Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
J Am Acad Audiol 3: 324-330 (1992) Effects of Cochlear Hearing Loss on the ABR Latencies to Clicks and 1000 Hz Tone Pips Cynthia G. Fowler* t Charlene M. Mikami* Abstract High-frequency hearing losses can substantially confound the interpretation of clickevoked auditory brainstem responses (ABRs) . One method proposed to circumvent the problem is to use frequency-specific tone pips to stimulate equivalent areas on the basilar membrane in normal and pathologic groups, or in both ears of asymmetrically impaired patients . This retrospective study investigated the ABRs to clicks and 1000-Hz tone pips from 90 asymmetrically impaired subjects with cochlear pathology. The 4000 Hz threshold significantly affected the wave V latencies from both clicks and tone pips for the most severely impaired subjects . The wave V latencies were highly correlated between ears for both stimuli, but the correlation was higher with the 1000-Hz tone pips . It was concluded that the use of 1000-Hz tone pips can supplement the interpretation of click-evoked ABRs, particularly in patients whose 2000- and/or 4000-Hz thresholds are worse than 75 dB HL . Key Words: Auditory brainstem response (ABR), cochlear hearing loss, wave V latency, asymmetrical hearing ne difficulty in obtaining auditory brainstem responses (ABRs) from patients 0 with high-frequency hearing losses is that the hearing loss shapes the effective spectrum of the click. The result is that different areas on the basilar membrane are stimulated depending on the configuration and degree of the hearing loss (Aran et al, 1975 ; Elberling and Salomon, 1976 ; Coats and Martin, 1977) . The different areas of the basilar membrane, in turn, produce different latencies, which leads to a diagnostic dilemma when determining whether a delayed latency is due to a cochlear or retrocochlear pathology, especially in cases in which wave I is absent. Difficulties are encountered in comparing ABR latencies of hearing-impaired patients to norms collected on normal-hearing subjects, as well as in comparing interaural latencies of patients with unilateral or asymmetrical hearing losses . Various methods have been suggested to compensate for the delay caused by cochlear *VA Medical Center, Long Beach, California t University of California, Irvine, California Reprint requests : Cynthia G . Fowler, Audiology-126, VA Medical Center, 5901 East Seventh Street, Long Beach, CA 90822 324 pathology, although none has achieved widespread use because of relatively poor predictive results for individual cases (Fowler and Durrant, in press) . These methods include adding a correction factor to the click-evoked latencies (Selters and Brackmann,1977 ; Jerger and Mauldin, 1978 ; Hyde and Blair, 1981) and using electrocochleographic methods to elicit wave 1 (Durrant, 1986 ; Ferraro and Ferguson, 1989). An alternative method suggested for circumventing the problems ofclick stimuli in hearingimpaired subjects is to use tone pips in an attempt to stimulate equivalent areas of the basilar membrane in normal and pathologic ears (Fowler and Noflsinger, 1983 ; Telian and Mleny, 1989). The following is a retrospective study of patients seen for routine ABRs in the Audiology Section at the VA Medical Center, Long Beach, California in 1983 and 1984 . The purpose of the investigation was to evaluate the clinical utility of 1000-Hz tone pips and clicks for eliciting the ABR in patients with asymmetrical hearing loss presumably of cochlear origin . METHOD S ubjects were 90 patients referred for ABR testing because of asymmetrical hearing Cochlear Hearing Loss and ABR/Fowler and Mikami loss or tinnitus . The subjects were male and ranged in age from 21 to 82 years, with a mean age of 56 years. The results of audiologic evaluations, consisting of air- and bone-conduction thresholds at octave intervals from 250 to 8000 Hz, word recognition at PBm. and 90 dB HL, tympanometry, and acoustic reflex thresholds and adaptation, were consistent with cochlear pathology. Subjects with confirmed retrocochlear pathology were not included in the subject sample . Stimuli were rarefaction clicks and 1000Hz tone pips . Clicks were produced from 100psec rectangular pulses