Objective diagnostics of tinnitus

Abstract

Background: Tinnitus is the phantom perception of sound reported by around 14% of the population. For over 120 million people worldwide tinnitus is perceived as a major problem. Still, tinnitus is a heterogenous condition with no stratified sub-types or biomarkers for objective assessment. This severely limits the potential for development and evaluation of novel therapies for tinnitus. This thesis aims to I) investigate the relationship between conditions commonly reported by tinnitus patients and II) evaluate the potential of electrophysiology or magnetoencephalography (MEG) to function as an objective biomarker for tinnitus.

Methods: Studies I-III retrospectively analyzed questionnaire data, in total n = 5 593, collected in the Swedish Tinnitus Outreach Project (STOP) between November 2015 and January 2018. Multivariate logistic regression models were implemented to investigate the association between tinnitus and related conditions - temporomandibular joint (TMJ) pains, hyperacusis and headaches. Study IV used longitudinal data from The Swedish Longitudinal Occupational Survey of Health (SLOSH), in total 20 439 participants, with 53 273 observations. The transition from occasional to constant tinnitus was investigated using Generalized Estimating Equation (GEE) models. The second part of Study IV used ABR data from STOP (n = 405) to evaluate measurements of wave I, III & V amplitude and latency in distinguishing constant from occasional tinnitus, or non-tinnitus controls. Study V recruited n = 22 normal hearing, non-tinnitus participants for optimization of a GPIAS (Gap Pre-pulse Inhibition of Acoustic Startle) protocol for MEG. In this exploratory study sound pulses of 20 ms were presented in 60 or 70 dBA carrier noise with a 50 ms silent gap preceding the pulse by 240, 120, 60 or 0 ms. All MEG were recorded by the Elekta Neuromag TRIUX 306-channel system at NatMEG, Karolinska Institutet.

Results: TMJ complaints increased to over 30% among those with severe tinnitus, compared to 19% in all participants with tinnitus. For headache, adjusted odds ratios (95% confidence interval) showed an association of OR: 3.8 (2.4-5.9) and a strong association with hyperacusis of OR: 12.1 (7.1-20.6) was found for those with severe tinnitus. Longitudinal analysis indicated that tinnitus progresses towards constant tinnitus and that once established is very unlikely to remit. Changes in the ABR response, particularly wave V latency, distinguished constant from occasional tinnitus and non-tinnitus controls, likely reflecting plastic changes related to this chronification. A GPIAS stimulation protocol with an inter-stimulus interval of 240 ms between silent gap and 90 dBA pulse produced N1-inhibition of ERF responses with much lower variability when compared to traditional EOG responses.

Conclusion: We identified TMJ complaints and hyperacusis as important factors to consider in future studies of tinnitus. ABR wave V latency can distinguish constant tinnitus from occasional or non-tinnitus at a group level, but is likely not sensitive enough for individual diagnostics. Instead, GPIAS together with MEG is a promising approach to developing a biomarker for tinnitus.