Am J Audiol. 2025 Jun 13:1-14. doi: 10.1044/2025_AJA-25-00027. Online ahead of print.
ABSTRACT
PURPOSE: Numerous individuals in the United States are bothered enough by tinnitus that it affects normal daily activities, including sleep and concentration. There is a high prevalence of self-reported bothersome tinnitus in the U.S. military, and therefore, it is important to assess the impact of tinnitus on functional performance. The primary aim of this study was to examine the effects of tinnitus on a range of subjective, objective, auditory, and cognitive measures. A secondary aim was to determine which factors best predict whether individuals ever self-reported tinnitus. A prospective study was conducted to address the primary aim, and an exploratory machine learning approach was used to address the second.
METHOD: The study included 463 active duty U.S. Service members being seen for annual hearing surveillance. Participants completed several auditory and cognitive tests, along with questionnaires regarding tinnitus, depression, anxiety, and sleep difficulty.
RESULTS: The primary aim found no significant performance differences on tasks of selective attention, short-term memory, or speech in noise between the tinnitus groups. However, there were significant group differences in extended high-frequency hearing. A machine learning algorithm was used to predict whether an individual self-reported experiencing tinnitus. The model had approximately 80% accuracy, excellent sensitivity, and modest specificity.
CONCLUSIONS: Study findings corroborated other evidence that extended high-frequency thresholds are poorer in those who reported ever experiencing tinnitus, regardless of tinnitus bother. Our results suggest that combining results from audiometric testing, high-frequency hearing, and subjective assessment of sleep achieved 80% accuracy in predicting tinnitus presence.
PMID:40512970 | DOI:10.1044/2025_AJA-25-00027
Recent Comments