through matched TDH39 earphones in MX-41/AR cushions . The clicks were delivered to 30 subjects each at 75, 85, or 95 dB nSL (normalized sensation level*), which were equivalent to 105, 115, and 125 dB peSPL (peak equivalent sound pressure level) . The clicks were presented at equal levels for both ears at an overall level 2!15 dB above the audiometric threshold at 4000 Hz in the worse ear or at 95 dB nSL if the 4000-Hz threshold was >_80 dB HL . The choice of click level, therefore, ensured that the high-frequency audiometric thresholds would be higher for subjects receiving the higher level stimuli. The three subject groups are referred to by click level (i .e ., the 75 group, 85 group, and 95 group) . Tone pips at 1000 Hz (1-msec linear rise-fall times, 0msec plateau, fixed onset phase) were delivered at 75 dB nSL (100 dB peSPL) to all subjects . This level was ?15 dB above the 1000-Hz threshold in the poorer ear. The acoustic spectra for the clicks and 1000-Hz tone pips are given in Figure 1. The ABR was recorded with gold cup electrodes attached to the skin at the vertex (noninverting input), ipsilateral earlobe (inverting input), and at the forehead (ground) . Interelectrode impedance was !_5000 Q and equal (±1000 S2). Physiologic activity was filtered (Nicolet, Model 501-A) between 150 and 3000 Hz (3-dB down points, 12-dB/octave rejection rates) . Responses were averaged (Nicolet, Model 812) over 3000 trials in a time window of 10 .48 msec . All responses were replicated to ensure reliability . Latencies were measured from stimulus onset to the positive peak of the wave . Averaged latencies from the two replications were used in the data analysis . *Sensation level, as defined by Sonn (1969), is "the pressure level of the sound in decibels above its threshold of audibility forthe individual subject or for a specified group of subjects ." The prefix "n" denotes that a group of normal hearing subjects is used as the referent . 0 -20 -40 m .60 v w 0 -ao F0 ~ I J a a -20 -40 -60 -a0 1 i 0 2500 5000 7500 10000 FREQUENCY (Hz) Figure 1 The average acoustic spectra (N = 32) for the clicks (upper panel) and 1000 Hz tone pips (lower panel) . Spectra were obtained through TDH-39 earphones with a spectrum analyzer (Brdel & Kjar, Type 2033). RESULTS he mean audiometric thresholds for ocT tave intervals from 500 to 4000 Hz for the 30 subjects in each group are shown in Figure 2. Each group was characterized by a sloping highfrequency hearing loss, and as expected, this slope was greater for the subject groups requirPOORER EAR BETTER EAR ~I I 75 GROUP 75 GROUP 85 GROUP 85 GROUP 95 GROUP ~ 500 1000 2000 4000 95 GROUP 500 1000 2000 4000 FREQUENCY(Hz) Figure 2 The means (squares) and ranges (bars) for the auditory thresholds for the three groups ofsubjects . Open symbols are thresholds from the better ear and filled symbols are thresholds from the poorer ear. 325 Journal of the American Academy of Audiology/Volume 3, Number 5, September 1992 Table 1 Clicks 1000-Hz Tone Pips Latencies of Waves I and V and the I-V Latency Differences Group Wave 1 Wave V 75 85 95 All 1 .87 (0 .17) 1 .81 (0 .21) 2 .15 (0 .53) 1 .92 (0 .35) 6 .00 (0 .32) 5 .98 (0 .22) 6 .12 (0 .36) 6 .03 (0 .33) 4 .10 (0 .29) 4 .17 (0 .26) 3 .89 (0 .43) 4 .06 (0 .35) 75 85 95 All 2 .87 (0 24) 3 .03 (0 .32) 3 .12 (0 .37) 3 .00 (0 .33) 6 .82 6 .99 7 .15 6 .99 3 .95 (0 .28) 3 .95 (0 .29) 4 .00 (0 .27) 3 .97 (0 .28) (0 .31) (0 .29) (0 .36) (0 .35) 1-V Means and standard deviations (in parentheses) in milliseconds . ing higher stimulus levels . The group means do not demonstrate the asymmetries because different audiometric frequencies were asymmetrical in each subject. Wave V was present in all conditions for all 90 subjects, whereas wave I was absent in 25 subjects with the click stimuli and 31 subjects with the 1000-Hz tone pips . The difference in numbers of wave I present for the two stimulus types, evaluated with a chi-square analysis (Guilford, 1965), was not significant . The 4000Hz thresholds of the subjects missing wave I for either stimulus ranged from 10 to 105 dB HL . Only eight subjects, all in the 95 group, were missing wave I for both stimulus conditions . The mean 4000-Hz threshold for these eight subjects was 86 dB HL, with a