Otoacoustic emissions and quinine sulfate - UT … emissions and quinine sulfate a) ... The pre-drug baseline measurements also included psychophysical determinations of sensitivity

Otoacoustic emissions and quinine sulfate a) Dennis McFadden and Edward G. Pasanen

Department of Psychology and The Institute for Neuroscience, The University of Texas, Austin, Texas 78 712

(Received 6 May 1993; accepted for publication 1 February 1994)

A moderate dose of quinine sulfate, administered to three young adult males, reduced or eliminated various forms of otoacoustic emission (OAE). The individual differences in response to the drug were substantial, but a number of generalizations did emerge. The time courses of onset and recovery were considerably more rapid than for the parallel effects produced by aspirin. Most spontaneous otoacoustic emissions (SOAEs) were eliminated within 7 h of the first 325-mg dose (about 3 h after the second dose). Most $OAEs showed partial or complete recovery about 24 h after the last dose, although considerable instability often remained. The functions relating the magnitude of a distortion-product OAE (DPOAE) to the sound-pressure level (SPL) of the primary tones producing it were displaced toward higher primary levels and became lower sloped following quinine administration. The magnitudes of $OAEs, DPOAEs, and nonlinear peaks in the click-evoked spectra declined and recovered with grossly similar time courses, but there were some partial dissociations. The ability of a DPOAE to suppress an $OAE lying about 50 Hz below it either increased slightly or remained about constant through the drug episode, even though the magnitudes of both DPOAE and $OAE were changing. On several occasions, increases in SOAE magnitude of as much as 10-20 dB were observed during the first 15-30 min of an SOAE measurement period (an initializing effect). Psychophysical measures revealed hearing losses of as much as 20 dB at some frequencies in some subjects. Several short-lived "enhancements" of OAEs are discussed relative to similar quinine-induced effects reported in an animal model.

PACS numbers: 43.64.Jb, 43.64.Kc, 43.64. Gz

INTRODUCTION

The ototoxic potential of quinine compounds has been recognized for at least three hundred years (see review by Hawkins, 1976). Typical symptoms include hearing loss and tinnitus, both of which are generally totally eliminated within 24-72 h of cessation of the drug--unless high doses or protracted administration were necessary. While a number of possible mechanisms of action have been proposed for quinine's ability to produce hearing loss and tinnitus (see Karlsson et al., 199 lb; Jastreboff et al., 1991 ), none is universally accepted.

Several recent reports strongly suggest that quinine can act directly at peripheral auditory sites. Karlsson et al. ( 199 lb) reported that the administration of quinine could increase the amplitude of vibration of the cochlear partition in their in vitro preparation of the cochlea. Karlsson and Flock (1990) demonstrated that quinine can alter the slow motile response of isolated cochlear outer hair cells, which raises the prospect that it might also affect the so- called active cochlear process or cochlear amplifier (Davis, 1983). Stypulkowski and Oriaku ( 1991 ) made a number of physiological measurements on chinchillas comparing the effects of quinine and salicylate, two drugs that affect the permeability of cell membranes in different ways, but which can produce similar losses in hearing (Puel et al.,

a)A preliminary report of these data was given at the 122nd Meeting of the Acoustical Society of America [J. Acoust. Soc. Am. 90, 2289 (A) (1991)].

1990). Stypulkowski and Oriaku ( 1991 ) reported that salicylate decreased an electrically evoked otoacoustic emission (just as it does other emissions; see McFadden and Plattsmier, 1984; Long et al., 1986; Long and Tubis, 1988), but quinine increased it by about 3 dB. This effect on an electrically evoked emission suggests that quinine may have the ability to alter other types of otoacoustic emission (OAE). Here, we report that spontaneous, distortion-product, and click-evoked OAEs in humans are all affected by moderate doses of quinine sulfate.

I. METHODS

A. Subjects

The subjects were three males, aged 24 to 35 years. All had hearing better than 15 dB HL at audiometric test frequencies ranging from 500-6000 Hz, as measured with an adjustment technique (Rudmose ARJ4A audiometer). The subjects were given a physical exam, including blood and urine tests, designed to identify any contraindications to the administration of quinine sulfate. Consent forms informed the subjects of the various risks and side effects commonly associated with quinine use (see Webster, 1990).

B. Procedures

Otoacoustic emissions of three types were studied-- spontaneous (SOAEs), distortion product (DPOAEs), and transient (click)-evoked (TEOAEs). All emissions

3460 J. Acoust. Soc. Am. 95 (6), June 1994 0001-4966/94/95(6)/3460/15/$6.00 ¸ 1994 Acoustical Society of America 3460

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 128.83.210.150 On: Thu, 16 Oct 2014 20:50:47

measurements were made with the subject lying on a camp cot in an electrically shielded, sound-proofed room having an intercom that allowed two-way communication. Emis- sions were monitored using an Etymotic ER-10 low-noise microphone system inserted into the outer ear canal. This system consists of a recording microphone plus two poly- vinyl sound-delivery tubes, each of which was attached to an ER-2 insert earphone. The output of the recording microphone was led to an Etymotic ER10-72 preamplifier and then passed through a low-noise amplifier/filter com- bination that provided about 30 dB of gain and high passed the microphone output at 400 Hz. The output of this amplifier/filter was led to a Nicolet/Wavetek 444a Mini- Ubiquitous spectrum analyzer located outside the sound- proofed room. For SOAEs and DPOAEs, the spectrum analyzer was set to sum eight "samples" (fast Fourier transforms) for each measurement. Emission magnitudes were read directly from the screen of the spectrum analyzer in units of dB re: 1 V, which were later transformed into dB SPL (re: 20/zPa). All SOAE and DPOAE measurements reported here were obtained with the spectrum analyzer set to its frequency-expansion mode, meaning that frequency resolution was in steps of 0.25 Hz. • Estimates of the noise level in the local vicinity of each emission were regularly noted and found to be quite stable. Ideally, these noise levels would correspond to the noise floor imposed by the electronics used, but in practice they are invariably determined by the ambient body noise generated by the individual subject. The electronic noise floors in 0.25-Hz bands centered at 1000 and 2000 Hz were about -- 13.5 dB

and --16.5 dB SPL, respectively, with the microphone placed in a small cavity. The actual noise floors were about --12 dB SPL at 1430 Hz (MP and LB), --12 at 1900 Hz (LB), and -- 10, -- 12, -- 14, and -- 17 dB SPL at 1075, 1585, 2020, and 4150 Hz, respectively (CW).

Baseline measurements of SOAEs, DPOAEs, and TEOAEs were obtained in the nondrug state for later comparison with data collected during drug administration. The frequency and level of the SOAEs were monitored on a number of occasions in the nondrug state, both before and after the drug episodes. Pre-drug baseline SOAE measurements were collected only after the subject had reclined on the cot, with the emissions-recording equipment in place, for at least 15 rain (Whitehead, 1991).

For the study of DPOAEs, a pair of primary tones having an f2/f• ratio of 1.15 was chosen for each subject individually such that the 2f•--f2 distortion product lay about 50 Hz above a strong SOAE. With this arrangement, suppression effects of the DPOAE on the SOAE could easily be observed at the same time the magnitude of the DPOAE produced by a particular level of the primaries was being determined. • The primary tones were obtained from two General Radio 1310A oscillators. The primaries were always presented via separate transducers, and were adjusted to be equal in level in the individual ear canals using the microphone in the probe assembly. The maximum level used for the primaries in these measurements was 54 dB SPL in order to stay below the point at which

the spectrum analyzer began to show inherent distortion at the 2f•--f• frequency.

For the transient-evoked OAEs, unfiltered condensa-

tion clicks of 100-/zs duration were presented at a maximum rate of 2 per second, and the responses were averaged as follows. Beginning 4 ms after click offset, a 40-ms sample of the microphone response ( 1024 points collected at a sampling frequency of 25.6 kHz) was collected using the Nicolet/Wavetek analyzer. This sample was retained in the time domain, and, immediately upon collection, was deliv- ered to a PDP 11/73 computer for summing with previously collected samples. The "averaged" waveform based upon 50 such time samples was ultimately converted to 16-bit digital format and saved as a computer file on which further analysis was performed. The 11/73 computer also generated the click stimuli and controlled the timing of their presentation. The levels of the click stimulus are spec- ified here as the SPL of a 1000-Hz tone whose maximum

amplitude corresponded to the magnitude of the 100-/zs electrical pulse at the output of the miniature microphone (in a 2-cc coupler) when both waveforms were displayed on a storage oscilloscope. Typically, data were collected for several levels of the click in rapid succession. Before each TEOAE (and DPOAE) measurement session, the level of the stimulus was set using the output of the microphone system, with the subject in place. The level of the clicks ranged from about 37 to 49 dB corresponding peak SPL, which equaled about 2- to 14-dB sensation level (SL) for subject CW.

Several procedures were implemented to reduce the noise level in the click-evoked time samples collected. First, the output of the microphone system was checked by the 11/73 computer just before the presentation of each scheduled click, and if the noise level was not below 25 dB SPL overall (from 0-10 kHz), the presentation of the click was postponed until it was. Thus, if a subject were ex- tremely quiet, all of the clicks would, in fact, be presented at a rate of 2 per second, but, more typically, the subjects' body movements and other noises caused the click presentations to be somewhat irregular. (By listening to the click sequence, the subject could determine when he was being sufficiently quiet.) Second, only the last 50 of 54 click presentations were used to produce the averaged response to the click. Third, although the bandwidth of the time samples was 10 kHz, only spectral components below 3 kHz were considered. Finally, only 20 ms of the averaged waveform was subjected to spectral analysis, and that segment began 5 ms into the computer file (i.e., beginning 9 ms following click presentation).

