Clinical Otolaryngology 1979, 4, 49-56

Temporary threshold shift, loudness, and auditory evoked potentials M . C. B O T T E , H . A . BEAGLEY* AND R . C H O C H O L L E Laboratoire de Neurophyysiologie, College de France, II Place Marcelin-Berthelet, 75231 Paris, Cedex 05,France

'4ccepted for pziblication 16 .November 1977

BOTTEM.C., BEACLEY H.A. & CHOCHOLLE R. (1979) Clinical Otolaryngology 4,49-56 Temporary threshold shift, loudness, a n d auditory evoked potentials T T S , of zodB was produced in a group of subjects using a 1000 Hz pure tone. All TTS and related measurements were made at 1500 Hz. The recovery of T T S , and of T L S or 'temporary loudness shift' (measured against a reference tone at 20db, 40dB and 6odB SL in the opposite ear), as well as N1-P2 amplitude of the Vertex potential, were plotted. There was an orderly recovery of both TLS and T T S although the rate of recovery of the latter was noticeably more rapid. Recovery of N,-P2 amplitude was much less orderly, although it did recover in much the same way. When the amplitudes were measured at 4odB SL the recovery was more orderly and followed the course of T T S more closely. It was concluded that any relation between N,-Pz amplitude and growth of loudness was indirect. Keywords V-potential amplitude temporary threshold shift loudness loss loudness recovery

In an earlier experiment,' the simultaneous recovery of TTS and of the amplitude of the auditory evoked potential from the vertex (V-potential) was studied over a period extending from z to 80 min after a fatiguing sound, in this case a 1000Hz pure-tone. The intensity and duration of the tone had previously been determined for each subject so as to produce a TTS at z min that fell as far as possible in a 5 dB interval centred on 20 dB. I n every case the TTS and the V-potential amplitudes were determined one half octave above the frequency of the fatiguing sound, at 1500Hz. As it was not feasible to observe and measure V-potential amplitudes at threshold in most cases, it was decided to deliver the succession of the 1500 Hz tone bursts used to elicit the V-potential at 40 dB SL, which would represent the most linear part of the input-output function of the V-potential amplitude for this particular interstimulus interval. This point will be mentioned further in the discussion. When the preliminary experiment was concluded it was decided to increase the scope of the investigation I by studying the recovery of the V-potential at 20 dB and 60 dB, as well as at 40 dB SL already used, and z by studying the recovery of loudness, following the same sound exposure as used in the earlier experiment. The loss of subjective loudness at 2 min which we

* Present address for H . A. Beagley: Royal National Throat, Nose and Ear Hospital, Gray's Inn Road, London WCI. Address offprint requests to M. C. Botte. o ~ o ~ - ~ ~ ~ ~ / ~ g$02.00 / o ~ o0o 1979 - o oBlackwell ~~ Scientific Publications

49

50

M . C. BOTTE, H . A. BEAGLEY A N D R . CHOCHOLLE

have called Temporary Loudness Shift (TLS2) was measured, and its recovery plotted over 80 min at the same levels, i.e. 20,40 and 60 dB SL. The same five subjects were used throughout.

Method VERTEX P O T E N T I A L

The 40 dB SI, data from the preliminary experiment were expanded so that each of the 5 subjects was tested 6 times at this level. I n addition the decrease in the V-potential and its subsequent recovery was also tested at 20 dB and at 60 dB SL. As the 20 dB and the 60 dB measurements were carried out successively, in each of the 6 experiments the order of presentation of the test tones was alternated from experiment to experiment. By counterbalancing in this way errors due to systematic differences in timing were minimized. Thus we have data on the amplitude of the V-potential in 5 subjects, each tested 6 times, before and after fatiguing sound and the depression and recovery of the amplitudes followed and recorded for 80 min using test tone intensities of 20, 40 and 60 dB SL.