range of 65 to 105 dB HL. The means and standard deviations for wave I and V latencies and the I to V latency differences for each subject group and for the total group for both clicks and 1000-Hz tone pips are given in Table 1 . Two-way analyses of variance (ANOVAs) (Northwest Analytical, 1986) with one repeated measure (subject group by stimulus type) indicated the same significant variables for both waves I and V. Significant group differences for the wave latencies (Table 2) precluded pooling of data for the three subject groups for most of the statistical analyses. As expected, the latencies elicited by clicks and 1000-Hz tone pips were significantly different. The interaction between the latencies for subject group and stimulus type was also statistically significant . For wave V, for example, the latency difference between the 75 and 95 groups was 0.12 msec for clicks and 0.33 msec for the 1000-Hz tone pips . The increase in latencies for the 1000-Hz tone pips across subject groups, which was not complicated by different stimu326 lus levels (as were the click latencies), demonstrates that an increasing latency delay was imposed by an increasing hearing loss . A linear regression analysis (Northwest Analytical, 1986) indicated that the latency of wave V varied significantly with the latency of wave I. For all subjects having both waves I and V, the wave I latency contributed less variability to the wave V latency for the clicks (r 2 = 0.20; F [ 1, 158] = 38 .54; p < .01), than for the 1000-Hz tone pips (r2 = 0.41 ; F [1, 147] =100 .52; p < .01) . Bivariate plots of the relations between wave I and V latencies for each subject group and both stimuli are shown in Figure 3. The highest correlations were in the 95 group, for which the hearing loss was the greatest. The wave V latencies from both the clicks and 1000-Hz tone pips were subjected to a linear regression analysis with respect to behavioral thresholds between 500 and 4000 Hz for all three subject groups . The linear regression lines deviated significantly from 0 (Table 3 ) only in the 95 group and only for the 4000 Hz thresholds . These relations are shown in Figure 4, in which the wave V latencies increased with increases in the auditory threshold at 4000 Hz for waves V elicited by both clicks and 1000-Hz tone pips . The 4000-Hz thresholds contributed 41 percent of the variability in the click-evoked wave V latencies, but only 17 percent of the variability in the 1000-Hz tone pip-evoked latencies, indicating the greater effect of the high-frequency hearing loss on the click-evoked latencies. Similarly, only in the 95 group, as the interaural threshold differences increased for both 2000 Hz and 4000 Hz, significant increases occurred in the interaural wave V latency differences from both clicks and 1000-Hz tone pips (see Table 3) . This relation is depicted in Figure 5, in which the interaural wave V latency differences are shown increasing with interaural Table 2 Results of the ANOVAs for the Latencies of Waves 1 and V for Subject Group versus Stimulus Wave I Subject group F Ratio df p Stimulus Interaction 1202 .92 7.04 20 .68 2,177 < ' 01 Wave V Subject group Stimulus Interaction 8 .79 3726 .35 18 .40 2,177 1,177 2,177 < .01 < .01 < .01 1,137 1,137 < .01 < .01 Cochlear Hearing Loss and ABR/Fowler and Mikami TONE PIPS CLICKS B R 2=0 .14 R 2=0 .28 s 7 J} C 75 GROUP 75 GROUP R~=0 .27 Rx=0 .08 1 4000 Hz THRESHOLD (dB HL) . .r " 85 GROUP 85 GROUP R~=0 .50 R x=0 .14 Figure 4 Bivariate plots of the wave V latency and 4000-Hz thresholds from the 95 group. The left panel includes latencies from wave V elicited by clicks and the right panel includes latencies from wave V elicited by 1000-Hz tone pips . The lines through the data are the best fit regression lines. 