The pre-drug baseline measurements also included psychophysical determinations of sensitivity in the quiet for test frequencies covering the range from 250-6000 Hz. The psychophysical procedure used was adaptive, two- interval forced-choice with feedback. The three-down/

one-up rule was used to track the 79% point on the psy- chometric function (Levitt, 1971). The signal level was initially adjusted in 4-dB steps, becoming 2-dB steps after the second reversal. Block length was 60 trials, signal duration was 200 ms (including 10-ms rise and decay times),

3461 J. Acoust. Soc. Am., Vol. 95, No. 6, June 1994 D. McFadden and E.G. Pasanen: Otoacoustic emissions and quinine sulfate 3461


and the headphones were TDH-39s mounted in circumau- ral cushions. The trial-timing sequence was warning interval and light ( 100 ms), pause (500 ms), first observation interval and light (200 ms), pause (500 ms), second observation interval and light (200 ms), response interval (1000 ms), and feedback interval and light (300 ms).

C. Dose schedule

The quinine sulfate was administered in standard tab- lets of 325 mg taken at the approximate rate of 1 every 4 h, in the experimenters' presence. Subject MP took three doses total, and subject LB took two doses total. For subject CW [who is not the CW who served in Wier et al. (1988)], two drug runs were conducted. During pass 1, CW's daily dose schedule was one tablet at 12:00 noon, 4:00 p.m., 8:00 p.m., 12:00 midnight, and 8:00 a.m., which corresponded to 1.625 g per 24-h period. During pass 2, CW took only two doses, also separated by 4 h and beginning at noon. Twenty-one days separated the two passes. For subjects MP and LB, the first dose was at 9:00 a.m. Each 325-mg tablet corresponded to about 4.4, 4.2, and 4.6 mg/kg for subjects MP, LB, and CW, respectively.

II. RESULTS

A. SOAE results

1. Decline and recovery of SOAEs

First, we will consider the effects of quinine on SOAEs. In the SOAE figures presented here, the levels of the emissions are expressed in decibels above the average noise floor of our recording system at the frequency of that emission and with the subject in place. These ordinate values can be converted back into SPL by using the correction factors provided in the captions of Figs. 1-4.

In Fig. 1 are the results for subject MP, who had a single SOAE at about 1425 Hz. At the far left are four pre-drug measurements showing that this emission was or- dinarily between 24 and 29 dB above the noise floor (about 12-17 dB SPL). About 6 h after the first quinine dose, the emission was abolished. The emission remained absent un-

til about 34 h after the first dose, or about 26 h after the last dose (the timing of subsequent doses is indicated by the arrows at the top of this and succeeding figures), but the emission did not return to its normal range until about 78 (70) h after the first (last) dose. (The solid diamonds arrayed along the tops of Figs. 1 and 2 indicate times at which data have been omitted in order to simplify the figure; see Sec. II A 3 and Figs. 5 and 6 below.)

In Fig. 2 are shown the SOAE results for subject LB, who had two emissions. For this subject, there was evidence of a decline in both SOAEs within about 2 h of the

first dose, and, just as for subject MP, the emissions were absent beginning about the hour-6 mark. For the stronger, 1895-Hz emission, recovery was apparently complete about 25 h after the first dose, and recovery was also beginning at that time for the weaker, 1420-Hz emission. This faster recovery for LB than for MP is presumably attributable in part to the fact that LB had only two doses of the quinine sulfate and MP had three. While the two

m • 0.0 LU LU -0.5

n- 28 O O -- 24

• 2O O z

> 16 0

< 12

c3 4

SUBSEQUENT DOSES I ' ' I ' ' I .... I ' I ' I ' I ' I/"1 ' I ' I ' .

I ' I ' I ' I ' I -

e---e ---e ............... DATA OMITrED ............ ß ........ ß---ß ........ ß

SOAE: 1425 H

ß

ß

-- vv

SUB J: MP

I I I v. i--v, l--V, v-v--v I I I .. I I .•. I I , , .•e, I , I , , , .. , , , PRE-DRUG 2 4 6 8 10 26 28 30 32 34 49 51 77 79

HOURS SINCE FIRST DOSE

FIG. 1. (bottom panel) The levels of the single spontaneous otoacoustic emission of subject MP before ("Pre-Drug") and after administration of quinine sulfate. Following the first 325-mg dose at hour 0, two subsequent doses were administered 4 and 8 h later, as indicated along the top of the figure. Data have been omitted at the points in time marked with solid diamond symbols; these data are shown in Fig. 5. The levels shown are decibels above the noise floor, which can be converted into sound-pressure level (re: 20/zPa) by subtracting 12 dB. (top panel) Changes in frequency of the SOAE from its pre-drug value. The pair of data points at the far left designates the range of frequency values seen pre-drug.

emissions in LB's ear generally behaved similarly in response to the drug, there were several instances of apparent independence of effect (dissociation).

Subject CW was our first subject in this experiment. He participated in two separate drug runs because, during the first run, some changes in emissions occurred so rapidly that they were missed. (We had established our test

•5 ø O n 'O -1 n- n -2

n- 24 0 0

u_ 20

0 co 12

Lu 8

•3 4

SUBSEQUENT DO•E '• /•' ' I ' I ' I ' I ß' I ' I ' I ' '/•' I '//•' I ' . .

. ' ............. ß ß ß ..... V- -v--v--w-•--- • .............................. v¾ ß ¾,- ............... .

ß - ß .

1420 Hz

' ' I ' I ' I ' I ' I/'t I ' I ' I ' .5' I ' ';t I ' - _

ß .......... •' .... ß ....... DATA ....... ß .... OMITTED ..... ß ......... ß

PRE-DRUG 0

s.-s•895 Hz

ß ß ß

1420 H

SUBJ' LB

I , I , I , I , I ,/.. I , I , I , .C., I , .•.. I , 2 4 6 8 24 26 28 106 128


FIG. 2. Similar to Fig. 1, but for the two SOAEs of subject LB. To convert the ordinate values shown in the bottom panel into SPL, subtract 12 dB for the SOAEs at both 1895 and 1420 Hz. In the top panel, frequency shifts are shown only for the 1420-Hz SOAE; the 1895-Hz SOAE was quite stable in frequency.



• 0.5 0 •o 0.0 n- n Lu -0.5

28

n- 24 0 0

u_ 20

0 z 16

0 rn 12

LU 8

Q 4

SUBSEQUENT DOSES '• (75, 11 5,195Hr) •. I ' I [ ' ' I ' I ' I ' [ . .5' , [ ' '•' I ' I '

--,?•. .... ; .................................................. •- ..... • .... • ......... - ' [ '• • ,,[ [ [ ' • [ , ] , _

1580 Hz .

0 •0 • O• -

.

//•/ SUB J: CW - •/•1 (PASS 1)

ß .½.. • , .½. I , I , 47 73 75

o

1070 Hz o•%... ø ß

2015 Hz AA

ß

4145 Hz

I , I m .•- I , m . PRE-DRUG 3 5 21 23 25


FIG. 3. Similar to Fig. 1, but for subject CW (pass 1). To convert the ordinate values shown in the bottom panel into SPL, subtract 12, 10, 14, and 17 dB for the SOAEs at 1580, 1070, 2015, and 4145 Hz, respectively. Not shown are new, short-lived SOAEs that appeared during this drug episode (see text). (top panel) Changes in frequency from their pre-drug values for four SOAEs. The four pairs of data points at the far left designate the ranges of frequency values seen pre-drug.

schedule, in part, on the basis of our experience with aspirin, which, we now know, acts much more slowly than quinine.) For this and other reasons that will become obvious, the data from these two passes will be presented separately.

Figure 3 shows the SOAE data from pass 1 for subject CW. This subject had four emissions of varying strengths, and on pass 1 he received a total of five doses of quinine sulfate. By 3 h after the first dose, two SOAEs (1070 and 2015 Hz) were already substantially reduced in level-- although the 1070-Hz emission later rebounded toward its normal value as the other emissions were further reduced.

24

o z 16

o

< 12

o 8

' I ] ' I * I ' '' I " [ .*' m m I I m I _ --

o

$

PRE-DRUG 0

o 1580 Hz

•>o_o?<• % øo

• ß o 1070 Hz

,'/ ?•/,•i• ' / •a a 2015 Hz

4145 Hz

SUB J: CW (PASS 2) I , I , I , I , I . .•. I . m .•. I , I .. I , I

2 4 6 8 28 30 52 5•'100 102 HOURS SINCE FIRST DOSE

FIG. 4. Similar to Fig. 3, but for subject CW (pass 2).

All the emissions were absent about 21 h after the first dose

(about 1.5 h after the last dose) and did not show recovery until about 46 h (26.5 h) after the first (last) dose. About 73 h after the first dose, all four emissions appeared to be about normal in level.

During pass 2, which began 21 days after the last dose of pass 1, the onset of effect was not as rapid as during pass 1. We believe that this difference is attributable in large part to the fact that on pass 1 the first dose was taken at noon and no food was consumed for about an hour, while on pass 2 the first dose was taken about an hour after a large breakfast. Presumably, absorption of the drug was retarded in the latter case, explaining the slower time course of decline in the SOAEs. Figure 4 shows that, during pass 2, there was little change in any emission until about 5 h after the first dose, at which time the 1070- and 2015-Hz emissions began to decline. At about the hour-6 mark, all emissions showed some decrease in level, and two emissions were absent about 7.5 h after the first dose. Re-

covery appeared to be essentially complete for all four emissions about 30 (26) h after the first (last) dose. (Note that different pre-drug values are shown for passes 1 and 2, the latter being obtained 10 days after the last dose in pass 1; the two sets of baselines are in good agreement.)

2. Supplementary observations

Frequency changes. There was a general tendency for SOAEs to drift toward lower frequencies as their levels diminished in response to the drug, and toward higher frequencies as their levels returned to normal. Some of these frequency shifts are illustrated in the top panels of Figs. 1-3. For subject LB, frequency shifts are shown only for one SOAE since the other was quite stable, and for subject CW, frequency shifts are shown only for pass 1 since the changes during pass 2 were so similar. Note that for subject CW (pass 1), the frequency of the 1070-Hz emission decreased monotonically even though its level was increasing. None of CW's emissions varied more than a couple of Hertz in the baseline measures. Long et al. (1991) have previously discussed frequency shifts in SOAEs and their relation to amplitude changes.

