sides. Panels (c) and (d) show the mean % change in hits and reaction time (RT) after the multisensory treatment. The dashed line represents the magnitude of procedural learning on the CAB. Error bars represent +/- SEM. * p < 0.05, ** p < 0.01. AVMT-A=Auditory-Visual Multiprocessing Tests – Auditory sub-type, AVMT-M = Auditory-Visual Multiprocessing Tests – Mixed sub-type, AVMT-V=Auditory-Visual Multiprocessing Tests – Visual sub-type, DRT-A=Discriminate Reaction Time – Auditory sub-type, DRT-M=Discriminate Reaction Time – Visual sub-type, DRT-M = Discriminate Reaction Time – Mixed sub-type. Error bars represent +/- SEM. * p < 0.05, ** p < 0.01.

Figure 4. The effect of multisensory training on saccade error rates. The percentage of correct saccades is shown before (Pre) and after (Post) the multisensory training for the integration.

(a) and attention diversion groups (b). Filled circles represent saccades directed towards the tinnitus side. Open circles represent saccades towards the non-tinnitus side. Error bars represent ± SEM. * indicates p < 0.05.

Figure 5. The effect of multisensory training on tinnitus characteristics.

The solid lines in panels (a) and (b) represent a mean change (pooled for both groups) in tinnitus pitch and loudness, respectively, as a function of days of training. The dotted lines represent a ± 1 SEM range.