51 " 2 95 GROUP 95 GROUP 3 4 1 2 3 4 WAVE ILATENCY (ms) Figure 3 Bivariate plots of the wave I and wave V latencies for the 75 group (upperpanels), 85 group (middle panels), and 95 group (lower panels) for the click (left panels) and 1000 Hz tone pip (right panels) stimuli . Lines through the data are the best fit regression lines. threshold differences for both 2000 Hz and 4000 Hz . The 2000-Hz and 4000-Hz threshold differences accounted for 23 percent and 52 percent, respectively, of the variability in interaural latency differences for click responses . In contrast, the 2000- and 4000-Hz threshold differences each accounted for 27 percent of the variability in interaural latency differences for 1000-Hz tone pips . Thus, the wave V latencies from clicks were more dependent on both absolute thresholds and interaural threshold differences than were the wave V latencies from 1000-Hz tone pips, and these dependencies were only significant for the severe high-frequency hearing losses . Of all the frequencies tested, 4000 Hz exerted the greatest effect on the wave V latencies. The wave V latencies between the left and right ears were subjected to a linear regression analysis . For the three subject groups combined, the left and right ears were significantly correlated both for clicks (r2 = 0.67; F [1, 881 = 177.41 ; p < .O1) and for the 1000-Hz tone pips (r' = 0.79; F [1, 881 = 343.93; p < .01) . The lesser interaural correlation for click latencies than for 1000-Hz tone pip latencies is undoubtably a result of the greater dependence of the click latency on the 4000-Hz auditory threshold, at which most of these subjects had asymmetrical hearinglosses . The interaural latency relations for wave V for both clicks and 1000-Hz tone pips are shown in Figure 6. DISCUSSION T he use of tonal stimuli has been proposed as a method for circumventing the latency Table 3 Linear Regression Analyses for the 95 Group Audiometric Thresholds and Wave V Latencies r r2 F of Absolute Click V vs 2000 Hz Click V vs 4000 Hz 1000 Hz V vs 2000 Hz 1000 Hz V vs 4000 Hz 0 .32 0 .64 0 .24 0 .41 0 .10 0 .41 0 .06 0 .17 6 .53 40 .05 1,58 1,58 p> .01 p < .01 12 .16 1,58 p< .01* Relative Click V vs 2000 Hz Click V vs 4000 Hz 1000 Hz V vs 2000 Hz 1000 Hz V vs 4000 Hz 0 .48 0 .72 0 .52 0 .52 0 .23 0 .52 0 .27 0 .27 Comparison 3.52 1,58 p p> .01 8.57 1,28 p< .01* 10 .34 1,28 p< .01* 30 .92 10 .44 1,28 1,28 p < .01 * p< .01* `Significant at p < .01 Absolute refers to the absolute wave V latencies compared to 2000- and 4000-Hz audiometric thresholds, and Relative refers to the interaural wave V latency differences compared to the 2000- and 4000-Hz threshold differences . 327 Journal of the American Academy of AudiologyNolume 3, Number 5, September 1992 0.6 0.4 0.2 0.0 -0 .2 -0 .4 -0 .6 -0 .8 0.4 0.2 0.0 -0 .2 -0 .4 -0 .6 -0 .8 -60 2000 Hz 4000 Hz R2-0 .23 R2=0.52 R2 -0.27 -40 ' .20 0 R2 20 40 -60 .0.27 -40 ' .20 0 20 I 40 60 THRESHOLD DIFFERENCE (dB) Figure 5 Bivariate plots of the interaural wave V latency difference and interaural threshold differences for the 95 group. The latencies from clicks (upper panels) and 1000-Hz tone pips (lower panels) are shown with the interaural threshold differences from 2000 Hz (leftpanels and 4000 Hz (right panels). Best fit regression lines are shown. delays imposed by high-frequency hearing losses on click-evoked ABRs . The present study indicates that this method has some advantages, but does not completely alleviate the problem . The use of the 1000-Hz stimuli was not necessary in the subjects with 2000-Hz and/or 4000Hz auditory thresholds below 80 dB HL . In these subjects, the ABR stimulus was >_20 dB relative to the auditory threshold at 4000 Hz and the click-evoked wave V latencies were within normal limits . In the 95 group, some of the high-frequency hearing losses were sufficiently severe to prolong the wave V latencies elicited by clicks and by 1000-Hz tone pips . The effect of the 4000-Hz thresholds is expected in the click-evoked ABR because neurons that respond to the high-frequency components of the stimuli dominate the latency of the click-evoked ABR (Don and Eggermont, 1978 ; Fowler, in press) . Because N 8 O 7 _E a U Z w fa J CLICKS TONE PIPS 6 55 6 7 8 5 6 LATENCY AS (ms) 7 8 9 Figure 6 Bivariate plots of the latency of wave Vin the left ear (abscissa) and the right ear (ordinate) for the click (left panel) and 1000-Hz tone pip (right panel) stimuli. Regression lines are shown in both panels . the spectral splatter of the 1000-Hz tone pips excites fibers beyond the 1000-Hz area on the basilar membrane in normal ears, a sufficiently severe high-frequency hearing loss can also prolong the wave V latencies to the 1000-Hz tone pips . This finding was consistent