Is the posterior cingulate cortex an on-off switch for tinnitus?: A comparison between hearing loss subjects with and without tinnitus.

As the human brain works in a Bayesian manner to minimize uncertainty toward external stimuli, the deafferented brain may generate tinnitus in an attempt to fill in missing auditory information, e.g. due to hearing loss. However, not everybody with hearing loss develops tinnitus. Understanding the differences between people with hearing loss who develop tinnitus versus those who do not offers a unique opportunity to unravel critical brain areas involved in the generation of a phantom sound. In this study, we compared resting-state quantitative electroencephalography between hearing loss patients with (HL-T) and without tinnitus (HL-NT) to identify cortical oscillatory signatures that may reveal prerequisites for the selective development of tinnitus in subjects with hearing loss. We enrolled 65 subjects with HL-NT and 65 subjects with HL-T whose tinnitus handicap inventory scores were <16 (grade 1) to minimize the bias induced by distress-induced cortical activity changes. Subjects in the HL-T and HL-NT groups were matched in terms of the bilateral hearing threshold (0.25-8 kHz) using nearest neighbor method. Compared to the HL-NT group, the HL-T group showed significantly higher activity in the right parahippocampus for the beta 1 frequency band, in the left inferior parietal lobule (IPL) for the beta 2 frequency band, and in the right IPL for the beta 3- and gamma frequency bands. Functional connectivity analyses revealed that the HL-T group had significantly higher connectivity than the HL-NT group between both parahippocampal gyri and the right IPL for the delta frequency band, and between the left posterior cingulate cortex (PCC) and right IPL for the beta 2 frequency band. These results suggest that tinnitus may be perceived only if auditory memory stored in the parahippocampus is actively linked to the IPL-based "circuit breaker" system and the circuit breaker signal is connected to the PCC-based default mode network (DMN). Thus, when the circuit breaker system regards tinnitus secondary to peripheral deafferentation as a salient event and then the DMN regards tinnitus as a norm, subjects with hearing loss may consciously perceive tinnitus. The results of this study further refine the recently proposed Bayesian model and decipher the neurobiological mechanism of the selective development of tinnitus in subjects with hearing loss.