Order of effect on SOAEs. Previous work in this lab indicated that aspirin affected the weakest SOAEs before affecting the strongest ones (McFadden and Plattsmier, 1984). Here, however, the weakest SOAEs did not neces- sarily respond first to the drug. For example, for CW (pass 1), one strong SOAE (1070 Hz) and one SOAE of me- dium strength (2015 Hz) were both greatly reduced 3 h after the first dose, while another strong emission (1580 Hz) and one quite weak emission (4145 Hz) were mini- mally affected at the hour-3 mark (Fig. 3). Study of Figs. 1-4 reveals that there was also not a simple relationship between emission frequency and the order, or time course, of decline of the SOAEs. By comparison with the present results, aspirin caused the weakest SOAEs to be gone within about 20 h of the first dose, while the strongest SOAEs took 40-70 h (McFadden and Plattsmier, 1984). While less can be gleaned from the present data about the time courses of recovery for SOAEs of differing initial mag-



nitude, there is a suggestion for both LB and CW (pass 1 ), that the initially strongest SOAEs recovered fastest. No clear-cut pattern of recovery could be determined for aspirin either. By comparison, the onset and rate of recovery of SOAEs from quinine may have been slightly more rapid than for aspirin; of course, many fewer quinine doses were given.

If we make the reasonable presumption that the SOAEs continued to originate from the same specific sets of (subtly abnormal?) cells throughout the drug episode, then the frequency shifts observed must be attributable to small, quinine-induced retunings of the cochlear structures in the local vicinities of those sets of cells. The primarily downward shifts in frequency suggest that quinine some- how produced changes corresponding to local reductions in the stiffness of the cochlear partition, possibly via changes in the length of outer hair cells (Karlsson and Flock, 1990).

New SOAEs. Subject CW produced new SOAEs during both of his drug runs. New, short-lived SOAEs of this sort have previously been reported during recovery from aspirin administration (McFadden and Plattsmier, 1984). During CW's pass 2, an emission was detected at about 1640 Hz about 2 h after the first dose and then again at about 48 and 96 h after the last dose. The other new

SOAE•at about 1490 Hz--was continuously detectable from about 5 to 26 h after the last dose during pass 1, and then for a short time during pass 2. This latter episode was particularly interesting and deserves special comment.

A short-lived quasi-stable state. At about 4 h after the second dose in CW's pass 2, his sole remaining SOAE began to alternate between two specific frequencies--1565 and 1495 Hz. Initially, the jumps from one frequency to the other were "spontaneous" in the sense that they occurred in the absence of anything we, or the subject, were (aware of) doing to initiate them. The emission would jump to 1495 Hz and dwell there for anywhere from 5- 120 s before returning to 1565 Hz, behavior that is remi- niscent of the quasi-stable-state behavior reported by Burns et al. (1984). The return to 1565 Hz was often accompanied by a large body-noise artifact on the spectrum analyzer. The level of the SOAE was about 12 dB SPL irre- spective of which frequency was evident. During this time, we noticed that the emission often jumped to 1495 Hz when we spoke to the subject over the intercom, so we tried whispering in his contralateral ear, which also produced jumps for a time. Attempts to obtain more objective information about this unusual effect were not particularly suc- cessful. At first, when tones several hundred Hertz lower in frequency than the emission were presented to the contralateral ear using an ER-10 earphone, a jump in the emission to 1495 Hz could be reliably obtained, but not when the contralateral tone was similar to, or above, the frequency of the emission. However, attempts to replicate this observation were inconclusive, as the emission ceased to jump for any reason and remained stable at about 1565 Hz. The contralateral effects make it difficult not to wonder

whether the crossed efferent system might be involved in this interesting behavior (Probst et al., 1991, pp. 2039 and

2047; Mountain, 1980; Mort etal., 1989; McFadden, 1993). It should be emphasized that 1495 Hz is quite close to the 1490 Hz of the newcomer seen during recovery in CW's pass 1.

3. Initializing effects

Reports are common of marked changes in SOAEs within the first half-hour or so following insertion of the probe tip into the external ear canal (e.g., Zurek, 1981; Rabinowitz and Widin, 1984). In a recent study of this effect, Whitehead ( 1991 ) typically observed increases of 2-3 dB in the magnitude of SOAEs over the first 20-30 rain following insertion of the probe, although increases of up to 20 dB were seen on occasion. Frequency changes of 6-10 Hz were also observed, although the direction of frequency shift was not as systematic as the level change. Sometimes, new SOAEs emerged over the course of tens of minutes. On the basis of a number of control conditions,

Whitehead argued that these changes in SOAEs originated not in the equipment or recording environment, but in a "central nervous system influence on the emission generator." Candidates for this CNS influence include changes in the efferent and/or cardiovascular systems (see McFad- den, 1993).

Over the years, we have also observed small "initializing effects" of this sort in our nondrug subjects. However, in the present experiment we saw several extreme instances. On a number of occasions during drug use or recovery, we saw no, or very weak, SOAEs immediately after insertion of the probe, and then a relatively rapid rise of SOAE magnitude over the course of minutes or tens of minutes. In two instances, SOAEs increased more than 20 dB within 15-30 min. A number of these initializing episodes are shown in Figs. 5 and 6 for subjects MP and LB,

rr 24

u_ 20

0 Z 16

Om 12

• 8

• 4

SUBSEQUENT DOSES • I ' I ' I ' ! ' ' I '' ' I '/""' I ' I ' I '/"" I ' I ' I '

SOAE: 1425 Hz SUBJ' MP

I , I • I , I , I , I , .$.. • I , I.•,,! I • I , I ..•.. I , I , I , 0 1 2 3 4 5 33' 34 49 50 51 78 79 80


FIG. 5. Initializing effect for subject MP. The levels of the single SOAE of subject MP during seven time intervals immediately following insertion of the probe tip into the ear canal. Only the solid data points were also shown in Fig. 1. The ordinate values can be converted to SPL by subtracting 12 dB from the values shown. The timing of the subsequent doses is indicated along the top of the figure.



28

rr 24 o o

u_ 20

o z 16

o m 12

• 8

c)

• 4

SUBSEQUENT DOSE,

.

1895 Hz

/

/

/ 1420 Hz

I , I , I , I , I , 1.5. I , I .•.,. , 0 1 2 3 4 5 25 26


SUB J: LB

I .5" I 106 128

FIG. 6. Initializing effect for subject LB. The levels of the two SOAEs of subject LB during six time intervals immediately following insertion of the probe tip into the ear canal. Only the solid data points were also shown in Fig. 2. The ordinate values can be converted to SPL by subtracting 12 dB from the values shown. The timing of the subsequent dose is indicated along the top of the figure.

respectively. No such episodes have ever been observed in subject CW, whose emissions we have studied on numer- ous occasions over the course of many months, with and without drugs. (To avoid possible confusions between initializing effects and drug effects, only the final point or two in each of these initializing runs was plotted in Figs. 1 and 2 for these two subjects.)

The initializing episodes around hours 3 and 128 for subject LB reveal that the levels of multiple emissions in the same ear may not fluctuate in a correlated manner, a fact that will have to be addressed by any proposed explanation of the initializing effect. During the initializing periods, the SOAEs of MP and LB changed only very slightly in frequency (typically downward by less than 0.2% ).3

B. DPOAE results

By positioning the 2fl-f2 distortion product about 50 Hz above an SOAE, both an input/output function for the DPOAE and a suppression function for the SOAE could be determined concurrently as the common level of the primaries was varied. The DPOAE data will be discussed in this section and SOAE suppression in the next. 4

For all three subjects, the general effect of quinine was to displace the DPOAE input/output functions toward successively higher primary levels and to successively reduce their slopes. The input/o•tput functions for subjects MP and CW are shown in the top panels of Figs. 7 and 8. For these data, it was not always possible to find a single level of the primary tones that yielded a nonasymptotic level of the DPOAE for all test times (functions). An alternative summary was obtained by fitting the linear segments of the input/output functions with straight lines, and determining the levels of the primaries required to produce a DPOAE of --3 dB SPL (horizontal cuts at a higher SPL value would have yielded larger effects because

15

o3 10 v

< 5 o

c:) 0

o .• -5

LU -10

-15

15

a_ 10

LU 5

O O3 0 O -• -5

. -10

-15

Pre-drug Pre-drug Subject: MP •' 1.5hr. x 25.3 hr. [] 2.7 hr. ß 28.5 hr.

b 4.5 hr. • o 33.3 hr. 0 6.8 hr. / •, ß 49.7 hr. , •,,

,,•j•'-.--.•,/ ,, 78.1 hr•,•• /M-- (Slope = 1.0) ............................... 0__ •_ •_/_ ..............

•0/, /

(b)

(c) ,% (d)

30 35 40 45 50 55 30 35 40 45 50 55

SPL OF EACH PRIMARY SPL OF EACH PRIMARY

FIG. 7. (a) and (b) DPOAE input/output functions for subject MP. The levels of a 2f•-f2 distortion product at 1475 Hz as a function of the common level of the primary tones. The parameter on the curves is the time, following the first dose, at which the data were collected. For reference purposes, a slope-l.0 line is shown. Abscissa values at the intersections of straight-line fits to the input/output functions and the dashed line at --3 dB were used to produce the summary data shown in Fig. 10 below. (c) and (d) Suppression functions. Levels of an SOAE at 1425 Hz as a function of the common level of the primary tones generating a DPOAE of 1475 Hz. (all) The data are shown as sound-pressure level. Unmeasurable DPOAEs and SOAEs are plotted at the SPL of the noise floor, --12 dB.



PASS ONE PASS TWO .

ß i .... i .... i .... i .... i .... i .... i .... i. ß i .... i .... i.... i .... i ß ß ß ß i i ß ß i i ß ß ß ß i ß

Pre-drug Pre-drug Subject: CW ß

15 [] 3.0 hr. ** 1.0 hr. , ,e-e O 21.8 hr. ,• .,..o,. [] 3.1 hr. ,•*

10 A 27.0 hr. J[ .-."' 13, . O 5.8 hr. O 28.3hr. /' _,.,.•iEI2 / [] : O 28.2hr. • /•'•/- (S,ope=l.0)

5 ß 45.6 hr. [ .'I•E• .-; - ß 29.7 hr. t • ./"

,I .t:-? 'u 0 / ß

-5 .,o 'o o/ ß (a) (b) -15

15

10

-10

-15

(c) (d)

** •3 0 AOA .

'.1 .... i .... I .... i .... i .... I .... i .... i.- '. i .... I .... i .... i .... i .... i .... i .... i.'