with the delay in wave V latencies to 1000-Hz tone pips that was seen in subjects with such severe highfrequency hearing losses that wave V could not be elicited with 4000-Hz tone pips (Fowler and Leonards, 1985) . A persistent basal response bias, therefore, may cause absolute latencies to 1000-Hz stimuli to be delayed in some patients with severe-to-profound high-frequency cochlear hearing losses . The tonal stimuli used in this study had linear ramps . Other ramps, such as the Blackman, may increase the frequency specificity of the stimulus in normal hearing subjects such that the latencies of more patients with high-frequency hearing losses will fall within the norms . Although the latencies of waves I and V were positively correlated, wave I contributed only 20 percent of the variability to the latency of wave V for click stimuli and 41 percent for 1000-Hz stimuli. Among the subject groups, the highest correlations were in the 95 group, in which the high-frequency hearing losses were the greatest . Further, I-V latency differences were shortest for the 95 group with clicks, but equal for all three subject groups with the 1000Hz tone pips . These results are consistent with three previous indications that the click-evoked wave I is derived from higher frequency responses than is wave V. First, Don and Eggermont (1978) demonstrated the different frequency responses of waves I and V by noting that wave I disappeared sooner than wave V as the derived-band ABRs decreased in frequency. Second, Coats and Martin (1977) reported reduced I-V latency differences in some subjects with high-frequency cochlear hearing losses . In these subjects, the click responses were derived from lower frequencies than in normal-hearing subjects. Third, Fowler and Noffsinger (1983) found shorter I-V latency differences in normal subjects for ABRs evoked by lower-frequency tone pips as compared to higher-frequency tone pips . In the latter two cases, a decrease in the effective stimulus frequency caused a greater prolongation of wave I than of wave V. One disadvantage in using the tone pips is the qualitative waveform degradation that can sometimes result from the longer rise time in the tone pips than in the clicks . This degradation is expected to affect elderly subjects and 328 rf~I Cochlear Hearing Loss and ABR/Fowler and Mikami those with hearing losses, because the waveforms in these subjects already have reduced amplitudes relative to those of young normalhearing subjects . This degradation may compromise waveform identification and accuracy in identifying peak latencies, and thus may limit the usefulness of tonal stimuli in some subjects . A further, and perhaps more significant, disadvantage of the use of tone pips is the possibility that low-frequency tone pips may reduce the sensitivity of the ABR to retrocochlear pathology compared to its sensitivity with clicks or high-frequency tone pips . Clemis and McGee (1979) reported a diagnostically significant difference in wave V latencies to tone pips at various frequencies in only 1/17 patients with acoustic neuromas . That case, however, was the only one in which all interaural wave V latency differences were under 1 .0 msec . Retrocochlear pathology that exerts lesser effects on the wave V latencies is more likely to be confused with cochlear pathology, and therefore, is the targeted patient group for the use of tonal stimuli . Fowler and Noffsinger (1983) reported that in retrocochlearly impaired subjects, wave V was more delayed with 4000-Hz stimuli than with 2000-Hz stimuli, and that this discrepancy was exacerbated with faster stimulation rates . This finding suggests that higher-frequency tone pips or clicks may be more sensitive in identifying retrocochlear disorders than are lower-frequency tone pips . Telian and Kileny (1989) found that 1000Hz tone pip ABRs were useful in confirming the diagnosis of acoustic neuroma made by clickevoked ABR in 17 patients with high-frequency sensorineural hearing losses . In many of these patients, the abnormally long interaural wave V latency differences with click stimuli were considered to be at least partially caused by the severity of the high-frequency hearing losses . The abnormally long interaural