Biomarkers of Subjective Tinnitus Presence and Clinical Burden: A Systematic Review

Hear Res. 2026 Aug 7;481:109770. doi: 10.1016/j.heares.2026.109770. Online ahead of print.

ABSTRACT

Despite being common and clinically burdensome, subjective tinnitus is diagnosed and characterised primarily through subjective reporting. Objective biomarkers could transform tinnitus care by enabling reproducible identification, stratification, and monitoring, but their clinical utility for detecting tinnitus and quantifying its severity remains unclear. We conducted a PRISMA-guided systematic review of tinnitus biomarkers, critically evaluating their clinical and diagnostic relevance, methodological quality, and readiness for clinical translation. PubMed, Scopus, and Web of Science were searched for human studies reporting objective measures linked to tinnitus presence, severity, or diagnostic accuracy. Risk of bias was assessed using Joanna Briggs Institute tools, and results were synthesised narratively due to heterogeneity. We included 69 studies spanning neurophysiological, neuroimaging, cognitive-behavioural, and molecular approaches. Although many studies reported tinnitus-related group differences, convergence across studies was limited. Common limitations included small cohorts, inconsistent phenotyping, heterogenous methodologies, and inadequate adjustment for major confounders, especially hearing loss. Cognitive-behavioural paradigms, particularly attention-modulated cortical responses, showed the strongest and most consistent association with tinnitus burden and the greatest translational potential. Molecular biomarkers, notably metabolomics, showed early promise but currently lack replication and external validation. Neurophysiological and neuroimaging markers provided mechanistic insights but were constrained by reproducibility and feasibility for routine clinical use. In summary, tinnitus biomarkers are advancing, but no single measure currently supports stand-alone clinical diagnosis. Progress will require harmonised methods, large well-characterised samples, severity-stratified designs, and validated multimodal panels aligned to mechanistic models. Systematic review registration: PROSPERO (CRD420261449639).

PMID:42612615 | DOI:10.1016/j.heares.2026.109770