Tracking population health based on self-reported impairments: Trends in the prevalence of hearing loss in US adults, 1976-2006.

Trends in the prevalence of hearing loss among US adults remain ambiguous because of variation across surveys in question wording and limited use of audiometric examinations. Pooling samples of participants aged 20-69 years in 4 nationally representative cross-sectional survey series conducted from 1976 to 2006 (N = 990,609), the authors performed logistic regression to quantify self-reporting biases compared with audiometric measurements. Statistically significant underreporting or overreporting of hearing loss was observed, with various patterns of bias across age groups and surveys. Substantial upward reporting biases appeared among young adults in the National Health and Nutrition Examination Survey since 1999 and in the National Health Interview Survey since 1997. Trends in age-standardized prevalence of bilateral hearing loss were estimated with corrections for self-reporting biases. Prevalence in men shifted from 9.6% (95% confidence interval (CI): 7.8, 11.8) in 1978 to 12.2% (95% CI: 10.1, 14.7) in 1993 and declined to 8.1% (95% CI: 7.0, 9.5) in 2000. In women, prevalence was relatively constant at approximately 6%-7% until the early 1990s and decreased from 7.0% (95% CI: 5.5, 9.1) in 1993 to 4.2% (95% CI: 3.4, 5.3) in 2000. Prevalence was stable in both sexes in the early 2000s. This approach to adjust for biases in self-reported impairments by using measured performance may be useful in various health domains.