This retrospective cross-sectional study of median hearing threshold levels of males employed in two specific occupations shows that the trend with decreasing noise-induced hearing loss in Sweden during the 1970s and 1980s continues into the 1990s. In the occupational categories mechanical work and wood processing men in age groups 30-39, 40-49 and 50-59 years old examined during the time period 1971-76, 1981-86 and 1991-96 were compared. Possible explanations to the improvement might be a wider use of hearing protectors at work and less exposure to noise during military service. The results show that the awareness of noise-induced occupational hearing loss has improved but the hearing conservation programs are still necessary as hearing threshold levels in these occupational groups continue to be poorer than expected in relation to age.
Keywords: Noise-induced hearing loss, hearing threshold level, hearing conservation, database, hearing protectors, occupational noise exposure
|How to cite this article:|
Johansson M, Arlinger S. The development of noise-induced hearing loss in the Swedish County of Ostergotland in the 1980s and 1990s. Noise Health 2001;3:15-28
|How to cite this URL:|
Johansson M, Arlinger S. The development of noise-induced hearing loss in the Swedish County of Ostergotland in the 1980s and 1990s. Noise Health [serial online] 2001 [cited 2020 Aug 3];3:15-28. Available from: http://www.noiseandhealth.org/text.asp?2001/3/10/15/31761
| Introduction|| |
In a developed country more than one third of the hearing impairments are partly caused by excessive noise exposure (Smith, 1998). In many countries the most prevalent irreversible industrial disease is noise-induced hearing loss as reported at a WHO meeting in 1997 on the prevention of deafness and hearing impairment relating to noise-induced hearing loss (Smith, 1998). According to the National Board of Occupational Safety and Health in Sweden (1999) noise is the third most common cause of occupational disease for men in Sweden. In a review of hearing conservation practices Barrenas et al. (1998) have reported that the decrease in prevalence of noise-induced hearing loss in Sweden during the 1970s and 1980s has stopped. One explanation to this according to Barrenas et al. is the reduced financial support from the state to the occupational health centres.
Previous studies of the population in the Swedish County of Ostergotland by Arlinger & Ivarsson (1985) and Ivarsson & Arlinger (1985) have shown an improvement in hearing threshold levels (HTLs) and proportion of normal audiograms for defined age groups from the early 1970s to the late 1970s and the early 1980s. Different occupational categories showed different degrees of improvement. These results were based on a database of HTLs established in the early 1970s. The database is still in use and now provides the opportunity to evaluate the development of hearing threshold levels during three decades.
The main aim of this study is to investigate the change over time in HTLs among employees in different occupations. The secondary aim is to compare the results to available reference material of normal hearing. The question is: has the improvement of hearing threshold levels during the 1970s and 1980s stopped during the 1990s?
| Methods|| |
This is a retrospective cross-sectional comparison between the 1970s, 1980s and the 1990s. The audiograms were selected from the existing HTL database in the county of Ostergotland in Sweden. The selection criteria were:
- Category of occupation: mechanical work, wood processing
- Age at test: 30-39; 40-49; 50-59 years
- Time period for audiometry: 01.01.71 to 31.12.76; 01.01.81 to 31.12.86; 01.01.91 to 31.12.96
- Noise exposure: >2 hours/day
- Gender: male
The category of mechanical work includes e.g. engineering workshop, car repair shop, and forging. Wood processing includes e.g. sawmill, carpentry, and forestry.
According to subjective assessment the subjects had been exposed to noise at least two hours a day at their present work. This criterion excludes men who belong to the occupational categories mentioned but work in low noise areas.
The dependent variable is hearing threshold level by air conduction. The test frequencies 2000, 3000, 4000, 6000 and 8000 Hz for the right and the left ears are included in the descriptive statistics, but in the statistical analysis the included frequencies are limited to 4000, 6000 and 8000 Hz for right and left ears. The frequencies 4000 and 6000 Hz best reflect the noise-induced hearing loss and 8000 Hz is included to distinguish between age-related sensoneural hearing loss and noise-induced hearing loss.
