Noise Health Home 

[Download PDF]
Year : 2004  |  Volume : 6  |  Issue : 24  |  Page : 35--41

Reference data for evaluation of occupationally noise-induced hearing loss

M Johansson, S Arlinger 
 Division of Technical Audiology, Department of Neuroscience and Locomotion, Linköping University, SE-581 85, Linköping, Sweden

Correspondence Address:
S Arlinger
Division of Technical Audiology, Department of Neuroscience and Locomotion, Linköping University, SE-581 85 Linköping


Relevant reference data are required in order to determine the effect from occupational noise exposure on hearing. Pure-tone averages (PTA) of hearing threshold levels simplify the evaluation for audiometric frequencies typically affected by noise. The present study provides reference data of high frequency (HF) PTA over 3, 4 and 6 kHz for a general adult population, aged from 20 to 79 years, not exposed to hazardous occupational noise. The results are presented as statistical distributions of HF PTA values as functions of age, and as prevalence of different degree of HF PTA in the age groups 20-29, 30-39, 40-49, 50-59, 60-69 and 70-79 years.

How to cite this article:
Johansson M, Arlinger S. Reference data for evaluation of occupationally noise-induced hearing loss.Noise Health 2004;6:35-41

How to cite this URL:
Johansson M, Arlinger S. Reference data for evaluation of occupationally noise-induced hearing loss. Noise Health [serial online] 2004 [cited 2023 Feb 3 ];6:35-41
Available from:

Full Text


One major constituent of a successful hearing conservation program is the examination of hearing thresholds of noise-exposed subjects. An early indication on noise-induced hearing threshold shift is important both for prevention of further deterioration for the exposed person and as an incentive for actions against hazardous working environment.

To evaluate the impact from occupational noise exposure on hearing thresholds relevant reference data are crucial. In ISO 1999 (International Organization for Standardization, 1999) the criteria for a relevant reference data base (data base B) are described, and Johansson and Arlinger (2002) present a reference data base of a general adult population in Sweden, not exposed to hazardous occupational noise, based on these criteria. The reference data are presented as hearing threshold levels (HTL) for 10 single test frequencies (0.125, 0.25, 0.5, 1, 1.5, 2, 3, 4, 6 and 8 kHz).

It is well known that noise-induced hearing loss initially appears as a threshold shift in the frequency region 3 to 6 kHz in pure-tone audiometry (Quaranta et al, 2001). These frequencies are obviously of most interest when evaluating hearing loss due to occupational noise exposure.

An indicator for noise-induced hearing loss is, beside single frequency hearing threshold shift, deterioration in pure-tone average of the frequencies 3, 4 and 6 kHz (referred to as HF PTA in the present paper). Audiometers are available on the market, which automatically determine and indicate the HF PTA, and this single value is easily applied when evaluating results from individuals as well as a group of subjects. Also, test-retest accuracy is somewhat better when considering the average of HTLs at several frequencies compared to a single frequency (Arlinger, 2003; Dobie, 1983).

Since the inter-frequency correlation of HTLs is less than 100%, the HF PTAs have to be determined from each subject's HTL. To our knowledge no reference data in terms of average HTLs with focus on the high frequency range have been presented. The aim of this paper is therefore to provide reference data on HF PTA for evaluation of occupational noise-induced hearing loss.

 Materials and Methods

The reference data are based on results from 603 subjects, 266 male and 337 female, aged from 18 to 80 years. The subjects were randomly selected from a general adult population in the province of Ostergotland in Sweden and the only exclusion criterion was hazardous occupational noise exposure. A questionnaire was used in which the subjects estimated whether the working environment was noisy. A four ­alternative question was used and the subject stated whether he or she could have a conversation with another person at a distance of 1 m. Those who answered 'Yes, with normal voice level' or 'Yes, with raised voice' were included in the study; while those who answered 'Yes, if I shout' or 'No, not at all' were excluded. In total 32.4% of the invited subjects were included in the analyses. The selection procedure is described and discussed in Johansson and Arlinger (2002).

Pure-tone audiometry was conducted using an ascending method with 5-dB steps according to ISO 8253-1 (International Organization for Standardization, 1989). A Grason-Stadler GSI 68 audiometer with EAR-tone 3A insert earphones was used to determine air conduction HTLs for the test frequencies 0.125, 0.25, 0.5, 1, 1.5, 2, 3, 4, 6 and 8 kHz on both ears. The equipment was calibrated according to ISO 389-2 (International Organization for Standardization, 1994).

The test battery also included otoscopic examination, tympanometry and a questionnaire. For further details see Johansson and Arlinger (2002, 2003).

