Noise-induced permanent threshold shifts in the occupational noise and hearing survey: An explanation for elevated risk estimates

OBJECTIVES: To (1) estimate noise-induced permanent threshold shift (NIPTS) values from the data of the 1968-1972 Occupational Noise and Hearing Survey (ONHS); (2) compare these NIPTS estimates to NIPTS predictions from an international standard (ISO-1999); (3) determine why excess risk estimates based on the ONHS are so much higher than those based on ISO-1999. DESIGN: The ONHS raw data were acquired from the National Institute for Occupational Safety and Health. Binaural average thresholds from 0.5 to 6 kHz were calculated for each of 1291 noise-exposed subjects (80 to 94 dBA, for up to 30 yr, all tested just before their shifts) and 665 non-noise-exposed control subjects (mostly office workers, tested throughout the work day). "Screened" subjects had had no prior significant noise exposure and had no evidence of other ear disease, whereas "excluded" subjects had failed one or more screening criteria. Twenty exposure groups were created (based on exposure level, exposure duration, and screened versus excluded status) and compared with 20 age-matched control groups. The median difference statistic estimated median NIPTS. RESULTS: Median NIPTS estimates in the 3 to 6 kHz region were generally consistent with the NIPTS predictions of ISO-1999. At lower frequencies, especially at 0.5 and 1 kHz, the ONHS estimates were significantly larger than the ISO-1999 predictions, even for exposures below 90 dBA, but these differences did not increase systematically with exposure level and duration. CONCLUSIONS: High-frequency median NIPTS estimates from ONHS data are consistent with the predictions of ISO-1999. Differences between exposed and control low-frequency thresholds in the ONHS are higher than predicted by ISO-1999, but these differences probably are related more to socioeconomic or test procedure effects than to occupational noise exposure. These low-frequency effects explain why excess risk estimates from the ONHS are higher than those based on ISO-1999.