LePage, Murray, Tran 1994_2

Better Hearing Australia Conference
Adelaide 7-11 August 1994

2nd presentation

Comparison of otoacoustic emission measures of cochlear damage in the Australian population with hearing loss in the Australian and British populations.

E. L. LePage, N. M. Murray and K. Tran,
Hearing Loss Prevention Research Unit
National Acoustic Laboratories
126 Greville Street, Chatswood, NSW, 2067



Otoacoustic emissions in response to clicks may be measured as a sound in the ear canal which diminishes throughout life, but which is not accompanied by hearing loss until a critically-low value is reached. By plotting the distribution of emission strengths as a function of age it is possible to calculate the percentages of the population reaching the critical value, i.e. who from our cross comparison audiometric data may be demonstrating initial symptoms such as mildly raised pure tone thresholds, problems with sorting sounds and tinnitus. A fully detailed study of the actual onset of hearing loss has not yet been conducted in Australia such as has been carried out in England by Davis (1989). The curve of percentile of the Australian population crossing a critical value of emission strength co-incides quite remarkably with the analogous curve describing distribution of the British population showing at least a mild hearing loss.


The object of this paper is firstly to show how emission strengths co-vary with pure tone thresholds. However, unlike the previous paper, in this group of people we have left in all the pathological cases. We have determined the three frequency pure tone average 1, 2 and 4 kHz and compared these with the emission strengths computed between 1 and 4 kHz. Figure 1 plots coherent emission strength (CES dB SPL) (LePage and Murray, 1993) with pure tone thresholds.

In Figure 1A a large grouping of people can be seen at the high values, but we find that most of those with hearing losses appear on the left-hand side of this Figure. This chart shows that there exists a “damage threshold for hearing loss” which is to be distinguished from the usual threshold sound levels for hearing acuity. This new threshold phenomenon is best understood as a critically-low value of coherent emission strength which, when passed, is associated with a rise in hearing threshold or loss of hearing acuity. We have determined this critical value is approximately -3 dB sound pressure level (SPL) (LePage and Murray, 1993).

Figure 1B shows the overall trend. Neonates are obviously not included in Figure 1 for lack of pure tone thresholds, but we do know neonates consistently produce the highest emission strengths, 38 dB SPL being the highest seen in our sample. As shown in the previous paper, the overall trend is a decrease in emission strength throughout life. When it reaches a critical value, around the knee of the curve in Figure 1B, hearing loss becomes evident so that cases of hearing loss tend to appear in the lower left quadrant. From these kinds of plots we can determine specificity and sensitivity for the test as a screening test.

Figure 1
Figure 1A and B. Interdependence of Coherent Emission Strength and Hearing Level.People with normal hearing occupy about 80 per cent of the total range of emission strengths. All those with hearing loss have emission strengths in the lowest 20 percent of the range (Figure 2). It is therefore possible to track the decline for many years before most individuals report a hearing problem. As seen in Figure 3 the overall population is declining at about 3 dB per decade from a mean value at birth of about 20 dB SPL. These changes that are occurring above the critical value are pre-clinical changes, i.e. we can watch trends in young people before there is any sign of hearing loss. This technique, therefore, in combination with our population statistics, gives us a potent early warning tool for tracking a person’s susceptibility to hearing loss.

Figure 2
Figure 2. Low emission threshold for the onset of hearing lossBASIS OF EARLY WARNING DETERMINATION

Figure 3 shows in addition to the decline in emission strength with age (Murray et al., 1994, Murray and LePage, 1993), a horizontal line representing the critical value of emission strength (-3dB SPL) determined from our audiometric data. Normal hearing levels occur for ears with CES values higher than this.

Figure 3Figure 3. Distribution of individuals affected by at least a mild hearing loss.The critical value of CES is assumed to be independent of age. This assumption is reasonable because the measurement is specifically targeted at distinguishing the tiny nonlinear or active component of response in the sound picked up by the microphone from the much larger passive response. Due to the frequency analyser properties of the cochlea, the emission arrives at the microphone with its frequency components separated in time, so that the spectral characteristics of the emission may, to a first approximation, be regarded as a map of damage. The recording algorithm thus affords an estimate of the number of working cells along the length of the cochlea, from which is derived the CES — our net measure of the total activity. The most potent factor likely to produce variability of this net measure is the existence of a disturbance to outer hair cell function, such as recent sound overexposure. However, cases of temporary threshold shift were not included in this analysis.The distribution of people crossing this critical value may be seen directly from the lines representing the mean value of CES versus age, and the mean ±1 standard deviation (filled points). Assuming the values in each age range are normally distributed (reasonable), 70 percent of the population lie between these curves. Accordingly, about 16 percent of the Australian population may suffer a hearing loss by age 50, while half the population will be suffering a hearing loss by age 70 and (by extrapolation) most of the population by age 90. It may be seen that due to the low slope of the aging curves small variations in the value taken as the critical value make large changes in the ages at which the population is first affected.

Figure 4 contains two curves drawn from data obtained in an extensive British epidemiological study by the Institute of Hearing Research, Nottingham, (Davis 1989). The curves represent the onset of hearing loss versus age for each person’s worse ear and better ear. If we plot the percentages of the population crossing the critical value for emission strength, we can place it over the top of the British results and find that, there is quite a remarkable coincidence in the shape and placement of the curves despite only having three points for comparison. The actual value chosen as the critical value (i.e. -1 dB, -2 dB or -3 dB) is seen to influence the agreement between the studies quite sensitively; a value of -2.5 dB would give best agreement. This value thus agrees favourably with the value determined as critical from our own internal study.

P2Fig4.gif - 11600 BytesFigure 4. Comparison of distributions of hearing loss in British community
with the distribution of critical values of emission strength in the Australian sample.CONCLUSIONS

  1. The distribution of current hearing impaired population is similar in England to Australia, largely weighted to higher age ranges.
  2. The similar distributions add weight to the notion that the parameter Cochlear Emission Strength is a viable inverse measure of OHC damage.
  3. The similar distributions add weight to the notion that there is a critical value of emission strength which represents a threshold of low emission strength or conversely a threshold for clinical manifestation of OHC loss.
  4. In turn, the similarity of the age distributions lends additional weight to the notion that the critical value is indeed independent of age and depends only on net cochlear activity.
  5. This critical value of CES is approximately -2.5 dB SPL. Once middle ear problems are ruled out, for individual values below this value the risk factor for sensorineural (permanent) hearing loss is high, far above this value the risk factor is low.


Davis, A. (1989). The prevalence of hearing impairment and reported hearing disability among adults in Great Britain. Intl. J. Epidemiol. 18, 911-917.

LePage, E.L. and Murray, N.M. (1993). Click-evoked otoacoustic emissions: comparing emission strengths with pure tone audiometric thresholds. Aust.J.Audiol.15, 9-22.

Murray, N.M., and LePage, E.L. (1993). Age dependence of otoacoustic emissions and apparent rates of ageing of the inner in an Australian population. Aust.J.Audiol.15, 59-70.

Murray, N.M. , LePage, E.L. and Tran, K., (1994). Aging characteristics of the Australian population in terms of otoacoustic emission strengths; global and individual picture. (This publication).

Last Modified: Tuesday, 30 November 2004