The results are described as HF PTA for male and female subjects at different ages. The HF PTA was determined as the mean value of the HTLs at 3, 4 and 6 kHz for each ear examined. Smooth percentile curves were determined based on the individual PTA values and the resulting curves were used to determine the parameters A', B', C and D in a curve fit to the expression:

PTA=A'+B'· tanh(C· age+D)

The method for the parameter estimation is described in Johansson and Arlinger (2002). Separate sets of the parameters A', B', C and D were determined for male and female subjects and for nine different percentile curves from 10 to 90%. Based on these estimated models, HF PTA data are presented for each year between 20 and 79 years for male and female subjects.

Prevalence data for different degrees of PTA values are also presented. The prevalence was determined for HF PTA in the subjects' better ears (BE) equal to or worse than from -5 to 65 dB in steps of 5 dB. The results are presented for male and female subjects in six age groups (20­29, 30-39, 40-49, 50-59, 60-69 and 70-79 years). A 95% confidence interval was also determined for each prevalence value based on the binomial distribution.


The values of HF PTA obtained are listed in [Table 1] for male subjects and in [Table 2] for female subjects. The results are described for nine percentiles, from the 10th to the 90th, and for each year of age from 20 to 79 years.

The age-related shift in PTA is illustrated for male subjects in [Figure 1] and for female subjects in [Figure 2].

The prevalence of different degrees of HF PTA hearing loss is described for male subjects in [Table 3] and for female subjects in [Table 4]. The results are divided into male and female subjects in six age groups and the PTA criterion ranges from -5 to 65 dB HL.


The results described in [Table 1],[Table 2], [Table 3] and [Table 4] are determined to provide reference data on HF PTA for evaluation of the influence from occupational noise exposure. On an individual basis the HF PTA for the subject under study could be compared to the reference PTA at the specific age. If the PTA for the test subject exceeds the 90th percentile of the reference data, the subject diverges from the non-exposed reference population with 10% risk of misinterpretation. If a history of occupational noise exposure is present and the subject's audiogram shows the typical pattern of noise-induced threshold shift with a high-frequency notch, it may be concluded that this threshold shift is likely to be an effect from occupational noise exposure.

Based on the approximation that the distribution of the reference PTA is a half normal distribution above the 50th percentile, the standard deviation (SD) equals the 84th percentile. The PTA for a specific age-value is based on between 22 and 160 observations. If the standard error (SE) is determined for 42 observations, which covers all cases above 29 years for male subjects and above 23 years for female subjects, the result equals 0.154· SD. For a one-tailed distribution and 5% risk of type-I error, 1.64· SE equals the 60th percentile of the distribution. Based on these approximations the 60th percentile of the PTA distribution would be a reasonable criterion- value for evaluating the influence from occupational noise exposure on results for a group of subjects.

Another approach to evaluate group results of investigated subjects is to compare the prevalence of a specific degree of PTA in this group to the prevalence of the same PTA for the reference data. If the prevalence in the investigated group exceeds the upper limit of the 95% confidence interval for the prevalence in the reference data, the investigated group is affected by the occupational noise exposure with 2.5% risk of misinterpretation.


The study was financed by the National Institute for Working Life in Sweden.[8]


1Arlinger S. (2003) Measurement of hearing thresholds. In Noise and its effects. Luxon L., Prasher D., eds. Whurr Publishers, London, in press.
2Dobie R. (1983) Reliability and validity of industrial audiometry: implications for hearing conservation program design. Laryngoscope 93(7): 906-27.
3International Organization for Standardization (1989) Acoustics-Audiometric test methods-part 1: Basic pure tone air and bone conduction threshold audiometry. ISO 8253-1. Geneva: ISO.
4International Organization for Standardization (1994) Acoustics-Reference zero for the calibration of audiometric equipment-part 2: Reference equivalent threshold sound pressure levels for pure tones and insert earphones. ISO 389-2. Geneva: ISO
5International Organization for Standardization (1999) Acoustics - Determination of occupational noise exposure and estimation of noise-induced hearing impairment. ISO 1999. Geneva: ISO
6Johansson M., Arlinger S. (2002) Hearing threshold levels for an otologically unscreened, non-occupationally noise­exposed population in Sweden. Int J Audiol 41: 180-194
7Johansson M., Arlinger S. (2003) Prevalence of hearing impairment in a population in Sweden. Int J Audiol 42: 18-28
8Quaranta A., Sallustio V., Quaranta N. (2001) Noise induced hearing loss: Summary and perspectives. In Noise induced hearing loss: Basic mechanisms, prevention and control. Henderson D., Prasher D., Kopke R., Salvi R., Hamernik R., eds. Noise Research Network Publications, London, pp 539-557