20 25 30 35 40 45 50 55 20 25 30 35 40 45 50 55

SPL OF EACH PRIMARY SPL OF EACH PRIMARY

FIG. 8. Similar to Fig. 7, but for subject CW (pass 1 on the left and pass 2 on the right). The 2fl--f2 distortion product was at 1625 Hz, and the SOAE at 1580 Hz. Different pre-drug baselines are shown for the two passes (see text). To convert the SPLs shown into decibels above the noise floor, add 12 dB.

of the reduced slopes of the input/output functions obtained under quinine). (Those summaries of the DPOAE data are shown in the top panels of Figs. 10-12.) The slopes of the least-squares fits to the linear segments of the pre-drug DPOAE input/output functions were 1.2, 1.1, 3.4, and 1.8 dB for MP, LB, CW (pass 1), and CW (pass 2), respectively. By comparison, Wier et al. (1988) reported slopes ranging from about 1.0 to 2.1 dB/1 dB increase in the level of the primary tones. When the pre-drug DPOAEs were --3 dB SPL, they were about 43, 40, and 36 dB down from the primaries for subjects MP, LB, and CW, respectively.

As can be seen from the top panel of Fig. 7, for subject MP, the DPOAE input/output function was shifted by about 4 dB toward higher primary levels about 2.7 h after the first dose. Then, beginning about 6.8 h after his first dose, his DPOAE could not be detected, and this remained true until 33.3 (25.3) h after the first (last) dose. Recovery was not complete until about 78 (70) h after the first (last) dose for this subject (see top panel of Fig. 10). Thus, for subject MP, the time courses of decline and recovery were similar for SOAE and DPOAE (compare Figs. 1 and 10).

The DPOAE input/output functions of subject CW, like his SOAEs, behaved somewhat differently during his two drug episodes. Another difficulty was that the no-drug input/output function measured just prior to pass 2 was shifted slightly toward weaker primary levels from the function measured just prior to pass 1 (determined about 12 days earlier). In ignorance of the basis for this shift, it seemed prudent to use as the pre-drug baseline, the function determined closest in time to each drug episode. Ac-

cordingly, different "pre-drug" functions appear in the two top panels of Fig. 8. During pass 1 [Fig. 8 (a)], CW's input/output functions were displaced toward higher primary levels beginning at least 3 h after the first dose, but like LB (and unlike MP), CW's functions never became flat---even after five doses. Recovery was apparently complete sometime between about 46 and 73 (26.5 and 53.5) h after the first (last) dose in pass 1. During pass 2 [Fig. 8 (b)], CW's DPOAE input/outpui functions flattened slightly and were displaced about 2 dB toward higher primary levels beginning at hour 5.8. The input/output functions remained slightly flattened for about 30 h after the first dose, and the input/output function obtained 52.2 h after the first dose was displaced slightly toward primary levels that were weaker than those used for the baseline

function, at least at higher levels of the primaries; that is, it was "enhanced." (A summary of the changes in the input/ output functions at the -3-dB intersection points during pass 1 appears in Fig. 12.) In general, the time course of effect on CW's DPOAE at 1625 Hz appeared quite similar to the time course of decline and recovery shown by his SOAE at 1580 Hz.

For subject LB, two pairs of primary frequencies were studied. One pair produced a DPOAE at 1942 Hz and the other at 1626 Hz; that is, one DPOAE was about 50 Hz above the 1895-Hz SOAE and the other about 270 Hz

below it. Input/output functions are not shown for subject LB, for space considerations (but summaries exist in Fig. 11). Half an hour after the first dose, the input/output function for the 1942-Hz DPOAE was shifted about 3 dB

toward lower primary levels (an "enhancement") and was



lower in slope. At the next measurement (hour 5.8) the input/output function was shifted about 8 dB toward higher primary levels and it remained there through the last measurement taken (hour 106). Important to note is that a DPOAE input/output function could be obtained at hour 5.8 even though the SOAEs had been eliminated by that time. The 1626-Hz DPOAE is unable to provide much information about suppression since its lower primary ( 1918 Hz) would produce far more suppression of the 1895-Hz SOAE than the DPOAE. However, the input/ output functions obtained with those primaries are of interest because the 1626-Hz DPOAE was further from an

SOAE than any other DPOAE reported here. 2 Those input/output functions were displaced toward higher primary levels and reduced slightly in slope, in accord with the general trend. However, unlike the 1942-Hz DPOAE, the 1626-Hz DPOAE had returned to normal by the last measurement session. Thus, for LB, there was evidence of a complex pattern of dissociation among SOAE and DPOAE (and TEOAE) measures (see Fig. 11 ).

increases in suppression. In the pre-drug condition for CW (pass 1), suppression increased by about 8 dB for each 1 dB of increase in the primaries; this fell to about 1.4 dB/1 dB at 3 h after the first dose. For MP, the corresponding numbers were about 5.8 dB/1 dB (pre-drug) and 1.5 dB/1 dB (at hour 33), and for LB, they were about 2.7 dB/1 dB (pre-drug) and 1.8 dB/1 dB (at hour 26.8). By comparison, Wier et al. (1988, Fig. 5) reported slopes ranging from 1.3 to 2.7 dB/1 dB of primary level. Also, Rabinowitz and Widin (1984) reported a change of about 4 dB for each 1 dB of change in the level of a single suppressor tone lying just above the frequency of the SOAE (their Fig. 3, subject PZ-R). However, this latter numerical comparison is potentially highly misleading because, here, the suppressor was not like Rabinowitz and Widin's fixed external

tone, but was a DPOAE whose level was diminished by the drug, as were the level, and frequency, of the SOAE being suppressed. This illustrates the peril that is associated with comparisons of this sort that depend upon the level of the primaries.

C. Suppression of an SOAE by a DPOAE

Because here the DPOAE was typically the frequency component closest to the SOAE, we regard it to be the principal suppressor of the SOAE (see Zurek, 1981; Rab- inowitz and Widin, 1984; Wier et al., 1988). (In control tests conducted on subject CW, we established that the lower primary tone by itself was incapable of producing any more than about 2 dB of suppression of the SOAE over the range of primary levels used here. )

The quinine-induced decreases in the magnitudes of both the SOAEs and the DPOAEs were accompanied by changes in the level of the primaries necessary to attain a fixed level of SOAE suppression. Examples of these suppression functions are shown in the bottom panels of Figs. 7 and 8, for easy comparison with the corresponding DPOAE input/output functions in the top panels of those figures.

The families of suppression curves in Figs. 7 and 8 (and LB, not shown) have different characteristics. For all subjects, the general shape of the suppression functions did clearly change following administration of quinine; they were displaced somewhat to the right and became somewhat lower sloped. In addition, for MP, the curves were displaced downward considerably (as the SOAE lost magnitude), an effect that was much smaller for subjects LB and CW. In part, this difference is attributable to the times in the drug episodes when suppression measurements were made. For LB and CW, the suppression measurements were, perhaps unfortunately, taken at times when the SOAEs were either absent or undiminished by the drug. Note that at hour 45.6 of pass 1, CW's SOAE at 1580 Hz was essentially normal in level (Fig. 3), and the DPOAE input/output function was only slightly displaced from the baseline [Fig. 8 (a) ], yet the Suppression function was still markedly abnormal [Fig. 8 (c) ].

In general, once the level of the primary tones (actually, the DPOAE) became high enough to initiate suppression, further small increases in the primaries led to large

1. Effectiveness of suppression

Because the magnitudes of the SOAEs and the DPOAEs were both changing, at different rates, following administration of quinine sulfate, it is difficult to determine by visual inspection of the suppression and input/output functions (bottom and top panels, respectively, of Figs. 7 and 8) whether, or how, the "effectiveness of suppression" was altered by the drug. In an attempt to resolve that problem, the following procedure was used. For each measurement period through the course of each drug episode, two values were noted for each setting of the primary tones--the level of the DPOAE produced and the level of the (suppressed) SOAE (top and bottom panels, respectively, of Figs. 7 and 8). The latter values were subtracted from the average magnitude of the (unsuppressed) SOAE at that stage in the drug regimen in order to obtain an estimate of the amount of suppression being produced by the DPOAE. These estimates of SOAE suppression were used as ordinate values and plotted above the corresponding level of the DPOAE then produced by that level of the primaries. To the extent that the same amount of SOAE suppression is produced by the same magnitude of DPOAE at different points in the drug regimen, one could conclude that the effectiveness of suppression is unchanged by the drug and, further, that the DPOAE is the suppressor of the SOAE.

The results of this analysis of the effectiveness of suppression are shown in Fig. 9 for all three subjects. For MP, there was a tendency for the DPOAE to produce more suppression of the SOAE with the drug than without it. A similar effect could be seen for CW (both passes), at least for the data collected within about the first 6 h of the first

dose. For subject LB, there was a clear appearance of greater suppression of the SOAE after the drug than before it, but that impression must be examined with care. Note that the most extreme data points for LB in Fig. 9 came from measurements obtained well into the recovery period. Figure 2 reveals that LB's SOAE at 1895 Hz was back



•. 25

z 20 0

• 15

n 10

• 5

0 0

25

z 20 0

• 15

n 10

• 5

0 0

Subiect: MP

Pre-drug [] V 1.5 hr. [] 2.7 hr. A 4.5 hr. o 33.3 hr.

ø/, ß 78.1 hr. oR

Subiect LB (C) Pre-drug

o 0.5 hr. B 26.8 hr. & 29.8 hr.

