Acta Oto-Laryngologica
ISSN: 0001-6489 (Print) 1651-2251 (Online) Journal homepage: http://www.tandfonline.com/loi/ioto20
The Effect of Acute Hypoxia on the Latency of the Human Auditory Brainstem Evoked Response Simon Carlile, Daphne A. Bascom & David J. Paterson To cite this article: Simon Carlile, Daphne A. Bascom & David J. Paterson (1992) The Effect of Acute Hypoxia on the Latency of the Human Auditory Brainstem Evoked Response, Acta OtoLaryngologica, 112:6, 939-945, DOI: 10.3109/00016489209137494 To link to this article: http://dx.doi.org/10.3109/00016489209137494
Published online: 08 Jul 2009.
Submit your article to this journal
Article views: 9
View related articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ioto20 Download by: [University of Sydney Library]
Date: 21 March 2016, At: 18:55
Acta Otolaryngol (Stockh) 1992; 112: 939-945
The Effect of Acute Hypoxia on the Latency of the Human Auditory Brainstem Evoked Response SIMON CARLILE, DAPHNE A. BASCOM and DAVID J. PATERSON From the University Laboratory of Physiology, Oxford, England
Downloaded by [University of Sydney Library] at 18:55 21 March 2016
Carlile S, Bascom DA, Paterson DJ. The effect of acute hypoxia on the latency of the human auditory evoked response. Acta Otolaryngol (Stockh) 1992; 112 939-945. Recent studies have shown a decrease in the amplitude and an increase in the threshold of the cat’s auditory brainstem evoked response (ABER) during severe hypoxia (P.0, of 20 to 30 Torr). In this study we have examined the effects of euoxia (end tidal PO, 100 Torr) and mild hypoxia (end tidal PO, of 45 to 50Torr) on the latency of the ABER in 6 human subjects. Hypoxia resulted in a blood 0, saturation of between 75 to 85% and caused a significant prolongation of the latency of wave V of the ABER by 0.185 & 0.045 ms (Mean f S.D; p c 0.01). The prolongation of the ABER during severe hypoxia has previously been attributed to a change in peripheral sensitivity. Using the stimulus level/response latency relationship obtained for each subject under normal breathing conditions, the change in latency produced by mild hypoxia can be interpreted as a mean shift in auditory sensitivity of 5.1 f 3.4 dB. These results suggest that the auditory system is sensitive to much smaller changes in blood 0, saturation than previously thought. Keywords: mild hypoxia, auditory sensitivity.
INTRODUCTION Recent studies of cochlear function have shown that hypoxia reduces a number of cochlear electro-chemical potentials [e.g. (1,2, 3)] thought to be important in maintaining normal cochlear sensitivity. Other studies of auditory system function during hypoxia have used the auditory brainstem evoked response (ABER) and shown that severe hypoxia (blood 0, saturations of 25-50%) produces a reduction in the amplitude (4, 5) and threshold sensitivity (6) of the response in the anaesthetized cat. A recent study in the rat has shown decreases in the threshold sensitivity of the ABER and increases in the latencies of waves I and V of the ABER following a 2-h exposure to 5% 0, in N, (7). In humans, Deecke et al. (8) have reported significant prolongation of the latencies of the middle and long latency components of the auditory evoked potential during hypoxia (alveolar PO, (PA02)of 55Torr). Furthermore, an early audiometric study of threshold sensitivity (9) showed that exposure to hypoxia (9% to 12% 0,in N,) for periods of up to 30 min produced a significant depression in auditory sensitivity. In contrast, Burkett & Perrin (10) failed to find any change in audiometric sensitivity when subjects were exposed for a short time to simulated altitudes of up to 6,100 m. In this study we found that 20 minutes’ exposure to mild hypoxia of between 45 to 50 Torr produced a blood 0, saturation of 75 to 85% resulting in a significant prolongation of the latency of wave V of the ABER. If this change in latency is interpreted as a reduction in auditory sensitivity at the periphery, this prolongation corresponds to an average reduction in auditory sensitivity of 5.1 dB. A preliminary report of these results have been published previously (11).
