Age-related hearing decline in individuals with and without occupational noise exposure.

This study was conducted to compare the pattern of age-related hearing decline in individuals with and without self-reported previous occupational noise exposure. This was a prospective, population-based, longitudinal study of individuals aged 70-75 years, from an epidemiological investigation, comprising three age cohorts. In total there were 1013 subjects (432 men and 581 women). Participants were tested with pure tone audiometry, and they answered a questionnaire to provide information regarding number of years of occupational noise exposure. There were no significant differences in hearing decline, at any frequency, for those aged 70-75 years between the noise-exposed (N= 62 men, 22 women) and the nonexposed groups (N = 96 men, 158 women). This study supports the additive model of noise-induced hearing loss (NIHL) and age-related hearing loss (ARHL). The concept of different patterns of hearing decline between persons exposed and not exposed to noise could not be verified.

Introduction

The interaction between noise-induced hearing loss (NIHL) and age-related hearing loss (ARHL) is still poorly understood. The traditional model to assess NIHL in older persons assumes that ARHL adds to permanent noise-induced threshold shift (NIPTS).[1] This additive model is embraced by the International Standards Organization (ISO) 1999,[2] and has the implication that the total hearing loss is the sum of ARHL and NIPTS minus a compression factor that varies depending on the degree of threshold shift. According to a less-than-additive model, the deterioration caused by ARHL is reduced within the noise frequencies in noise-damaged ears, i.e., sub-additivity.[34567] Super-additivity, a positive synergism between aging and noise exposure has also been described, on one occasion only, in an animal model.[8]

In the Framingham study, a combined effect of noise and aging has been described.[9]

According to this more recent study, the aging process in a noise-damaged cochlea is different from that in a cochlea not exposed to excessive noise. They found that elderly men with noise notches in their audiograms had a reduced progression of hearing loss over time at 3 kHz, 4 kHz, and 6 kHz, and an accelerated rate of hearing loss in frequency areas adjacent to noise-damaged frequencies, especially at 2 kHz. In men without typical noise-notches, this pattern was reversed. This finding could not be reproduced by Lee et al.,[1011] who studied 97 men aged 60-81 years at entry. In their study, the rates of threshold change for subjects with a positive noise history were not significantly different from those with a negative history of noise exposure.

Patterns of change in hearing in men and women were studied in the Baltimore Longitudinal Study of Aging.[12] However, in this study the participants were selected and had no evidence of otological disease, unilateral hearing loss, or NIHL.[12]

Kujawa and Liberman[13] studied CBA/CaJ mice: One group exposed to noise at different ages, and a control group. Animals with previous exposure demonstrated ARHL and histopathology fundamentally, unlike unexposed, aging animals, or old animals exposed for 2 weeks only. There was a substantial ongoing deterioration of cochlear neural responses, without additional change in preneural responses. There was also histologic evidence of neural degeneration throughout the cochlea. Data suggest that pathologic but sublethal changes initiated by early noise exposure render the inner ears significantly more vulnerable to aging.

Results from the Gothenburg Gerontological and Geriatric Population Study can provide more information about this issue.[14] In this study, men with and without occupational exposure to noise were followed longitudinally from age 70, to ages 75 and 79. The most profound deterioration of hearing levels between age 70 and age 75 was found at 2 kHz for both exposed and nonnoise-exposed men, but the deterioration was apparently more pronounced for the exposed group than for the control group. However, the longitudinal design was not impeccable, there was a dropout rate over time with more participants at age 70 than at age 75, and the number of participants was still smaller at age 79 for natural reasons. A comparison between groups supported the concept of accelerated ARHL at 2 kHz.[15]

The aims of this study were to describe the development of hearing decline in a population-based cohort of men and women aged 70-75, and to evaluate if the rate of hearing decline in this age period differs depending on the extent of previous occupational noise exposure. The study is an extension of the earlier one,[15] with more participants included, and with a longitudinal design.

Methods

The study design was a prospective, longitudinal, population-based epidemiological investigation, in which age-related hearing decline was evaluated with regard to previous occupational noise exposure. Altogether six age cohorts have been studied in the Gerontological and Geriatric Population Study since 1971 and three of these cohorts were included in this study. These three age cohorts were studied at 70 years and 75 years of age in a strictly longitudinal design. Results from these three age cohorts were pooled together. Both men and women were included; however, there were relatively few women with occupational noise exposure.