• 106.6 hr. . .... i,,,, i,,,,i,,,, i,,,, i,,,, i,,,, i,,,,

-20 -15 -10 -5 0 5 10

Subject: CW (Pass 1)

Pre-drug / [] 3.0 hr. 4 ß 45.6 hr. *[

ß 73.Ohr. %,/ D,

[] [] .. [] ..%

(•)

Subject: CW (Pass 2) (d) Pre-drug V

v 1.0 hr. [] 3.1 hr. o De o 5.8 hr. o [] ß O 28.2 hr. ß 29.7 hr.

ß 52.2 hr.

, V *oD,• *

-10 -5 0 5 10 15 20

SPL OF DPOAE SPL OF DPOAE

FIG. 9. Effectiveness of suppression. The relationship between the magnitude of a DPOAE and the corresponding suppression of an SOAE at various times in the drug episodes. See text for explanation.

within its normal range at hours 26.8, 29.8, and 106.6 after the first dose, yet the most errant points in Fig. 9 come from the first two of those conditions. Said differently, LB's SOAE seemed to show greater-than-normal suppression as an aftereffect of exposure to the drug--an aftereffect that had not fully dissipated more than four days after the first dose.

When interpreting these data, it is necessary to keep two facts in mind. First, the frequencies of the SOAEs often drifted downward slightly following administration of the drug (see Sec. II A 2), meaning that the frequency separations between the SOAE and its DPOAE suppressor were not constant across all measurement periods shown in Figs. 7 and 8. Second, as the level of the primaries was increased (within a given measurement session), the frequency of the SOAE was invariably displaced even further downward; that is, there was a "repulsion" from the DPOAE frequency of as much as --0.7%. Presumably, these increased frequency separations contributed to slight reductions in the power of the DPOAE to suppress the SOAE. The increased frequency separations also reveal that the suppression reported here was not based on the entrainment observed by van Dijk and Wit (1988).

O. Click-evoked OAE results

The averaged click-evoked responses (see Sec. I B above) were analyzed spectrally using the computer pro- gram w^v^x developed by W. S. Geisler. In order to identify spectral peaks unequivocally associated with the nonlinear active mechanism (Davis, 1983; Probst et al., 1991 ), the baseline averaged time waveform obtained with the click at one level was doubled in amplitude, inverted, and

added to the averaged time waveform obtained with a click 6 dB more intense. This procedure functions to eliminate from the spectrum any frequency components that are growing linearly with stimulus level (e.g., Probst et al., 1991 ). Only frequency components that appeared nonlinear by this preliminary analysis of the pre-drug baseline data were studied in the click-evoked responses obtained during drug use, and of greatest interest were peaks at frequencies other than SOAE frequencies. While spectra were always obtained for several levels of the 100-/•s click, for the presentations shown here, only spectra obtained with a single click level (43 dB) were used. The spectra produced across the drug episode by that single click level were examined for changes in the amplitudes (linear plus nonlinear) of the previously identified "nonlinear" spectral components. The w^v^x analysis yielded amplitude values in 25-Hz bins (re: 0.25 Hz for SOAEs). From each spectrum, the largest amplitude peak in any of the bins within one bin of the bin covering the nominal frequency of each spectral component of interest was taken as the estimate of that spectral peak at that point in the drug episode. The results are shown in the top panels of Figs. 10-12, where increases in the common level of the two primary tones necessary to produce the same magnitude of DPOAE, and decreases in the magnitude of peaks in the TEOAE spectra obtained with the same click level are both plotted downward in order to indicate a weakening of those emissions.

The time course of change in click-evoked spectral peaks was best documented in subject MP, who, with only one SOAE, offered the best chance to find a nonlinear peak



SUBSEQUENT

DOSES •f • .",//' / •' / •' I ß I ß I ß I ' I ß , ' I ß I ' I ' I ' I •'• I ' I/'! I ' I '

-4 -

a_ .... ';fi: .................................................................. .: ......

z • • •", ', ..• ....... ;'•'• O 12 (a) •7-n .... • _ '• . • D 1425Hz • -- [no mea•ra•n•apoae) • i , i [ i i i i i i i i //I i i i i i , i ] i// i i i //i i i i

• (b)

• o ............ .• .................. :A ...................................

• 500

• 12 D 10• n 2o• SUB J: MP

18 , • so• . I , i , i , i , i ,,•x, i , i , i . i , ixx i , i xvl , i , 0 2 4 6 8 10 26 28 30 32 34 49 51 77 79


FIG. 10. Effects of quinine on TEOAE, DPOAE, and hearing level for subject MP. (a) Decreases (relative to the pre-drug baseline) in the magnitude of two nonlinear peaks in the (click-evoked) TEOAE spectrum are plotted downward (squares), as are increases (relative to the pre-drug baseline) in the SPL of the primary tones necessary to produce a DPOAE of --3 dB SPL (triangles). The spectral peak at 1425 Hz coincided with an SOAE. A click level of approximately 43 dB corresponding peak SPL was used to collect all TEOAE data shown. (b) Losses in heating in the quiet (relative to pre-drug baselines) are plotted downward. The standard errors of the mean difference hovered between

about 2.5 and 3.5 dB for the 250- and 500-Hz test frequencies, but were generally less than 1.0 dB for the other frequencies.

remote from an SOAE. Shown in Fig. 10(a) are data for one spectral peak at the frequency of MP's 1425-Hz SOAE and another 400 Hz below it. The spectral peak at 1425 Hz appeared to be affected more than the one at 1025 Hz, and with a time course more similar to that of the DPOAE

(and SOAE; compare Fig. 1 ). Karlsson et aL ( 1991 a) also reported a quinine-induced decrease in the magnitude of a

-4

z ,9, 8 o 12

SUBSEQUENT • DOSE • I , , I , I , I , I , I// I ' I ' I , I/•// I//// I

A SPECTRAL PEAK

AT 1100 Hz - _[] ............ 13 ................... _:. __. _ • _--_-_ _-_ _"_ _-_ _- ............... ,• ........... .... '•7', [] ..... / 1626Hz / NN.• // A/' DPOAE

i • //

(a)

! . [ . ! . , . l/•/ J . I . I . I /•,// . I/•/ . I

:• (b) ß

f: ' o .... .................... • 6h •25 o • / • i u •ooo • / • 12 • •oo • ••

• A •oo A SUB J: LB / i , i , i , i , i •/i , i . i . i • i , • i ,

0 2 4 6 8 24 26 28 30 106 128


FIG. 11. Similar to Fig. 10, but for subject LB. (a) The 1942-Hz DPOAE was about 50 Hz above an SOAE and the 1626-Hz DPOAE was about

206 Hz from the nearer SOAE. (b) For the hearing measures, there were only four conditions in which the standard errors of the mean difference were greater than 1.5 dB, and the remainder averaged 1.01 dB.

click-evoked response and a shift of relative spectral strength toward lower frequencies. (The reader will recall that all DPOAE summary values shown here are conser- vative because they were obtained from cuts at --3 dB SPL on input/output functions whose slopes were generally decreasing. )

The click-evoked OAEs for subject LB were less frequently monitored than those for subject MP, but, from the information available, it appears that his data were rather different from subject MP's. The magnitude of the spectral peak at 1100 Hz alternately declined and recovered during the first few hours after the first dose, and then appeared to be fully recovered at hour 29.5 (25.25) after the first (last) dose. Thus this TEOAE peak followed a markedly different time course from LB's DPOAEs and SOAEs (compare Fig. 2). Long et al. (1986) also reported different time courses for SOAEs and TEOAEs following administration of aspirin.

For subject CW, the click-evoked emissions were affected more on pass 1 than on pass 2 (not shown), just as was true for his DPOAEs, and, again, this difference was surely attributable to the difference in number of doses taken. In Fig. 12 (a), pass- 1 data are presented for a nonlinear peak at 1225 Hz, which was about 155 Hz from the nearest SOAE.

E. Hearing loss results

Of all the measures reported here, it may be most difficult to extract across-subject generalizations from the data on hearing sensitivity. A number of factors surely contributed to this circumstance. Among them are differences in the number of doses administered, differences in the timing of the measurements, the collection of only relatively few blocks of trials per test frequency per test period (because of time limitations), and, apparently, large individual differences in response to the drug. As noted above, each 325-mg tablet administered here corresponded to about 4.4, 4.2, and 4.6 mg/kg of body weight for sub-

0

<• 4 o

•' • 8

(5 < 12 z o

T •- 16 o

o

z

• ø, 6 12

• 18

SUBSEQUENT •/ /•.(7.5.11.5, 19.5 Hr) DOiSES: i i//' ,/•'", ' ' I ' Ill , I ' I ' I I [ i , i ,

ß

D"-----___.i_ I .......... /_ ........ D DPOAE SPECTRAL PEAK AT laa5 Hz

(a) '' ' ' '' '.5'' ' ' ' ' ' ' '.•"' '.5"'' '

(b) I:]\

____ - . %,. ' ß 500 • - '

, [] / ß ; : 260• ß ?SUB J: CW(PASS 1) . • , , , . , . ,.;., , . , . , . ,.$-, , ..½.[, [ .-

0 2 4 6 22 24 26 28 47 73 75


FIG. 12. Similar to Fig. 10, but for subject CW (pass 1 and left ear). (b) For the hearing measures, only one standard error was larger than 1.9 dB and the remainder averaged 1.4 dB.



jects MP, LB, and CW, respectively. Past research indi- cates that relatively little hearing loss should be expected from such low doses. Alvfin et al. (1991) saw only very small hearing losses with a single dose of 5 mg/kg. At the other extreme, Karlsson et al. (1990) gave doses of 10 mg/kg three times per day for 7-9 days, and observed essentially flat hearing losses ranging from 15-45 dB de- pending upon the plasma concentration and individual subject.