940 S. Carlile et al.
Acta Otolaryngol (Stockh) 112
Downloaded by [University of Sydney Library] at 18:55 21 March 2016
MATERIAL AND METHODS Under normal (euoxic) conditions there is a systematic reduction in the latency of the ABER with increasing stimulus level (12). The relationship is essentially sigmoidal with the latency changing most rapidly over the middle portion of the dynamic range. The selection of a test stimulus level corresponding to the steepest portion of this relation would provide the most sensitive stimulus for detecting any changes in ABER latency, particularly if these changes result from change in the sensitivity of the auditory system. In this study our strategy was to first determine the stimulus-leve1,kesponse-latencyrelation for each subject, and then to select a test stimulus level based on the criteria outlined above. Using a constant intensity stimulus, ABERs were then determined during euoxia and mild hypoxia. Ten experiments were performed on 6 healthy adult volunteers (2 females and 4 males) ranging in age from 23 to 50 years (mean 36, s.d. 10 years). Subjects reclined comfortably on a couch and were required to remain still with their eyes closed to reduce muscle artifacts. The skin was cleaned to reduce electrical resistance (Omniprep) and the ABER to binaural click stimuli was recorded using three EEG cup electrodes filled with conductive jel (Camjel); one positioned at the vertex and one over each mastoid (left ear active, vertex reference and right ear ground). The signal was amplified (preamp: Iselworth Type A102, high-gain amp: Neurolog NL104), band pass filtered (20 Hz to 5 kHz: Neurolog NL12.5) and delivered to a computer interface (Cambridge Electronic Design: CED1401) for analog to digital conversion. A conventional signal averaging software package (SigAvg: Cambridge Electronic Design) running on a 80286 PC was used to average the responses to between 500 and 2,000 repetitions of the click stimulus presented at 15 pulses per second. The number of repetitions were determined by the amount of averaging required to obtain a clear wave V component in the response (see below). Auditory stimuli were produced using a 50ps transient positive voltage pulse generated by a function generator (Wavetek Model 182A) driving calibrated circumaural headphones (Bayer; DT202). Because of the requirements of the breathing control apparatus, these recordings could not be achieved in an electrically isolated environment. This unfortunately resulted in a relatively noisy EEG signal. In addition, as hypoxia results in increased ventilation, the increased inspiratory effort produced an increase in the respiratory muscle artifact which also contributed to the relatively noisy recordings. Under these recording conditions the latency of the ABER could only be assessed reliably using the latency of wave V (V,J which proved to be the most robust feature of the human ABER [See (13) for wave nomenclature]. The gas delivery system used for producing hypoxia has been described in detail elsewhere (14). In brief, hypoxia was produced by varying the concentration of inspired N2 in air. Each subject had his or her nose occluded and breathed the gas mixture through a mouthpiece. By measuring end tidal O2 partial pressure this mixture was adjusted to result in a PA02of either 50 Torr or 45 Torr. Blood O2 saturations were estimated using a pulse oximeter (PhysiolControl Lifestat 1600) with a finger probe. A P A 0 2 of 50 Torr produced average blood O2 saturation levels of 83.2 f 1.4% (SEM), while reducing PA02to 45 Torr resulted in an average blood 0,saturation of 80.9 5 1.15%. PAO, and PAC02 were held constant throughout the experiment by a computer driven gas mixing system (14). After the stimulus-level/response-latencyrelation was determined, the subject was placed on the breathing apparatus and allowed to breathe normal room air (euoxic condition). Once breathing and blood 0,saturation stabilized, a number of control ABER recordings were taken. This was followed by a period of hypoxia which lasted 30min. During the hypoxic period the ABER was obtained as many times as possible (12 to 22 measures); the precise
ABER latency during acute hypoxia 941
Acta Otolaryngol (Stockh) 112
number of samples were determined by the number of stimulus repetitions required to obtain each sample. For the data analysis, latencies obtained during the period of hypoxia were pooled over three 10-min epochs and compared with the pooled control records.