LOUDNESS

T h e same 5 subjects were also submitted to a series of tests to measure the effect of auditory fatigue upon the subjective sensation of loudness at the same hearing levels. The same 1000 Hz fatiguing tone was used; the intensities and durations chosen were those known to produce a TTS at 2 min of 20 dB in each particular subject and the right ear was exposed to the fatiguing sound throughout the experiment. Measurements of loudness and its subsequent recovery were made, using a I 500 Hz tone, before and after exposure to the 1000Hz fatiguing tone. Loudness was measured 3 times at each level (20, 40 and 60 dB SL), in each subject (twice only on one subject). This made a series of 45 tests, but only 42 are recorded, one subject being unable to complete the series. The loudness measurements were carried out by the method of comparison. The subject was asked to compare the subjective loudness of a sound presented first to the left ear at a constant hearing level (20, 40, or 60 dB SI, according to the session) with that presented to the right ear at a level which could be varied. The subject was asked to say whether the sound presented to the right ear was ‘louder’ or ‘softer’ than that presented to the left. For each comparison, the experimenter varied the sound level in the right ear decibel by one decibel first in an ascending, then in a descending series of intensities. In this way the subjective level of equal loudness was determined between this tone and reference tone presented to the left ear. The difference was next determined between this level, which had been obtained before exposure to fatiguing sound, and those determined at specified times following exposure. This gave a measure in decibels of the reduction in loudness of the 1500Hz tone following auditory fatigue, as well as the manner and extent of its recovery.

Loudness and evoked responses

51

Results Tables I and 2 and Figure I show the mean reduction in loudness after sound exposure for the whole groups of subjects and also the threshold recovery curve (i.e. T T S ) from these subjects in the preliminary experiment. Tables 3-5 and Figure 2 show the mean depression of the V-potential amplitude and its subsequent recovery. Table I Reduction of loudness (dB). Means from 5 subjects, each tested 3 times Post-exposure (min)

Pre-exposure

I0 20 40 13.43 10.14 7.96 6.79 t = 12.44 t = = 10.24 t = 9.59 t = 7.44 ( < 0.001) ( < 0.001) ( < 0.001) ( < 0.001) ( < 0.001) 9.04 7.71 7.00 5.93 4.86 t = 12.05 t = 10.56 t = 10.94 t = 10.05 t = 7.36 ( < 0.001) ( < 0.001) ( < O . O O I ) ( < 0.001) ( < 0.001) 3.61 3.04 2.57 2.14 I .89 t = 5.92 t = 6.20 t = 4.28 t = 3.82 t = 3.44 ( < 0.001) ( < 0.001) ( < 0.001) ( < 0.01) ( < 0.01) 2

2odB SL 4odB SL 6odB SL

5 11.89 8.94 t

80 5.89 t = 8.54 ( < 0.001) 3.43 t = 6.73 ( < 0.001) 1.54 t = 2.66 ( < 0.01)

Table z Analysis of variance of TTS and TLS *

Sources of variation

Between subjects Between delta-t Between levels: Residual error Total

d.5 4 5

20 dB vs 40 dB Total

I

3 '07 119

Sums of squares 173 754 138 '576 5'5 3017

Mean squares 43.3 150.8 138.0 525.3 4.8

F 9.02 31.42 28.75 109.44 -

P 0.01 0.01

0.01 0.01 -

* This table shows significant differences for loss of loudness between subjects, between times at which loudness was measured, and between levels-in particular between 20 and 40 dB and, a fortiori, between o and 60 dB. Discussion

The reduction in loudness shows highly significant differences, and the magnitude of the reduction is in inverse relation to the level at which it was measured, either 20, 4.0 or 60 dB SL.. Thus 20 dB of T T S z produces a TLSz (Temporary Loudness Shift) of 13.4 dB, 9 dB and 3.6 dB at the hearing levels 20 dB, 40 dB and 60 dB respectively. When the loudness measurements are made 2 min after the cessation of the auditory fatigue, it can be seen that

M . C . BOTTE, H . A. BEAGLEY A N D R . CHOCHOLLE

52

Table 3 Reduction of NL-P2amplitude (pV). Means from 5 subjects, each tested 6 times)

Pre-exposure

Post-exposure (min)

5 6.67 11.49

2

20

dB SL

40 dB SL

60 dB SL

20 40 80 4.39 3.11 I .67 1.48 t = 7.16 t = 6.14 t = 3.60 t = 3.38 (< 0.001) ( < 0.001 ( < 0.01) ( < 0.01) I .64 I .70 I .oo 0.44 t = 2.37 t = 3.13 t = 1.78 t = 0.76 ( < 0.05) ( < 0.01) (NS) (NS) 1.11 0.58 0.39 0.81 t = 1.58 t = 0.88 t = 0.70 t = 1.44

I0

8.10 7.32 s.d. = 2.17 t = 13.99 t = ( < 0.001) ( < 0.001) 2.53 11.20 1.97 s.d. = 3.00 t = 2.76 t = 4.18 ( < 0.01) ( < 0.001) 11.01 0.57 I .70 s.d. = 2.83 t = 0.79 t = 2.62