Two of the present subjects did show some hearing loss at at least one test frequency within 6-8 h after the first dose. For subject MP [Fig. 10(b)], there were early losses at all test frequencies (except one) which peaked at 6-10 dB about 8 h after the first dose, and small hearing losses still remained at some frequencies 69 h after the last dose. For subject LB [Fig. 11 (b)], there were early improvements in hearing at some frequencies even though hearing at 6000 Hz was impaired by about 14 dB at hour 7. The subject taking the drug the longest [CW, pass 1; Fig. 12(b)], also showed some early improvements in hearing at two frequencies, but hearing losses of about 2-20 dB then developed by hour 23.5 after the first dose (4 h after the fifth dose). On pass 2 (not shown), subject CW exhibited no early improvements in hearing, but did exhibit losses of up to 11 dB by hour 9 (4.75) after the first (last) dose. 5

While the temporal sampling of data makes it difficult to be certain, it appears from Figs. 10-12 that hearing loss did not occur until well after the distortion-product, click- evoked, and spontaneous emissions had begun to diminish, an outcome that may be attributable to the same mechanisms that caused Alvfin et al. (1991) to observe hearing loss lagging behind quinine plasma concentration. Rela- tively small hearing losses can also coexist with significant decreases in emissions following administration of aspirin (McFadden and Plattsmier, 1984; Long et al., 1986; Long and Tubis, 1988). While some of the heating losses observed here were larger than those observed by Alvfin et al. (1991 ), those authors did use single doses, while our subjects received 2-5 doses. That study also differed from this one in its use of female subjects (see McFadden and Mishra, 1993) and quinine hydrochloride.

F. Phenomenology

Most of the subjects were aware of transient elevations of pulse rate within a couple hours of the first dose of quinine sulfate, and all experienced short episodes of warm-feeling skin or a flushed face. Within a few hours of the first dose, two of the subjects mentioned having a bitter or metallic taste in their mouths, which disappeared within 24 h. Also, beginning about 6-7 h after the first dose, the subjects all reported a weak tinnitus that was variously described as "soft roaring" and "distant waterfall"; for subject MP, this tinnitus became more tonal before disap- pearing. At about this time, subject CW mentioned that some of our test tones sounded distorted or "hollow," and subject LB mentioned that the 200-ms tones used to test for hearing loss sounded as if they had a long reverberation time--"like the 'boing' of sonar pings." These latter effects

were largely gone within about 24 h of the last dose. Dur- ing recovery, two subjects reported hearing mid- or high- pitched tinnitus at about the same time their SOAEs were returning, or new emissions were present, but time did not permit pursuing these possible correlations.

III. DISCUSSION

A. Enhancements

Using a variant of a technique described by Hubbard and Mountain (1983), $typulkowski and Oriaku ( 1991 ) produced an otoacoustic emission similar to those evoked by sound by applying sinusoidal electrical stimulation through a round-window electrode (e.g., Probst et al., 1991 ). Stypulkowski and Oriaku ( 1991 ) found that quinine increased this electrically evoked OAE by about 3 dB. 6 This finding primed us to look for "enhancements" in the various measures obtained in this experiment, and some do exist. For example, ( 1 ) subjects LB and CW (pass 1 ) had heating sensitivity slightly better than baseline levels at some frequencies soon after administration of the drug [Figs. 11(b)and 12(b)]; (2) DPOAE input/output functions obtained within the first hour after drug administration for subjects LB and CW (pass 2) were shifted toward smaller primary levels [Fig. 8(b)]; (3) on pass 1, two of CW's four SOAEs increased in level between hours

3 and 4.5 (Fig. 3); and (4) new SOAEs emerged from the noise floor for CW (both passes).

However, while these examples are of some interest individually, they do not constitute evidence for a generalized enhancement because: ( 1 ) typically, concomitant effects were absent from the other measures obtained at

about the same times in the drug episode, and in some cases, other measures were actually shifted in the opposite direction (dissociated); (2) for LB, heating sensitivity appeared heightened at different test frequencies at different points in the drug episode; (3) CW showed similar DPOAE "enhancements" at 73 h (pass 1) and 52.2 h (pass 2) after the first doses [Fig. 8 (a) and (b) I--long after all quinine effects on SOAEs had dissipated (at least for pass 2); (4) given that noticeably different "baseline" DPOAE functions were obtained before passes 1 and 2 for CW, we are reluctant to argue for the existence of enhancement on the basis of small deviations from one or the other

of those baseline functions; and (5) new SOAEs have been seen with aspirin as well (McFadden and Plattsmier, 1984), and there is no physiological evidence for enhancements with that drug (Stypulkowski, 1989).

The failure to observe an obvious, generalized enhancement could be attributable to any of a number of factors, including individual differences, species differences, mode of drug administration, dose level, and fine- ness of temporal sampling. Another possibility is that the electrically evoked OAE studied by Stypulkowski and Ori- aku (1991) simply behaves differently from other OAEs. Hubbard and Mountain (1990) suggested that the cochlear-amplification function performed by the outer hair cells actually consists of two processes--forward and reverse transduction--and that different ototoxic agents



can differentially affect the two, and, thereby, differentially affect different types of OAE. According to this model, then, there is no necessary reason to expect to see enhancements in other OAEs just because there is an enhancement of the electrically evoked OAE.

Note that between hours 3-5 of CW's pass 1, the level of the SOAE at 1070 Hz (Fig. 3, solid triangles) increased markedly while the levels of the other SOAEs were generally declining. We have observed inversely correlated behavior like this in other settings in the past. It is as if there exists a fixed, common pool of energy against which the various SOAEs in a particular frequency region draw, and any agent that acts to diminish one SOAE provides others with the opportunity to draw more energy and increase in level. The quasi-stable-state behavior observed by Burns et al. (1984) can be interpreted as another form of this effect; the size of the energy pool limits the number of SOAEs exhibited at any given moment to one of the two subsets of possible SOAEs. Also in accord with such a view is the appearance of new SOAEs (Sec. II A 2) once the ability of established SOAEs to draw upon the fixed energy pool has been reduced or eliminated.

B. Dissociations

As noted, the present data contain several instances of apparent dissociations, or partial dissociations, among the various dependent variables. For example, (1) one DPOAE and all other OAEs had fully recovered by hour 106 after the first dose but another DPOAE had not (LB, not shown); (2) a click-evoked spectral peak at 1942 Hz showed a different initial time course of response to the drug than did the SOAEs or DPOAEs [LB, Fig. 11 (a)]; (3) hearing sensitivity and the DPOAE appeared slightly "enhanced" at the same time the SOAEs were diminished

by the drug (LB); (4) one SOAE was increasing in level while another was decreasing [LB, Fig. 2 and CW (pass 1 ), Fig. 3]; (5) DPOAE input/output functions could often be obtained at times when the SOAEs had been abol-

ished [LB, MP, and CW (pass 1 )]; (6) at hour 45.6 of pass 1, CW's SOAE at 1580 Hz was essentially normal in level (Fig. 3), and the DPOAE input/output function was only slightly displaced from the baseline [Fig. 8(a)], yet the suppression function was still markedly abnormal [Fig. 8 (c)]; and (7) at times of no evident hearing loss, the SOAEs or other OAEs could be diminished [LB and CW (pass 1 )]. These facts suggest both that the different types of OAE do not involve exactly the same cochlear structures to the same degree, and that different cochlear (frequency) regions can be affected differently by agents like quinine.

The fact that SOAEs were more vulnerable to quinine than DPOAEs or TEOAEs parallels previous work with aspirin. Similar partial dissociations have been reported by Long et al. (1986), Long and Tubis (1988), Martin et al. (1988), Wier et al. (1988), and Brown et al. (1994). Fur- ther, the general similarity seen here (Figs. 10-12) in the time courses of effect for DPOAEs and TEOAEs (at least for subjects MP and CW) is in accord with the findings of Smurzynski and Kim (1992), who reported significant cor-

relations between these two types of emission obtained from the same normal-hearing ears (compare Lonsbury- Martin et al., 1991 ).

C. Sites of action

As noted in the Introduction, the findings of Karlsson and Flock (1990) and Stypulkowski and Oriaku ( 1991 ) suggest that quinine has the ability to affect the cochlear amplifier (CA) or active process (Davis, 1983), which is widely regarded as involving the outer hair cells.

Stypulkowski and Oriaku ( 1991 ) compared the effects of quinine and salicylate in chinchillas and found that both drugs reduced the AP response and the spontaneous activ- ity recorded with a round-window electrode. However, other effects went in opposite directions with the two drugs. For example, salicylate increased the cochlear microphonic, at least at low sound-pressure levels, whereas quinine decreased the cochlear microphonic across the full range of levels studied. As noted, the onset of effect was always faster for quinine than for salicylate. Puel et al. (1990) studied the effects of salicylate and quinine on the compound action potential (CAP) in guinea pigs. They found that salicylate decreased the amplitude of CAP and increased the latency to the N1 peak, but only at low sound-pressure levels. By contrast, quinine decreased the amplitude and increased the latency at all SPLs tested (also see Jastreboff et al., 1991), although the effects did appear to grow smaller at high SPLs. Since both salicylate and quinine reduce or eliminate OAEs, both drugs clearly operate at a (perhaps different) site crucial to the normal functioning of the CA--a mechanism that operates at low SPLs and clearly involves the outer hair cells. The implication of the Puel et al. results is that quinine operates on mechanisms in addition to the CA, and thus, at sites perhaps unaffected by salicylate. Of course, one obvious can- didate for the additional site is the inner hair cells. Were

they, as well as the outer hair cells, affected by quinine, there should be CAP effects over the entire range of stimulus levels. However, it is also worth noting in this regard that any action that increased the effectiveness of both the medial and lateral components of the efferent system should also affect CAP amplitude and N1 latency at high SPLs as well as low. Also note that it is possible that some of the effects of quinine on the ear may be indirect rather than direct. For example, quinine might produce changes in vascular or perilymphatic pressure, which in turn might produce a change in auditory function. However, the Karlsson and Flock (1990) demonstration of quinine affecting slow motility in outer hair cells shows that, at least in vitro there are direct effects as well as any indirect effects that might be discovered. Both Puel et al. (1990) and Karlsson and Flock (1990) argued that quinine acts on a K + channel in outer hair cells, while the mode of action of aspirin is not yet clear. Jastreboff et al. ( 1991 ) have shown that a calcium-channel blocker can offset some effects of

quinine. Further information about possible sites of action of

quinine comes from Karlsson et al. (1991a) who studied pure-tone audiomerry, psychophysical tuning curves



(PTCs), the acoustic reflex, and click-evoked OAEs in six normal-hearing females. Single doses of quinine hydrochloride of different magnitude were administered during three test sessions. Hearing losses of 15-24 dB were observed, but there were no significant changes in the acoustic-reflex thresholds. The PTCs were reportedly less sharply tuned following quinine administration, an effect commonly seen with hearing loss and alterations in the cochlear amplifier.