differences from 1000-Hz tone pips, however, were attributed to neural pathology because the 1000-Hz thresholds were better than the higher-frequency thresholds, and thus, presumably were less likely to produce a cochlear delay . Further studies that compare the success rate in identifying retrocochlear lesions with clicks and tone pips of various frequencies are warranted. In summary, the use of 1000-Hz tone pips to elicit the ABR is helpful in distinguishing between cochlear and retrocochlear pathology in some cases of high-frequency, and particularly asymmetrical, sensorineural hearing loss . Gen- erally, the use of clicks is sufficient if the audiometric threshold at 4000 Hz is less than 80 dB HL and the stimulus is delivered at least 20 dB above the audiometric threshold at 4000 Hz . For patients with more severe cochlear hearing losses, the supplemental use of 1000-Hz tone pips can reduce the number of equivocal ABRs if only click stimuli are used . Still to be determined is the likelihood that cases of retrocochlear pathology will fall within normal limits with low-frequency tonal stimuli, but beyond normal limits with clicks or high-frequency tonal stimuli . Further research is necessary to determine the comparative diagnostic sensitivity of the ABR elicited by clicks and tone pips at various frequencies. Acknowledgment. This study was supported by a Merit Review from the Medical Research Service, Department of Veterans Affairs, Washington, DC . REFERENCES Aran JM, Darrouzet J, Erre JP . (1975) Observation of click-evoked compound VIII nerve responses before, during, and over seven months after kanamycin treatment in the guinea pig. Acta Otolaryngol 79 :24-32 . Clemis JD, McGee T. (1979) Brain stem electric response audiometry in the differential diagnosis ofacoustic tumors . Laryngoscope 89 :31-42 . Coats AC, Martin JL . (1977) Human auditory nerve action potentials and brain stem evoked responses : effects of audiogram shape and lesion location . Arch Otolaryngol 103:605-622 . Don M, Eggermont J. (1978) Analysis of the click-evoked brainstem potentials in man using high-pass noise masking. J Acoust Soc Am 63 :1084-1092 . Durrant JD . (1986) Combined ECochG-ABR versus conventional ABR recordings . Semin Hear (Electrocochleography) 7:289-305 . Elberling C, Salomon G. (1976) Action potentials from pathological ears compared to potentials generated by a computer model. In : Ruben RJ, Elberling C, Salomon G, eds. Electrocochleography . Baltimore : University Park Press, 439-456. Ferraro JA, Ferguson R. (1989) Tympanic ECochG and conventional ABR: a combined approach for the identification ofwave I and the I-V interwave interval . EarHear 10 :161-196. Fowler CG . (in press) Effects of stimulus phase on the normal auditory brainstem response . J Speech Hear Res. Fowler CG, DurrantJD. (in press) The effect ofperipheral hearing loss on the ABR. In : Jacobson J, ed . The Auditory Brainstem Response . Needham, MA: Allyn & Bacon. Fowler CG, Leonards JS . (1985) Frequency dependence of the binaural interaction component of the auditory brainstem response . Audiology 24 :420-429 . 329 Journal of the American Academy of Audiology/Volume 3, Number 5, September 1992 Fowler CG, Noffsinger D. (1983) The effects o£ stimulus repetition rate and frequency on the auditory brainstem response in normal, cochlear-impaired, and VIII nerve/ brainstem-impaired subjects . JSpeech Hear Res 26 :560567. Guilford JP . (1965) Fundamental Statistics in Psychology and Education. New York : McGraw-Hill. Hyde ML, Blair RL . (1981) The auditory brainstem response in neurotology: perspectives and problems . J Otolaryngol 10 :117-125 . Jerger J, Mauldin L. (1978) Prediction of sensorineural hearing level from the brainstem evoked response . Arch Otolaryngol 104:456-461 . Northwest Analytical . (1986) Statpak, V.4.1 . Portland, OR: Northwest Analytical, Inc. Selters WA, Brackmann DE . (1977) Acoustic tumor detection with brain stem electric response audiometry . Arch Otolaryngol 103:181-187 . Sonn M. (1969) Psychoacoustic Terminology. Portsmouth, RI : Ratheon Co . Telian SA, Kileny PR. (1989) Usefulness of 1000 Hz toneburst-evoked responses in the diagnosis of acoustic neuroma. Otolaryngol Head Neck Surg 101:466--471. 330 , '1`'