(NS)

(NS)

(i0.02)

(NS)

(W

(N.9

Table 4 Analysis of variance of reductions of N,-P, amplitudes compared with pre-exposure amplitude*

Sources of oariation Between subjects Between delta-t (post-stim.) Between levels of test-tone:

Sums o j Mean 4

min vs 5 min Total 4odB vs 60 dB Total

I

2

5 I 2

Residual error Total

78 89

squares squares 86 21.5 0.90 0.9 98 19.6 7.3 7.3 180 90.0 150 I .9 513

F

P

11.32 0.01 0.47 NS 10.32 0.01 3.84 NS 47.37 0.01 -

-

~

* This table shows that the differences of amplitude reductions are significant in respect of the different subjects, in respect of the time differences (delta-t), except that the 2-5 min difference is not significant, and in respect of the level of the test-tones, except that the difference between 40 and 6odB is not significant. Table 5 Analysis of variance of N,-P2 amplitudes*

d.$

Sources of variation Between subjects Between times of measurements: Between levels of test-tones: Residual error Total

4

Pre- vs postTotal 4odB vs 6odB Total

Sirmsof squares 265

I

61

6 2

159 6 706

92

215

'04

I345

I

Mean squares 66.25 61 26.50 3 353 2.34

F

P

28.31 26.07 11.32 1.28 150.85

0.01 0.01

0.01

NS 0.01

-

* This table shows that the amplitude differences are significant in respect of the different subjects, in respect of the times at which the measurements were made and especially between pre-exposure and post-exposure measurements, and in respect of the levels of the test-tones used, except that there is no significant difference between the 4odB and the 6odB levels.

Loudness and evoked responses

53

shift of loudness rises by 5 or 6 dB for each 20 dB drop in the sensation level. However, these differences in the loss of loudness are progressively narrowed as the recovery time increases, and at the end of 80 min the differences amount to only 2 or 3 dB, while there is no significant difference between the loss of loudness at 20 dB hearing level and depression of threshold (TTS), the loss in each case being 5.9 and 6.1 dB respectively. It is also evident that the curves of the recovery of loudness in decibels are essentially linear when plotted against the logarithm of time (Figure I). This relation is well known in the case of TTS2.’ However, there is another important aspect of the recovery curves for loudness, and this is in relation to their slopes. On calculating a regression line for each of the 4 experimental curves, 0,20, 40 and 60 dB SL slopes of 8.65, 5.01,3.42 and 1.30 respectively, are obtained. These values permit calculation of the theoretical minimum time for complete recovery, namely for threshold (0 dB), 360 min, or 6 h, whereas for 20 dB, 40 dB and 60 dB SL the minimum recovery times are respectively 1ooo,960 and 1073 min, or about 16 or 17 h (Figure I). This observed difference in the recovery times between the threshold and suprathreshold curves does not seem to be an artefact. We have tried to verify experimentally in 2 subjects the persistence of a T L S after practically complete recovery of T T S . This was done by measuring both threshold and loudness after a sufficiently long recovery period (360 min for subject D.B. and 150 min for subject R.A.). The values obtained as a result of 6 trials for each Table 6 Comparison of longterm TTS and TLS. Means from subjects, each tested 6 times

TTS ( I 50) (360) Mean s.d.

- 1.00 I .49

20dB SL -2.63 2.06

TLS (150) (360) 4odB SL - 1.59 2.03

2

6odB SL - 0.30 I4 8

t

TLS vs TTS

3.70 ( < 0.01)

I .32 (NS)

1.14

(W

subject are shown in Table 6. These data indicate the persistence of TLS as long as there is a measurable TTS, at least at 20 and 40 dB SL. The T L S is actually larger than the T T S , at these 2 levels, although it is only at the 20 dB level that statistical significance is observed. These results are not confirmed by most earlier published results because the effect of auditory fatigue has usually been measured at threshold, and when loss of loudness has been the recovery of loudness was not measured concurrently with that of the T T S . Comparing the recovery curves of the V-potential amplitudes and of loudness (Figures I & 2) it is clear that while the 2 measures follow a similar and roughly parallel course, they are by no means superimposed. Whereas for the 3 sensation levels considered (20, 40 and 60 dR) the curves of loudness recovery are quite distinct and nearly equidistant, it is clear that the

54

M . C . BOT'I'E, H . A . B E A G L E Y AND K. C H O C H O L L E

I

I

I

1

!

I

I

I

I

-2O.#----.-

4-

m

-

9

SL