Screening audiometry was performed according to procedure, which agree with those specified in ISO 8253-1 (1989). Most of the earlier examinations are screening measurements at a screening level of 20 dB HL combined with HTL determinations when the screening level was inaudible. Later on screening levels of 10 dB HL or 0 dB HL have also been used [Table - 1].
The data have been collected over 26 years and audiologically trained nurses performed the audiometric tests. Hearing threshold levels were measured either at different occupational health centres or in a mobile unit equipped with a soundproof booth and a pure-tone audiometer. Different audiometers, all equipped with Telephonics TDH-39 or TDH-49 earphones, have been used at different occupational health centres. All audiometers have been calibrated according to ISO 389-1 (1998) or relevant earlier editions annually. Questions regarding noise exposure and use of hearing protection at work, noise exposure and use of hearing protection during military service and subjectively worse hearing were included in the examinations.
The data are summarised as median, 1st and 3rd quartile values for grouped distributions according to Winkler and Hays (1975). When the values are given in a grouped distribution with class width of 5 dB, as in audiometry according to ISO 8253-1 (1989), this method may be used to determine an interpolated median (or quartile) within a class. The value determined is based on the proportion of observations between the median and the lower limit of the class and the total number of observations in the median class.
where n=sample size, cf=cumulative frequency, f=frequency. To determine the 1st and the 3rd
The results from different occupations, age groups, test frequencies and ears have been compared over the three time periods for audiometry. The difference between hearing threshold levels at 6000 Hz and 8000 Hz has been investigated in order to separate noiseinduced hearing loss from presbyacusis. The left and the right ears have been compared to assess any ear differences.
To study differences between the three time periods Kruskal-Wallis test was used. When a significant difference (p<0.01) occurred, the same test with Bonferroni correction was used as post hoc test to compare two groups at a time. Answers to the questionnaire were compared with chi 2 test (p<0.01) and post hoc tests were corrected according to Bonferroni.
If data from one person occurred more than once in the same time period mean values of HTLs from at most 8 test occasions were determined. These mean values were then used to determine median and quartiles and to compare the groups.
A total of 15,058 audiograms were included in the statistical analysis unequally distributed over 18 groups [Table - 2]a,b. In the groups investigated during the 1970s and the 1980s about 7% of the subjects occur in both time periods. About 15% of the subjects during the 1980s and the 1990s occur in both time periods.
In order to calculate reference data, ISO 7029 (1984) for males was used to obtain median, 10th and 90th percentile. The hearing threshold levels of this highly screened otologically normal population have been corrected according to Passchier-Vermeer (1993) to estimate the corresponding values for an otologically unscreened population. The 10th percentile is increased with 6 dB, the median is increased with 2 dB and the 90th percentile is kept unchanged. The 1st and 3rd quartiles for the otologically unscreened population are determined from the 10th and 90th percentiles according to two different normal distribution curves. This set of reference data was preferred to the example of database B given in ISO 1999 (1990), since we have previously found (unpublished) the ISO 1999 (1990) database B to overestimate HTLs in corresponding Swedish populations.
To estimate the number of normal hearing subjects a classification according to Klockhoff et al. (1974) was used. Normal hearing was defined as HTLs less than or equal to 30 dB HL at 500 Hz and less than or equal to 25 dB HL at 1, 2, 3, 4 and 6 kHz for both the left and the right ear. An estimation of the prevalence of normal hearing in the reference data according to this definition was based on the fact that 6 kHz is the test frequency showing the poorest hearing thresholds. Thus, we have simply looked at the distribution of hearing threshold levels at 6 kHz, corrected according to Passchier-Vermeer (1993) and assumed this to represent the prevalence of normal and abnormal hearing as defined by Klockhoff et al. (1974).
| Results|| |
Improvement over time
The group of mechanical workers shows a continuing improvement of hearing threshold levels during the 1980s and 1990s for all three age-categories [Figure - 1]a-f. The group working with wood processing shows an improvement from the 1980s to the 1990s in the youngest age group of 30-39 years. The age group 40-49 years only improves significantly (p<0.01) at 4 kHz, but there is an improvement at 6 kHz from the 1970s to the 1990s with p<0.05. The age group 50-59 years only improves at 4 kHz from the 1970s to the 1990s with p<0.05.