D. Initializing effect

There have been previous reports of SOAEs rising in magnitude over the course of minutes following insertion of the probe tip (e.g., Zurek, 1981; Rabinowitz and Widin, 1984; Whitehead, 1991), but the examples shown in Figs. 5 and 6 rank among the largest yet documented. The relationship between these large changes in SOAE magnitude and the typically more subtle ones reported by others is yet to be established, but it may be relevant that our subjects reclined on a small cot during all emissions measurements, while the subjects in the other studies cited were apparently all seated (also see Macrae, 1972). The existence of these initializing effects creates a procedural constraint for future experimenters interested in obtaining accurate estimates of the level of SOAEs. The fact that some subjects show little or no initializing effect might constitute a useful screening criterion for experiments in which it would be inconvenient to allow 15-30 min to elapse prior to the first measurement of each session.

Bell (1992) observed initializing-like increases in SOAE magnitude even after 20 min of recording when the subject's body orientation was changed from lying flat to lying at a 30 ø angle with head down. While many conditions in Bell's experiment are clearly free from concern about initializing effects, others may have been affected by this problem.

K6hler and Fritze (1992) recently reported poor test/ retest consistency in SOAE presence and frequency over inter-test time intervals that averaged 68 months. While SOAEs were unchanged in 3 of the 22 ears tested, they were changed in 14, having decreased in number in 10 of those 14 ears. We believe that this outcome may be attributable to the absence of an initializing period in that experiment ("recording time never exceeded 5 min," their p. 55). There can be considerable variation in the presence and magnitude of SOAEs during the first 5 min of measurement, and this variability would work to produce the appearance of poor reliability. In our experience, SOAE presence and frequency are highly reproducible over long inter-test time intervals (in accord with Probst et al., 1991).

E. Suppression of SOAEs

The suppression of SOAEs by external tones and distortion products has been studied previously (see Probst et al., 1991, p. 2039, for review; Zurek, 1981; Rabinowitz and Widin, 1984; Wier et al., 1988). In general, the most effective suppressor is a tone or distortion product just above the frequency of the SOAE, but suppression grows

more rapidly for suppressors on the low-frequency side of the SOAE than for those on the high-frequency side. Here, for all three subjects there was some evidence of greater suppression of the SOAE by the DPOAE at various times during or after the drug period (Fig. 9), even though the frequency separation between SOAE and DPOAE was typically greater with the drug than without it [counterar- guing a van Dijk and Wit (1988) interpretation]. Had suppression magnitude been constant for a constant DPOAE level, the implication would have been that suppression behaves like a multiplicative reduction in the level of the SOAE, and that the DPOAE is the primary source of the suppression. The deviation from constant suppression re- quires adjustments to this simplest view. 2

While it is unclear to what extent the mechanisms

responsible for SOAE suppression overlap those responsible for psychophysical suppression (e.g., Zwicker and Har- ris, 1990), the latter appear to be more vulnerable to tem- porary and permanent losses of hearing (e.g., Mills, 1982) than was the SOAE suppression measured here. The report of PTCs being broadened following quinine (Karlsson et al., 1991a) is in accord with a loss of psychophysical suppression.

Because the common level of the primary tones used here was always below 55 dB, only the low-level generator of distortion products was activated. The high-level generator seen in animal models (see Whitehead et al., 1992), and the suppression it produces, also deserve study with quinine.

IV. SUMMARY

The main points of this paper are the following. ( 1 ) In addition to its previously known ability to pro-

duce hearing loss, quinine sulfate is capable of reducing or eliminating SOAEs, DPOAEs, and TEOAEs in humans.

(2) Quinine's effects on OAEs can be seen within a couple of hours of the first dose and they generally proceed much faster than the similar effects of aspirin. This suggests that quinine may be a better choice than aspirin when there is a desire to induce a reversible cochlear hearing loss in certain physiological preparations. This rapid onset of effects is in accord with the rapid physiological effects observed by Stypulkowski and Oriaku (1991).

(3) Unlike certain physiological experiments, no con- vincing evidence was found for an enhancement of any OAE (see Sec. IIIA).

(4) SOAEs were greatly reduced or eliminated in all three subjects by about 6 h after the first dose, and it took 1-3 days after the last dose for the SOAEs to return to their normal levels. There was evidence of aftereffects last-

ing as long as 4 days (subject LB, Fig. 2). (5) The time courses of effect on the different OAEs

were slightly different, with the SOAEs generally appear- ing to be most susceptible to the quinine.

(6) Both changes of SOAE frequency with no corresponding change ili level, and changes in SOAE level with no corresponding change in frequency were observed. Also, individual SOAEs in the same ear often behaved oppositely in regard to changes in frequency and/or level.



The tendency of SOAEs to shift downward in frequency following administration of the drug and then upward in frequency during recovery parallels previous drug findings (e.g., McFadden and Plattsmier, 1984; Long et al., 1991 ) and the findings of Furst et al. (1992) following exposure to intense wideband noise.

(7) In several instances, marked increases in the level of an SOAE were observed during the first 15-30 min after a subject was placed in position for measurement (see Sec. III D). We have called this the initializing effect, in part because that term seems neutral theoretically.

ACKNOWLEDGMENTS

This research was supported by a grant from the Na- tional Institute on Deafness and Other Communication

Disorders (DC 00153). We thank Dr. Melinda C. Mc- Michael, of the UT Student Health Center, and Dr. Paul J. Godley of the UT College of Pharmacy for their assistance with medical screening and preparation of the consent forms. Also, Rakesh Mishra and Joy Runnels provided valuable technical assistance. We thank W. S. Geisler for

the use of w^v^x, and for helpful discussions, and R. L. Diehl for use of his computer. G. R. Long, P. J. Jastreboff, and an anonymous reviewer made many valuable com- ments about the initial version of this paper.

•Having such high-frequency resolution is an advantage for some purposes, but it raises a problem when trying to characterize the level of an SOAE. The reason is that many SOAEs appear as relatively broad regions of elevated level rather than as sharp spectral peaks. [It is yet to be determined whether this is because these SOAEs originate from truly narrow-band sources or from sources that are instantaneously sinusoidal but varying slightly in frequency across time, although a report by Rao and Bilger (1991) suggests the latter.] An experimenter desiring to obtain a precise measure of the overall level of such a broad-peaked SOAE would have to sum across this region of elevated frequencies, and across only those frequencies (which raises a difficult problem of definition). If individual labs are doing this, they are not describing the process in their published works. [Some experimenters may have attempted to circum- vent the problem by making measurements with less-than-maximal frequency resolution, some have used the "improved accuracy" mode available on certain spectrum analyzers (G. R. Long, personal communication), and one lab has recently begun to use a curve-fitting procedure to solve the problem (Talmadge et al., 1993).] From reading Methods sections of published papers, one must infer that the typical procedure is simply to take as the level of the emission the level in the most intense frequency bin (whatever its width in the measurement system being used) in the local frequency region of the spontaneous emission. That was the procedure used for the present experiment, meaning that our measures always underestimated the "overall levels" of SOAEs. Because a common reaction to a drug is for an. SOAE to increase in bandwidth as its level is reduced (also see Talmadge et al., 1993), the measurements made following drug administration are more subject to underestimation than are the baseline measurements. Note that a drug- induced change only in the variability of an SOAE's frequency during the measurement period would result in changes in the magnitude of the most intense bin, and thus, to under- or overestimates of its "true" "overall level" at that time. Clearly, comparisons of SOAE levels obtained in different laboratories and with different measurement systems must be made with great care. This general measurement problem also exists for DPOAEs, but to a much lesser extent because their energy is invariably concentrated in only one or two adjacent 0.25-Hz bins and is in a known frequency region. Note that the use of a relatively large frequency bin is not a panacea, for a number of reasons; most notably, the levels of highly narrow-band SOAEs will be overestimated, particularly for SOAEs that are just above the noise floor.

2As an anonymous reviewer has emphasized, the effects of quinine on the

behavior of the DPOAE and on its suppression of the SOAE may have been complicated here by the proximity of the DPOAE to the SOAE. DPOAEs are known to be stronger when in the vicinity of an SOAE (see Probst et al., 1991 ), and strong DPOAEs are thought capable of giving rise to a stimulus-frequency otoacoustic emission (SFOAE; see Probst et al., 1991 ). To the extent that there was a strong SFOAE component in the DPOAE, and that it played a role in SOAE suppression in the pre-drug condition, there was the potential for quinine to alter suppression and DPOAE expression in a number of direct and indirect ways. For example, there may have been a drug effect on the mechanism(s) ( 1 ) producing the DPOAE, (2) producing the additional strength of the DPOAE, (3) producing the SFOAE at the DPOAE frequency, and/or (4) enabling the DPOAE to suppress the SOAE.

3Unfortunately, we have little information about the existence, or magnitude, of initializing effects in these two subjects obtained prior to the administration of quinine. The reason is that standard practice in this lab has long been not to collect emissions data until the subject has been allowed to lie quietly in the test room for at least 15 min after insertion of the probe tip. The large initializing effects shown in Figs. 5 and 6 were discovered when collecting data during the drug episodes, when it was important to obtain SOAE measurements as frequently as possible. What is available are data from test runs conducted several days after the final dose, and these suggest that the drug did magnify the initializing effect, at least for subject LB.

4Initializing effects had minimal impact on the DPOAE and TEOAE data because those conditions were typically not tested until after the subject had been lying in the test room for tens of minutes while SOAEs were being measured. 5In all cases but one, the hearing data shown were obtained from the ear having the SOAEs described above; the contralateral ear was tested for CW (pass 1).

