Tinnitus has been the No. 1 disability at the Veteran Administration for the last 15 years, yet its interaction with hearing loss secondary to etiologies such as age, noise trauma, and traumatic brain injuries remains poorly characterized. Our objective was to analyze hearing loss and tinnitus, including audiogram data, of the Million Veteran Program within the context of military exposures in an aging population.
Health records, questionnaires, audiograms, and military data were aggregated for 758,005 Veteran participants in the Million Veteran Program 2011 to 2020, with relative risks (RR) calculated for ancestries, sex, hearing loss and military exposures such as combat, blast, and military era served. A multivariate model with significant demographic measures and exposures was then analyzed. Next, audiogram data stratified by sex were compared for those with and without tinnitus by two methods: first, mean thresholds at standard frequencies were compared to thresholds adjusted per ISO 7029:2000E age and sex formulae. Second, levels for those ≤40 years of age were compared with those 41 and older. Finally, a proportional hazards model was examined to ascertain the timing between the onset of tinnitus and hearing loss, calculated separately for electronic health record diagnoses (ICD) and self-report.
Tinnitus was either self-reported, diagnosed, or both in 37.5% (95% CI, 37.4 to 37.6), mean age 61.5 (95% CI, 61.4 to 61.5), range 18 to 112 years. Those with hearing loss were 4.15 times (95% CI, 4.12 to 4.15) as likely to have tinnitus. Americans of African descent were less likely to manifest tinnitus (RR 0.61, 95% CI, 0.60 to 0.61), as were women (RR 0.65, 95% CI, 0.64 to 0.65). A multivariate model indicated a higher RR of 1.73 for traumatic brain injury (95% CI, 1.71 to 1.73) and daily combat noise exposure (1.17, 95% CI, 1.14 to 1.17) than age (0.998, 95% CI, 0.997 to 0.998). Subjects ≤40 years of age had small but significantly elevated hearing thresholds through all standard frequencies compared to Veterans without tinnitus, and the effect of tinnitus on hearing thresholds diminished with age. In the hazard model, those >40 with new onset of tinnitus were at risk for hearing loss sooner and with greater incidence than those who were younger. The rate of hearing loss following tinnitus approached 100%. In contrast, only approximately 50% of those who self-reported hearing loss initially were at risk for later hearing loss, in contrast to ICD comparison, where those with ICD of hearing loss were more likely to sustain an ICD of tinnitus subsequently.
Evidence suggests that the occurrence of tinnitus in the military is more closely related to environmental exposures than to aging. The finding that tinnitus affects hearing frequencies across the audiogram spectrum suggests an acoustic injury independent of tonotopicity. Particularly for males >40, tinnitus may be a harbinger of audiologic damage predictive of later hearing loss.