The statistically significant average decrease in median hearing threshold levels at 4 and 6 kHz from the 1980s to the 1990s in the youngest age group is 1 to 3 dB. The improvement is greater in the category with wood processing work than mechanical work. In the age group 40-49 the average decrease is 3 to 6 dB. In the oldest age group the average decrease is 5 to 8 dB. The median hearing threshold levels at 4 kHz and 6 kHz in the 1990s do not differ significantly between wood processing workers and mechanical workers in any age group.
There are also improvements at the frequency 8 kHz. Both among mechanical workers and wood processing workers the youngest age group of 30-39 has improved by 1 to 3 dB. In the age groups 40-49 and 50-59 years only the mechanical workers have changed significantly from the 1980s to the 1990s. The 40-49 year olds improve by between 3 and 4 dB and the 50-59 year olds by between 5 and 9 dB. There is no significant difference between wood processing workers and mechanical workers at 8 kHz in any age group in the 1990s.
Frequency and ear differences
All median threshold levels are significantly better at 8 kHz than at 6 kHz in the group of mechanical workers. The largest difference in the 1970s was about 10 dB in the group of 40-49 years. In the 1980s the largest difference was about 7 dB and in the 1990s about 5 dB in the same age group. In the group working with wood processing the result is the same during the 70s and 80s but there is no significant difference between 6 and 8 kHz in the 1990s.
There is also a difference between the left and the right ear in many groups. For all groups with such a difference the median threshold level is poorer for the left ear than for the right ear. The largest differences in the 1980s were about 5 dB at 6 kHz and 7 dB at 8 kHz. In the 1990s the largest differences were about 3 dB at 6 kHz and 4 dB at 8 kHz.
The results from the classification according to Klockhoff et al. (1974) shows that the percentage of normal hearing subjects increases over time [Figure - 2]a,b. In the 1990s the percentage of mechanical workers classified as normal hearing on both ears was 66% among the 30 to 39 year age group, 34% among 40 to 49 years and 13% among 50 to 59 years. In the group of wood workers the figure was 57% among 30 to 39 years, 28% among 40 to 49 years and 11% among 50 to 59 years. Expected percentages with normal hearing according to ISO 7029 (1984) corrected according to Passchier-Vermeer (1993) are 82% at an age of 39 years, 61% at an age of 49 years and 35% at an age of 59 years.
Hearing protectors at work
The analysis of the answers to the question about use of hearing protectors at work shows that among the mechanical workers in all three age groups there was a significant increase over time in use of hearing protectors [Figure - 3]a. In the group of wood processing workers, on the contrary, there was a decrease in the stated use of hearing protectors in the youngest and the oldest age groups from the 1980s to the 1990s [Figure - 3]b. In both the 1980s and the 1990s the percentage of hearing protection users differs between the age groups among wood workers. A significantly lower percentage of workers in the age group 50-59 years use protectors. Among mechanical workers there are no differences between the age groups. In the 1990s the percentage of hearing protection users in the different age groups does not differ between the occupational categories.
Noise exposure during military service
On the question regarding noise exposure during military service, 8742 subjects have answered that they were exposed. The percentage among the exposed answering that they used hearing protectors is shown in [Figure - 4]. The use of hearing protectors differs between men born before and after 1950. Data from subjects born before 1930 are missing.
Subjective decrease in hearing ability
Totally 8428 subjects have answered a question if they consider their hearing ability decreased since the previous hearing test. During the 1980s and the 1990s the proportion answering "yes" was larger in the older groups than in the younger, on average 8% among the 30 to 39 year olds, 14% among 40 to 49 year olds and 22% among 50 to 59 year olds. However, there is no significant difference in proportion answering "yes" between the 1970s, 1980s and 1990s in any age group.
| Discussion|| |
The database in the county of Ostergotland in Sweden consists of audiograms arranged in 15 different categories of occupation. The subjects included in this study have been selected from the categories mechanical work and wood processing, which were the two categories that included sufficient number of audiograms for a reliable statistical analysis. Reorganisation of the occupational health care has drastically decreased the input of audiograms to the database after 1997. That is the reason why time periods for audiometry are not the entire decades.
