This study showed that day-shift work with continuous 8-h noise exposure in farm machinery factory workers produced higher level of hearing loss at 2, 3, and 4 kHz in both ears than nonstandard shift work in the comparison firefighters group. This finding was statistically significant after a 10-year work history in the stratified analysis, though duration was not significant in the multivariate model including age.
The most obvious explanation for the difference in risk is differential noise exposure. While we have documented exposures among farm factory workers in excess of 85 dBA, we do not have data to make a direct comparison with firefighters. However, a previous study which evaluated noise exposure among Korean firefighters according to their time-dependent activity patterns reported that firefighters were exposed to a mean noise level of 76 dBA during work time . This result was similar to another study of 16 firefighters by Bryan et al. which reported a mean of noise exposure level of 78.7 dBA . Based on the data available, the mean level of noise exposure for both groups evaluated may be reasonably comparable, but the pattern and duration of continuous noise exposure is likely very different.
The exposure pattern of the two groups differed in that farm machine factory workers were continuously exposed to a relatively constant level of noise during work time and the noise exposure of the firefighters usually varied by activity. Lee reported that firefighters spent 67% of total work time in "inside" areas at the fire station, including offices and waiting rooms and 23% of their time outside the station attending to fires and emergencies or transporting to and from these events. The noise exposure levels for inside and outside areas were 65-72 dBA and 79-85 dBA respectively . The data we collected via questionnaire were consistent with these findings. Firefighters included in this study spent 20-30% of total work time in "outside" areas. Of this time, a portion was spent on fire trucks being transported to fires or other emergencies. These trips occurred approximately 10 times per duty day and lasted between 10-20 min per trip. The sources of noise exposure in the outside areas were fire engines, horns/sirens and pumps. Several noise surveys conducted by NIOSH to determine the magnitude of noise exposures among U.S. firefighters found that exposure levels varied from low to intense exposures according to OSHA or NIOSH noise criteria . Firefighters traveling in emergency vehicles were exposed to noise ranging from 103.4 to 114.5 dBA. Mechanical equipment used by firefighters can produce up to 115 dBA with a mean duration of 30 min [14, 15]. Though these are very high levels of exposure (over 90 dBA), the duration of this exposure was less than 10% of total work time for the firefighter group. Globally, firefighter's exposure to noise was intermittent in both intensity and duration, with both factors dependent on emergency codes during their shift. The range of noise exposure varied from 65-115 dBA. This intermittent exposure pattern is very different from the continuous noise at a relatively fixed level experienced by the farm machine factory workers. It may be that continuous noise exposure carries a greater risk of hearing loss than intermittent exposure even if the mean range in dBA is similar.
Shift schedules also differ between the groups. The shift length of farm machine factory workers was 8 h per day, 5 days per week, so non-exposure time was 16 h per day with up to 63 continuous hours of exposure-free time on weekends. Firefighters were free from noise exposure for 9 h during night-work duty, twice a week. Noise free time for the twice weekly day-shift was 15 h and up to 48 h of noise free time during weekends. Insufficient time between work shifts to allow workers to recover from temporary hearing deficits may affect hearing level as temporary threshold shifts generally last 24 h or more after cessation of excessive noise exposure for employees who work regularly . Other previous studies  report that hearing loss is entirely preventable by administrative controls such as periodic shift rotation and limiting exposure to noise when levels exceed 85 dBA. We replicate these findings in that farm machine factory workers who were not given sufficient time to recover from temporary threshold shift experienced a higher level of permanent threshold shift (hearing loss) than firefighters exposed to similar levels of occupational noise. The duration of non-exposure periods related to shift type may be a contributing factor to this discrepancy, though to be sure, studies controlling for other known risk factors for hearing loss and measuring both noise magnitude and duration as well as hearing loss both groups in an identical fashion, would have to be carried out. Clark and Bohl evaluated hearing loss in firefighters compared with age-matched, non-occupationally exposed groups of individuals and reported that firefighters are not at risk for occupational noise-induced hearing loss, even though they work nonstandard shifts and are occasionally exposed to high levels of noise . Our results differ slightly in that firefighters who had worked 20 years or more showed statistically significant hearing loss compared to other subgroups of firefighters when age and other risk factors were controlled for, particularly at 4 kHz (Data not shown). A possible explanation is that firefighters in Korea do not generally wear hearing protection despite frequent exposure to noise levels over 90 dBA. Over the long term this may inflict hearing loss though it would not be detected in shorter term studies or studies which did not consider long duration separately. Occupational exposure to high heat at fires may also impact noise-induced hearing loss . A NIOSH investigation reported that health hazards exist for firefighters and recommended steps to the department to reduce noise exposure to help prevent further hearing loss . In the case of farm machine factory workers, mean hearing levels were over 25 dB despite being exposed to a mean noise level below the current accepted threshold of 85 dBA according to environmental noise exposure monitoring. This finding is consistent with previous studies on chronic exposure to moderately high amounts of occupational noise. Rabinowitz et al. reported that the majority of 10 dB standard threshold shifts occurred in workers whose calculated mean ambient noise exposures were less than or equal to 85 dBA . This may be partially explained by the greater number of individuals employed in environments with noise levels below 85 dBA as well as a decreased likelihood of hearing protection at noise levels not deemed to be dangerous . Hearing loss in farm machine factory workers appears to increase with duration of exposure. Both groups experienced work related hearing loss with a duration of work longer than 20 years though the factory workers showed a greater degree of hearing loss than the firefighters at this time point.
There are several limitations to this study beyond the lack of concurrent exposure assessment of the firefighter group. Some factors which have been shown in the literature to affect noise-induced hearing loss such as alcohol consumption and the use of organic solvents were not controlled for. Although firefighters are exposed to mixed organic solvents during fire suppression , the exposure time is irregular and short, approximately 30 min per duty day, and fell below the NIOSH recommended occupational hazard threshold. Exposure to heat experienced during fire suppression may also be a significant risk factor for noise-induced hearing loss  and we did not control for this in our study. Only three farm machine factory workers reported a work period less than 10-years. Therefore, a much larger sample would be necessary to carry out a robust analysis of the effects of duration on hearing loss in employees with a work history of less than ten years. This merits investigation because there is evidence that there may be an initial, relatively rapid, phase of hearing loss, followed by a leveling off . Previous research noted that 20% of firefighter audiograms showed threshold losses of 40-60 dB in hearing 3, 4, and 6 kHz test frequencies in one or both ears and 14% with still greater losses . We used the average of hearing levels at 2, 3 and 4 kHz instead of high frequencies, which is the OSHA "recordable hearing loss" case definition and the standard metric of hearing loss progression.
Neither group showed evidence of decline in hearing during the 4 years of continuous observation. It is likely that the farm machinery factory workers were compliant in wearing hearing protection. The farm machinery factory hearing conservation program was formally monitored and the reported rates of adherence were approximately 85% over the study period. This is consistent with the literature which has noted a higher rate of hearing protection use in noisy industries . The firefighters were noted to have very low rates of hearing protection use, so the absence of measured progression likely reflects either low exposure or the value of recovery from exposure between shifts.
A final important limitation which the above highlights is that the losses of interest for our study occurred in both populations largely before the observation began and we did not have sufficient data to control for previous occupational noise exposure. Though it is likely that both populations experienced considerable job stability due to the nature of the work environment in Korea and the size and stability of these employers, it does not follow that the level of noise exposure was consistent as assigned tasks may have changed during the course of employment. That having been said, it is certainly likely that hearing protection was used less regularly in the farm machinery workers before the hearing conservation program began, but the impact of this cannot be directly tested.