The talk emphasizes key take-home messages derived from results based on a Canadian dataset and the available quality evidence.
First, we report findings from a meta-analysis of Extended High Frequency Audiometry (EHFA) in adults with normal audiograms and chronic tinnitus. We found significantly elevated EHF thresholds in the tinnitus group compared with the control group, highlighting a degree of cochlear mechanical dysfunction that is not reflected in conventional audiometry. This suggests that establishing EHFA norms may enable earlier identification of individuals at risk for tinnitus and support preventive counseling strategies.
Second, we report the characteristics of tinnitus and mental health outcomes in adults with chronic tinnitus, comparing individuals with long-term noise exposure (LNE) to those without. Participants with LNE exhibited lower Uncomfortable Loudness Levels, higher tinnitus loudness and handicap, and greater self-reported anxiety, depression, and insomnia. These findings demonstrate that LNE exacerbates both tinnitus severity and associated mental health burden.
Third, we present artificial intelligence-driven insights from Machine Learning (ML)-based analyses of a large dataset of adults with age-related hearing loss (ARHL) or occupational noise-induced hearing loss (NIHL). LNE was associated with accelerated auditory aging by approximately two decades. Constant (28.0% vs. 9.0%) and intermittent (19.0% vs. 11%) tinnitus were more prevalent in NIHL compared with ARHL. ML-based models predicted tinnitus with 70.0% accuracy and differentiated NIHL from ARHL with 82.0% accuracy, highlighting the potential of AI to enhance early prediction and diagnostic precision in audiologic practice.
We conclude by summarizing the key research findings and take-home messages.
Keywords
- chronic tinnitus
- auditory aging
- occupational noise exposure
- extended high-frequency audiometry
- machine learning