The hearing threshold data collected over the years were measured using different types of audiometers equipped with earphones of type Telephonics TDH-39 as well as TDH-49. It is likely that they have all been calibrated using the same 6 cm 3 acoustic coupler and the same reference values. There is some evidence of differences in electroacoustic characteristics between these two transducer types (Smith and Lutman, 1992). However, Michael & Bienvenue (1977) found no significant differences in average hearing threshold levels of a group of test subjects when comparing the two types of transducers, both calibrated on a 6 cm 3 acoustic coupler. Thus, we conclude that this factor is not likely to have affected our results to a significant degree.
Improvement over time
The youngest age groups in both occupational categories have median hearing threshold levels in the 1990s as low as is possible to measure at the mixed screening levels of 0, 10 and 20 dB HL which have been used on different occasions. The significant improvement from the 1980s to the 1990s in the youngest age group is probably partly an artefact caused by the increased use of screening level at 10 dB in the 1990s compared to the 1980s. Even if the significant improvements are small in median values the improvements in 3rd quartile, which reflects the most sensitive or most exposed part of the population, are obvious.
The reasons for improvement over time in HTLs have not been deeply investigated in this study. We do not have data to answer the question why the trend has continued. It is of course an important issue to evaluate specific actions, but the aim of this study was to focus on the overall effect that different actions have on the HTLs.
Frequency and ear differences
The fact that all median hearing threshold levels at 8 kHz are lower than at 6 kHz and that the improvements in median hearing threshold levels are less at 8 kHz than at 6 kHz supports the notion that the noise-induced hearing loss has decreased. One exception is the latest time period of wood processing workers, where no significant difference between 6 and 8 kHz was found. This might imply that there is less noiseinduced hearing loss in that category today.
There is no obvious explanation for the difference in median HTL between the left and the right ears found in several groups. In the groups not protected against exposure during military service gun firing could partly be an explanation. Laboratory experiments have shown larger average TTS on the left ear than on the right ear caused by symmetrical broad-band noise (Pirila, 1991). Epidemiological studies have shown that noise-induced asymmetry with poorer HTL on the left ear is largest at 4 kHz and independent of handedness (Pirila et al., 1991a). There is also an increased asymmetry with poorer hearing threshold levels (Pirila et al., 1991b; Chung, 1983). However, in the present study the maximum ear difference is rather at 6 kHz than at 4 kHz, just as the maximum hearing loss.
ISO 7029 (1984) corrected according to Passchier-Vermeer (1993) has been used as reference material of an unscreened population with no occupational noise exposure. There are other studies of otologically unscreened populations available, e.g. Robinson (1988) and Davis (1995). Robinsons (1988) results are based on two different studies, Glorig and Roberts (1965) who investigated a random sample of the US general population and Sutherland and Gasaway (1978) who investigated civilian employees of the US Air Force. A systematical difference was observed between these two populations in the 30-, 40-, and 50-year bands, and in these age groups only the better HTLs of the two populations were used. These data presented by Robinson (1988) show a large discrepancy from seven more recent studies of unscreened populations. This discrepancy may be explained by occupational noise exposure of part of the test subjects, audiometric test conditions and differences in information and motivation of the test subjects (Passchier-Vermeer, 1988). These more recent studies correspond reasonably well with each other and the corrected data in ISO 7029 (1984) according to Passchier-Vermeer (1988). Davis (1995) presents data from a random population in Great Britain. The material is divided into data sets with no exclusion criteria and data sets with subjects who had no significant noise immission from occupational, gunfire or social noise exposure. However, neither of these data sets are ideal for our purpose. To review the effect of occupational noise-induced hearing loss it seems most reasonable to use an unscreened reference material with occupational noise exposure as the only exclusion criterion.
Median hearing threshold levels of the investigated populations are still in the 1990s poorer than reference data. In the younger age groups the explanation is partly the too insensitive screening levels as already discussed. In the older age groups the differences are presumably related to noise-induced hearing loss.
In spite of the trend with continuing decrease in median HTLs and increasing percentage of subjects classified as normal hearing there are still significant differences between the investigated population and the reference material.