6The increased amplitude of vibration in the cochlear partition observed by Karlsson et al. (1991b) following quinine administration is difficult to interpret as an enhancement of this same type, for a number of reasons. For example: (1) The time course was quite different. Their increase did not begin until 30-50 min after the first administration of the drug, and it often lasted as long as 120 min. (2) The cochlear amplifier was surely not involved. In the Karlsson et al. (1991b) in vitro preparation, the cochlea is isolated from its supply of blood and oxygen, which are known to be crucial to the normal functioning of the cochlear amplifier (e.g., Khanna and Leonard, 1986). Further, Karlsson et al. (1991b) used a sound-pressure level that was too high to allow the cochlear amplifier to reveal itself.

Alvfin, G., Karlsson, K. K., Hellgren, U., and Vil16n, T. (1991). "Hear- ing impairment related to plasma quinine concentration in healthy volunteers," Brit. J. Clin. Pharmacol. 31, 409-412.

Bell, A. (1992). "Circadian and menstrual rhythms in frequency variations of spontaneous otoacoustic emissions from human ears," Hear. Res. 58, 19-100.

Brown, A.M., Williams, D. M., and Gaskill, S. A. (1994). "Acoustic distortion as a measure of frequency selectivity: II. The effect of aspirin," J. Acoust. Soc. Am. (in press).

Burns, E. M., Strickland, E. A., Tubis, A., and Jones, K. (1984). "In- teractions among spontaneous emissions. I. Distortion products and linked emissions," Hear. Res. 16, 271-278.

Davis, H. (1983). "An active process in cochlear mechanics," Hear. Res. 9, 79-90.

Furst, M., Haran, I. R., and Attias, J. (1992). "Manifestations of intense noise stimulation on spontaneous otoacoustic emission and threshold microstructure: Experiment and model," J. Acoust. Soc. Am. 91, 1003- 1014.

Hawkins, J. E. (1976). "Drug ototoxicity," in Handbook of Sensory Phys- iology: Auditory System, Clinical and Special Topics, edited by W. D. Keidel and W. D. Neff (Springer-Verlag, New York), Vol. 3, pp. 707- 748.

Hubbard, A. E., and Mountain, D.C. (1983). "Alternating current de- livered into the scala media alters sound pressure at the eardrum," Science 222, 510-512.

Hubbard, A. E., and Mountain, D.C. (1990). "Haircell forward and reverse transduction: Differential suppression and enhancement," Hear. Res. 43, 269-272.

Jastreboff, P. J., Brennan, J. F., and Sasaki, C. T. (1991). "Quinine-



induced tinnitus in rats," Arch. Otolaryngol. Head Neck Surg. 117, 1162-1166.

Karlsson, K. K., and Flock, A. (1990). "Quinine causes isolated outer hair cells to change length," Neurosci. Lett. 116, 101-105.

Karlsson, K. K., Berninger, E., and Alvfin, G. (1991a). "The effect of quinine on psychoacoustic tuning curves, stapedius reflexes and evoked otoacoustic emissions in healthy volunteers," Scand. Audiol. 20, 83-90.

Karlsson, K. K., Hellgren, U., Alvfin, G., and Rombo, L. (1990). "Au- diometry as a possible indicator of quinine plasma concentration during treatment of malaria," Trans. R. Soc. Trop. Med. Hygiene 84, 765-767.

Karlsson, K. K., Ulfendahl, M., Khanna, S. M., and Flock, 3,. (1991b). "The effects of quinine on the cochlear mechanics in the isolated temporal bone preparation," Hear. Res. $3, 95-100.

Khanna, S. M., and Leonard, D. G. B. (1986). "Measurement of basilar membrane vibrations and evaluation of the cochlear condition," Hear. Res. 23, 37-53.

K/Shler, W., and Fritze, W. (1992). "A long-term observation of spontaneous otoacoustic emissions (SOAEs)," Scand. Audiol. 21, 55-58.

Levitt, H. (1971). "Transformed up-down methods in psychoacoustics," J. Acoust. Soc. Am. 49, 467-477.

Long, G. R., and Tubis, A. (1988). "Modification of spontaneous and evoked otoacoustic emissions and associated psychoacoustic microstructure by aspirin consumption," J. Acoust. Soc. Am. 84, 1343-1353.

Long, G. R., Tubis, A., and Jones, K. (1986). "Changes in spontaneous and evoked otoacoustic emissions and corresponding psychoacoustic threshold microstructure induced by aspirin consumption," in Periph- eral Auditory Mechanisms, edited by J. B. Allen, J. L. Hall, A. Hubbard, S. T. Neely, and A. Tubis (Springer-Verlag, Berlin), pp. 213-230.

Long, G. R., Tubis, A., and Jones, K. (1991). "Modeling synchroniza- tion and suppression of spontaneous otoacoustic emissions using van der Pol oscillators: Effects of aspirin administration," J. Acoust. Soc. Am. 89, 1201-1212.

Lonsbury-Martin, B. L., Whitehead, M. L., Henley, C. M., and Martin, G. K. (1991). "Differential effects of sodium salicylate on the distinct classes of otoacoustic emissions in rabbit and in monkey," Assoc. Res. Otolaryngol. Abstr. 14, 67.

Macrae, J. H. (1972). "Effects of body position on the auditory system," J. Speech Hear. Res. 15, 330-339.

Martin, G. K., Londsbury-Martin, B. L., Probst, R., and Coats, A. C. (1988). "Spontaneous otoacoustic emissions in a nonhuman primate. I. Basic features and relations to other emissions," Hear. Res. 33, 49-68.

McFadden, D. (1993). "A speculation about the parallel ear asymmetries and sex differences in hearing sensitivity and otoacoustic emissions," Hear. Res. 68, 143-151.

McFadden, D., and Mishra, R. (1993). "On the relation between hearing sensitivity and otoacoustic emissions," Hear. Res. 71, 208-213.

McFadden, D., and Pasanen, E.G. (1991). "Quinine sulfate reduces otoacoustic emissions," J. Acoust. Soc. Am. 90, Pt. 2, 2289 (A).

McFadden, D., and Plattsmier, H. S. (1984). "Aspirin abolishes spontaneous oto-acoustic emissions," J. Acoust. Soc. Am. 76, 443-448.

Mills, J. H. (1982). "Effects of noise on auditory sensitivity, psychophysical tuning curves, and suppression" in New Perspectives on Noise-

Induced Hearing Loss, edited by R. P. Hamernik, D. Henderson, and R. Salvi (Raven, New York), pp. 249-263.

Mott, J. B., Norton, S. J., Neely, S. T., and Warr, W. B. (1989). "Changes in spontaneous otoacoustic emissions produced by acoustic stimulation of the contralateral ear," Hear. Res. 38, 229-242.

Mountain, D.C. (1980). "Changes in endolymphatic potential and crossed olivocochlear bundle stimulation alters cochlear mechanics," Science 210, 71-72.

Probst, R., Lonsbury-Martin, B. L., and Martin, G. K. (1991). "A review of otoacoustic emissions," J. Acoust. Soc. Am. 89, 2027-2067.

Puel, J-L., Bobbin, R. P., and Fallon, M. (1990). "Salicylate, mefena- mate, meclofenamate, and quinine on cochlear potentials," Otolaryngol. Head Neck Surg. 102, 66-73.

Rabinowitz, W. M., and Widin, G. P. (1984). "Interaction of spontaneous otoacoustic emissions and external sounds," J. Acoust. Soc. Am. 76, 1713-1720.

Rao, P., and Bilger, R. (1991). "Spectral instabilities of spontaneous otoacoustic emissions," J. Acoust. Soc. Am. 90, 2289 (A).

Smurzynski, J., and Kim, D. O. (1992). "Distortion-product and click- evoked otoacoustic emissions of normally-hearing adults," Hear. Res. $8, 227-240.

Stypulkowski, P. H. (1989). "The mechanism of salicylate ototoxicity," unpublished doctoral dissertation, University of Connecticut, Hartford, Connecticut.

Stypulkowski, P. H., and Oriaku, E. T. (1991). "A comparison of the ototoxic effects of salicylates and quinine," Assoc. Res. Otolaryngol. Abstr. 14, 79.

Talmadge, C. L., Long, G. R., Murphy, W. J., and Tubis, A. (1993). "New off-line method for detecting spontaneous otoacoustic emissions in human subjects," Hear. Res. 71, 170-182.

van Dijk, P., and Wit, H. P. (1988). "Phase-lock of spontaneous otoacoustic emissions to a cubic difference tone," in Basic Issues in Hear- ing, edited by H. Duifhuis, J. W. Horst, and H. P. Wit (Academic, New York), pp. 101-105.

Webster, L. T. (1990). "Drugs used in the chemotherapy of protozoal infections," in Goodman and Gilmans' The Pharmacological Basis of Therapeutics, edited by A. G. Gilman, T. W. Rall, A. S. Nies, and P. Taylor (Pergamon, New York), 8th ed., pp. 978-998.

Whitehead, M. L. (1991). "Slow variations of the amplitude and frequency of spontaneous otoacoustic emissions," Hear. Res. $3, 269-280.

Whitehead, M. L., Lonsbury-Martin, B. L., and Martin, G. K. (1992). "Evidence for two discrete sources of 2f•- f2 distortion-product otoacoustic emission in rabbit: I. Differential dependence on stimulus pa- rameters," J. Acoust. Soc. Am. 91, 1587-1607.

Wier, C. C., Pasanen, E.G., and McFadden, D. (1988). "Partial dissociation of spontaneous otoacoustic emissions and distortion products during aspirin use in humans," J. Acoust. Soc. Am. 84, 230-237.

Zurek, P.M. (1981). "Spontaneous narrowband acoustic signals emitted by human ears," J. Acoust. Soc. Am. 69, 514-523.

Zwicker, E., and Harris, F. P. (1990). "Psychoacoustical and ear canal cancellation of (2f•- f2)-distortion products," J. Acoust. Soc. Am. 87, 2583-2591.



Documents

Otoacoustic emissions and quinine sulfate - UT … emissions and quinine sulfate a) ... The pre-drug baseline measurements also included psychophysical determinations of sensitivity