Participants were selected through a systematic sampling of 70-year-olds born in 1901-2 (cohort 1), 1906-7 (cohort 2), and 1930 (cohort 6), and living in the city of Gothenburg. Altogether 1013 persons participated in this study: 432 men and 581 women. From these subjects we chose 1) all who had participated in pure tone audiometry at 70 and at 75 years of age. Most of these had also 2) answered a questionnaire, including a question regarding self-reported occupational noise exposure. Two subjects with unreliable audiograms were excluded, as well as one subject with bilateral conductive hearing loss. Subjects with unilateral conductive hearing loss in the worse ear were included as only the better ear was assessed here. Altogether 171 men and 194 women were included, [See Table 1].

Table 1

Number of subjects in original cohorts at 70 years of age, and number included in present study (in bold print), by gender, cohort and noise exposure groups

	Men				Women				Total
Cohort number	1	2	6		1	2	6
Original N	161	132	139	432	197	176	208	581	1013
Noise group
No noise exposure	71	12	13	96	108	32	18	158
1-10 years of noise	11	1	1	13	11	1	2	14
>10 years of noise	42	15	5	62	20	1	1	22
	124	28	19	171	139	34	21	194	365

The final numbers represent all subjects who had performed audiometry at both 70 years and 75 years of age, and had answered the questionnaire about noise exposure

Pure tone audiometry was performed in a standardized manner in all three cohorts, as much as was possible, at all three test sessions. Otoscopy was performed before testing, and occluding cerumen was removed.

Air conduction thresholds for both ears were registered and stored in a database.

The following frequencies were tested: 0.25 kHz, 0.5 kHz, 1 kHz, 2 kHz, 4 kHz, 6 kHz, and 8 kHz, with the exception of cohort 1, where 6 kHz was not included. The ascending method was used, and ISO 389 and ISO 8253-1 standards were applied. The tests were performed in quiet office rooms in the same building. The background noise was at most 32 dB(A) at 0.25 kHz, and it did not exceed 20 dB(A) in the frequency range 1-8 kHz. The ambient noise levels were not measured in connection to the testing of cohort 6. Madsen OB-70 (Originally Madsen of Denmark, now (after series of mergers) part of GN Otometrics A/S, Taastrup, Denmark) audiometers with TDH-39 (Telephonics Corp., Farmingdale, NY, USA) headphones were used to test the participants in cohorts 1 and 2. Trained clinical audiologists performed the tests (air and bone conduction thresholds). Masking was performed when applicable. For cohort 6, an Interacoustics AD25 audiometer (Interacoustics A/S, Middelfart, Denmark) with TDH-39 headphones was used. Pure tone audiometry (air conduction thresholds) was performed by registered nurses with special training in audiometry. No masking was performed.

Pure tone thresholds at 0.25 kHz, 0.5 kHz, 1 kHz, 2 kHz, 4 kHz, and 8 kHz were included in the present study. The average annual rate of hearing decline in dB/year was calculated as the difference in hearing thresholds for each frequency between the two test occasions divided by 5 years.

Questionnaires were addressed to the participants of the cohorts. The questionnaires varied, but some of the questions were the same for all three cohorts. In cohorts 1 and 2, the participants were asked to state how many years they had spent working with occupational noise. For the third cohort, the question was of a multiple choice type, with the following options: None, less than 10 years, more than 10 years.

The occupational noise exposure was in most cases considerable: Most of the men had been working in very noisy work places, such as an automobile factory, a shipyard, or a ball bearing factory, whereas the women had mostly worked in the textile industry or in other industrial factories. Moreover, measures to prevent NIHL, as well as the awareness of the problem, had not been fully developed in the early and mid-1970s.

The median hearing thresholds for the three cohorts were compared separately for men and women, using the Kruskal–Wallis test of nonparametrical data, and did not differ more than what can be expected within a Type I error; hence the three cohorts were pooled and calculations were made group-wise based on the extent of noise exposure. Between-group differences of the rate of hearing decline were calculated separately for men and women for each frequency using the Mann–Whitney U test of non-parametrical data. The software Statistica 10, StatSoft® (StatSoft Inc., Tulsa, OK, USA) was used.

The Regional Board of Ethics of Gothenburg University approved the study of cohort 6 (Dnr S 227-00). The earlier cohorts were studied with the following ethical approvals: Dnr 52/76 760322 and Dnr 263/90 900920.