Hearing protectors at work
One reason for the relatively positive results of this study with overall improving hearing thresholds, especially at the frequencies related to noise-induced hearing loss, might be increased knowledge about hazardous noise and noise-induced hearing loss among both employers and employees. Barrenas et al. (1998) reported that the sale of hearing protectors has increased substantially according to the producers. The answers to the question about hearing protector use show a similar increasing trend in the group of mechanical workers. This increase in the group of mechanical workers has led to about the same percentage of users in both occupational categories in the 1990s. The fact that there is a lower percentage of hearing protector users in the oldest age group of mechanical workers may be due to the attitude that it is already too late, which is not uncommon in that age range.
Noise exposure during military service
According to the questionnaire the noise exposure during military service has decreased over the time periods studied. The percentages answering that they were exposed were larger in the 1980s than in the 1990s. One would expect the group of 30-39 year olds in the 1970s to reflect the same time period of military service as the 40-49 year olds in the 1980s and the 5059-year olds in the 1990s, but the percentages answering that they were exposed to noise in these three groups have decreased over time. That could mean that the criteria for defining noise exposure might have changed over time. When the subjects are reorganised according to year of birth the trend of decreasing exposure over time is clearer and is a possible part of the explanation of the improved hearing thresholds. An additional factor, independent of the subjective answers about exposure, is the fact that there was no use of hearing protectors during military service before the 1960s. In the group answering that they were exposed to noise during their military service, the part using hearing protectors differs between subjects born before and after 1950.
Subjective decrease in hearing ability
There is a difference in subjective change of hearing ability between the youngest and the oldest age groups in the 1980s and the 1990s.
This difference is probably caused by presbyacusis. The fact that this difference did not occur during the 1970s could probably be explained by that time period being the first time of examination. The absence of differences between the three time periods does not support the hypothesis that the noise-induced hearing loss has decreased. However, studies have shown that the correlation between subjective assessment of hearing ability and audiometry is small (Hallberg, 1998).
In the process of performing this study it became clear that the possibility of following population changes with regard to occupational hearing loss has become very limited. Several large databases that were kept active into the early 1990s have since been closed down. Since prevention of occupational disease requires constant monitoring and feedback, the lack of such databases may present a substantial impediment within the near-term future in regard to serious research into the epidemiology of occupational noise-induced hearing loss.
To evaluate the deviation from normal hearing in a specific population there is a need for an otologically unselected reference material. The International standard ISO 1999 (1990) suggests an age dependent HTL database for both men and women based on a representative population for the country. A material of a Swedish unscreened population is not yet available and leaves the evaluations today to HTLs in ISO 7029 (1984) corrected according to PasschierVermeer (1993) or other relevant reference data of unscreened populations. A more accurate picture of the influence of occupational noise in Sweden will be possible to present when such reference material becomes available.
| Conclusion|| |
The improvement of hearing threshold levels during the 1970s and 1980s has continued into the 1990s. However, noise-induced hearing loss is still significant among exposed groups. These results show that the hearing conservation programs continue to be meaningful. Although the awareness of the risk for noise-induced occupational hearing loss has improved, there is still a need for continuous monitoring and supervision in order to avoid a regression to earlier high levels of noise-induced hearing loss.
| Acknowledgements|| |
The study was financed by the National Institute for Working Life in Sweden.
