Impaired cortical gap-mediated inhibition in chronic tinnitus but with limited individual diagnostic accuracy

Tinnitus lacks an objective marker, limiting diagnosis and treatment evaluation. We tested whether gap-mediated inhibition of cortical sound-evoked responses provides one. In a prospective, multicentre case-control study at seven sites in the United States and Europe, adults with and without chronic tinnitus heard brief pulses preceded or not by a silent gap, embedded in broadband or narrowband carriers, during electroencephalography. The primary outcome was inhibition of the N100 response. Of 584 eligible participants, 455 yielded high-quality recordings with complete covariate data (232 cases, 223 controls; mean age 45.4 years, 52% women) that entered the adjusted analyses. Inhibition was strongest for broadband noise whereas narrowband carriers gave limited dynamic range. Participants with tinnitus showed reduced broadband gap-mediated inhibition (β = -19.27%, d = -0.413, cluster-corrected p = 0.002), scaling with tinnitus severity (covariate-adjusted √THI model: partial R 2 = 0.049, p = 0.0015). In contrast, discrimination of individuals was limited (AUC 0.612, 95% CI 0.560-0.663). Cortical gap inhibition is a reproducible group-level effect rather than an individual diagnostic test.