(1) To present hearing threshold data from a recent nationally representative survey in the United States (National Health and Nutrition Examination Survey, 1999–2004) in a distributional format that might be appropriate to replace Annex B in international (ISO-1999) and national (ANSI S3.44) standards and (2) to compare these recent data with older survey data (National Health Examination Survey I, 1959–1962) on which the current Annex B is based.
Better-ear threshold distributions (selected percentiles and their confidence intervals) were estimated using linear interpolation. The 95% confidence intervals for the medians for the two surveys were compared graphically for each of the four age groups and for both men and women. In addition, we calculated odds ratios comparing the prevalences of better-ear hearing impairment (thresholds > 25 dB HL) between the two surveys, for 500, 1000, 2000, and 4000 Hz, and for their four-frequency average.
Across age and sex groups, median thresholds were lower (better) in the 1999–2004 survey at 500, 3000, 4000, and 6000 Hz (8000 Hz was not tested in the 1959–1962 survey). For both men and women, the prevalence of hearing impairment was significantly lower in 1999–2004 at 500, 2000, and 4000 Hz, but not at 1000 Hz.
For men and women of a specific age, high-frequency hearing thresholds were lower (better) in 1999–2004 than in 1959–1962. The prevalences of hearing impairment were also lower in the recent survey. Differences seen at 500 Hz may be attributable at least in part to changes in standards for ambient noise in audiometry. The National Health and Nutrition Examination Survey 1999–2004 distributions are offered as a possible replacement for Annex B in ISO-1999 and ANSI S3.44.