| References|| |
|1.||Arlinger S. Ivarsson U. (1985) A data bank on hearing and use of hearing protection among 45.000 workers in small industries. Proc. Inter-Noise 85. Bundesanstalt ffir Arbeitsschutz, Dortmund, pp. 1375-1378. |
|2.||Barrenas M.L., Hellstrom P.A., Starck J. (1998) Hearing conservation. Advances in Noise Research Volume 2: Protection Against Noise. Edited by Prasher D, Luxon L., Pyykko I. Whurr Publishers Ltd, London, pp. 211-218. |
|3.||Chung D.Y., Mason K., Gannon R.P., Willson G.N. (1983) The ear effect as function of age and hearing loss. J Acoust Soc Am, 73: 1277-1282. |
|4.||Davis A. C. (1995) Hearing in Adults. Whurr Publishers Ltd, London |
|5.||Glorig A., Roberts J. (1965) Hearing levels of adults by age and sex, United States 1960-1962. National Center for Health Statistics, Ser 11, No 11. US Government Printing Office, Washington DC |
|6.||Hallberg L. (1998) Evaluation of a Swedish version of the hearing disabilities and handicaps scale, based on a clinical sample of 101 men with noise-induced hearing loss. Scand. Audiol. 27: 21-29. |
|7.||ISO 389-1 (1998) Acoustics - Reference zero for the calibration of audiometric equipment - Part 1: Reference equivalent threshold sound pressure levels for pure tones and supra-aural earphones. International Organization for Standardization, Geneva. |
|8.||ISO 1999 (1990) Acoustics - Determination of occupational noise exposure and estimation of noiseinduced hearing impairment. International Organization for Standardization, Geneva. |
|9.||ISO 7029 (1984) Acoustics - Threshold of hearing by air conduction as a function of age and sex for otologically normal persons. International Organization for Standardization, Geneva. |
|10.||ISO 8253-1 (1989) Acoustics - audiometric test methods - part 1: Basic pure tone air and bone conduction threshold audiometry. International Organization for Standardization, Geneva. |
|11.||Ivarsson U., Arlinger S. (1985) Results from repeated audiometry on 45.000 Swedish industry workers during 14 years of hearing conservation. Proc. Inter-Noise 85. Bundesanstalt ffir Arbeitsschutz, Dortmund, pp. 1387-1390. |
|12.||Klockhoff I., Drettner B., Svedberg A. (1974) Computerized classification of the results of screening audiometry in groups of persons exposed to noise. Audiology, 13: 326-334. |
|13.||Michael P. L., Bienvenue G. R. (1977) Real-ear threshold level comparisons between the Telephonics TDH-39 earphone with metal outer shell and the TDH-39, TDH-49, and TDH-50 earphones with plastic outer shells. J Acoust Soc Am, 61:1640-1642. |
|14.||National Board of Occupational Safety and Health in Sweden/The occupational Injury Information System (1999) Statistical report: Occupational injuries 1998: preliminary report, Stockholm, Am 69 SM 9901. |
|15.||Passchier-Vermeer W. (1993) Noise and health. Health Council of the Netherlands; publication no. A93/02E, The Hague. |
|16.||Passchier-Vermeer W. (1988) Occupational noise exposure and hearing, TNO Institute of Prevention Health Care, publication no. 88056, Leiden. |
|17.||Pirila T. (1991) Left-right asymmetry in the human response to experimental noise exposure. Acta Otolaryngol (Stockh) 111: 677-683. |
|18.||Pirila T., Jounio-Ervasti K., Sorri M. (1991a) Hearing asymmetry among left-handed and right-handed persons in a random population. Scand Audiol 20: 223-226. |
|19.||Pirila T., Sorri M., Jounio-Ervasti K., Sipila P., Karjalainen H. (1991b) Hearing asymmetry among occupationally noise-exposed men and women under 60 years of age. Scand Audiol 20: 217-222. |
|20.||Robinson D. W. (1988) Threshold of hearing as a function of age and sex for the typical unscreend population. Br J Audiol 22:5-20 |
|21.||Smith A.W. (1998) The World Health Organisation and the prevention of deafness and hearing impairment caused by noise. Noise & Health 1: 6-12. |
|22.||Smith P. A., Lutman M. E. (1992) Consequences of the change to standards for air-conduction hearing levels. A cautionary note. Br J Audiol, 26: 59-61. |
|23.||Sutherland H. C., Gasaway D. C. (1978) Current hearing threshold levels for noise-exposed US Air Force personnel: One year's reportings. USAF School of Aerospace Medicine, Report SAM-TR-78-39 Brooks Air Force Base, Texas |
|24.||Winkler R.L., Hays W.L. (1975) Statistics: Probability, Inference and Decision, 2nd ed. Holt, Rinehart and Winston Inc, New York. |
Division of Technical Audiology, Department of Neuroscience and Locomotion, Linköping University, S-581 85 Linköping
Source of Support: None, Conflict of Interest: None
[Figure - 1], [Figure - 2], [Figure - 3], [Figure - 4]
[Table - 1], [Table - 2]