Prevalence and predictors of problematic tinnitus

To determine the prevalence and characteristics of problematic tinnitus in a representative United States (US) sample. Cross-sectional study of the NHANES database from 2011–2012 and 2015–2016. 8029 records were included for analysis. The prevalence of any tinnitus was 16.32%. Only 1.38% reported tinnitus as a big problem or worse. Univariate analysis revealed a statistically significant relationship between tinnitus constancy and duration with severity, where more constant perception and longer duration was related to increased severity (p < 0.0001). Multivariable models showed a clear relationship of increased odds ratio (OR) of problematic tinnitus with constancy of tinnitus but not duration. Furthermore, multivariable models showed higher Body Mass Index (OR= 1.013, 95% CI: 0.99–1.03), sleep trouble (OR = 2.016, 95% CI: 1.52–2.66) and comorbidities (OR = 1.43, 95%CI: 1.06–1.94) were all associated with increased problematic tinnitus, but not non-problematic tinnitus. Our study demonstrated that the prevalence of problematic tinnitus is much lower than the prevalence of any tinnitus. There was also a relationship between BMI, sleep trouble, other comorbidities and increased odds of problematic tinnitus. Tinnitus severity was significantly related to constancy of perception but had a less clear relationship to duration.