Results

The mean pure tone thresholds for men and women at baseline (at age 70) and at follow-up (at age 75) are shown in Figure 1a–f. The hearing thresholds of the total group is shown in Figure 1a (men, N = 171) and 1b (women, N = 194). The thresholds of the subjects with no reported occupational noise exposure are shown in Figure 1c (men, N = 96) and 1d (women, N = 158), and those of exposed men (N = 62) and women (N = 22) are shown in Figure 1e and f respectively. The subjects with 1-10 years of reported noise exposure were rather few and therefore not shown separately [Figure 1].

Figure 1

Mean hearing threshold levels in the better ear, men (a-c) and women (d-f) At 70 and at 75 years of age. Subdivision by noise groups: Total group (a and b) No reported occupational noise exposure (c and d) and more than 10 years of occupational noise exposure (e and f)

The rate of hearing decline between 70 years and 75 years of age did not differ significantly between noise-exposed and nonexposed individuals, of either gender [Figure 2a and b]. For both genders the decline was most pronounced in the high frequencies, at 2-8 kHz, where it ranged from about 1.3 dB/year to more than 2 dB/year for men (mean values). For women the corresponding values ranged 1.6 dB/year to almost 3 dB/year. In the lower frequencies, 0.25-1 kHz, the decline was about 0.5 dB/year for men and about 1 dB/year for women. Between-group differences were calculated using the Mann–Whitney U test of nonparametrical data, and no significant differences in the rate of hearing decline depending on noise history could be seen at any of the investigated frequencies [Figure 2].

Figure 2

(a and b) Mean annual threshold change in dB/year, in the better ear, from 70 years to 75 years of age, by gender and noise exposure groups: No reported noise (No N) or >10 years of reported occupational noise exposure (N)

Adapted data for subjects of the same age for reference, from Lee et al. (2005) with negative or positive noise history, and from Pearson et al. (1995), are supplied. Mean values, rather than medians, are shown for illustrative purposes

Discussion

In this longitudinal study of hearing between 70 years and 75 years of age, the pattern of longitudinal decline of the pure tone thresholds did not differ between occupationally noise exposed and nonexposed groups, of both genders. The result supports the additive model favored by Lee et al.[10] Moreover, the rate of the decline is consistent with that reported by Lee et al.[10] An exception is the high frequency range, where the decline in this study was larger in the female group than that reported by Lee et al.[10] When the results from this study were compared to results from the Baltimore Longitudinal Study of Aging,[12] the annual decline was rather similar in the male group. However, in the female group the decline was much smaller, in accordance with Pearson et al.,[12] than what was found in this study in the comparable female group without noise exposure. There are possible explanations for the discrepancies between the decline of this study (especially for women, high frequencies) and those of the other two studies. One reason is different selection criteria. Another reason may be different follow-up periods. Still another possibility is that the decline in our epidemiological study appears to vary at different ages. In the study by Jönsson and Rosenhall[16] an age cohort was followed from age 70 to age 90. The study was not strictly longitudinal as new participants from the general population were added to compensate for the considerable loss of participants during the span of the study. According to this study, the decline was twice as large in the eighth decade (70-80 years), as compared with the ninth (80-90 years).

Gates et al.[917] reported that frequencies adjacent to supposed noise notches exhibit an accelerated annual decline, compared to the same frequencies measured in individuals with no such notches. We found no support for this concept, provided that these dips reflect NIHL, which has been called into question. According to Nondahl et al.,[18] there is a poor correlation between audiometric notches and a positive noise history.

In an earlier study from this research group,[15] some support for accelerated hearing decline affecting frequencies outside the noise frequencies/area was reported. However, this finding was based on a different method of analysis, i.e., comparisons of subgroups exposed and not exposed to occupational noise, at ages 70 years and 75 years. The present analysis was carried out on partly the same material, but the current design was strictly longitudinal, which explains why the study yielded a different result.

Conclusion

There were no significant differences in hearing decline, at any frequency, from 70 years to 75 years of age between the noise-exposed and the nonexposed subjects of either gender. The concept of different patterns of hearing decline between persons exposed and not exposed to occupational noise could thus not be verified.

Financial support and sponsorship

This study was part of the Gothenburg Gerontological and Geriatric Investigation, and was supported by The Tysta Skolan Foundation and Hörselforskningsfonden.

Conflicts of interest

There are no conflicts of interest.