Downloaded by [University of Sydney Library] at 18:55 21 March 2016
RESULTS The stimulus-level/response-latencyrelation was determined from ABER threshold to between 65 and 75 dBHL in 5 dB steps. Fig. 1 shows the changes in V,, associated with the variation in stimulus level obtained for 3 subjects. V,,, was found to be a systematic function of the stimulus level, showing a decrease of around 1.5 ms over a dynamic range of between 30 dB and 40 dB in stimulus level. Vlat was only measured where wave V could be clearly identified as a prominent peak preceding the large negative component around 9 to 10 ms latency commonly referred to as SN,,[see Fig. 2a and (1311. The change in response latency was most rapid for the middle 20 dB to 30 dB of the dynamic range of the response. The test stimulus for each subject was then selected so as to be centred in this range. In Fig. 2, four ABERs obtained from one subject during a euoxic control period (Fig. 2u) have been compared with 4 ABERs obtained from the same subject during the last 10 min of hypoxia (mean blood 0, saturation: 81.4%; Fig. 2b). Although some aspects of these ABERs are quite variable there is virtually no variation in the V,, within each condition. Fig. 2 shows a significant prolongation in V,,, obtained during hypoxia when compared with the latency of the response obtained during the control period. The latencies obtained during the 30 min of hypoxia were pooled into three 10-min epochs to examine the time course of the hypoxic effect. As the actual latencies recorded were dependent on the test stimulus level, which was varied from subject to subject, the hypoxic latencies were normalized with respect to the control latencies for each subject before pooling for analysis. The hypoxic stimulus produced a significant prolongation of V,,, by an average of 0.184 ms, s.d. 0.045 ms [Fig. 3; p < 0.01: One way ANOVA with repeated measures (1511. There was no significant difference between the control mean V,,, and the mean V , , obtained for the first two 10-min epochs of
7.5-
7.0-
6.5-
6.0 -
5.58
.
I
.
I
. '
.
'
'
Fig. I . Wave V latency (Vlat) is plotted against stimulus level for 3 subjects. Note that the relationship is essentially sigmoidal, changing most rapidly over the 40 to 60 dB range in stimulus level.
' . 0
Downloaded by [University of Sydney Library] at 18:55 21 March 2016
942 S. Carlile et al.
Acta Otolarmeol (Stockhl 112
Hypoxia ms
Fig. 2. (a) Four ABERs obtained during the euoxic control period have been overlaid and the mean latency of V , for all the control records for this subject is indicated by the dotted line at wave V. Note that despite the variation in other components of the waveform the wave V component is clearly identifiable and the latency is relatively constant. (b) Four ABERs obtained during the last 10min of a 30-min hypoxic period (blood 0, saturation of 81.4%) for the same subject as in (a). Note the delay in the latency of the wave V component when compared to the control condition. Stimulus level for both sets of records is 55 dBHL.
** p < 0.01
0.0 0-10 mln
10-20
min
Fig. 3. The mean prolongation of V , compared to the control V,, in each subject is plotted for the three 10-min epochs of hypoxia. Data are pooled from the 10 experiments in this study and the error bars indicate one standard error of the mean. The latency prolongation during the third 10min epoch of hypoxia was found to be significantly different to the control (see text).
20-30min
Hypoxia
hypoxia (p < 0.05). However, the difference in the mean V , obtained during the third epoch compared to the control and the first two epochs was highly significant (p < 0.01). The hypoxia used in this study resulted in blood 0,saturations of between 75% and 85%. Surprisingly, no significant correlation was found between blood O2 saturation and V,,, in our population [Spearman correlation: p = 0.513 (16)]. However, inspection of the individual results revealed that there was considerable individual variation in both the blood 0, saturation that produced the prolongation of V,, and the extent of the prolongation at any , at a PAOzof 50 Tom one level of saturation (Table I). One subject showed no change in V (mean blood 0,saturation of 83.3%) but showed a significant change in latency a t a P,02 of 45Torr (mean blood 0,saturation of 78.6%). Furthermore, one subject showed no change in latency for a blood O2 saturation of 74.6% (PAO, 45 Torr).
DISCUSSION In this study there was a significant increase in the ABER wave V latency following 20 min of mild hypoxia. The time course of this effect is consistent with the findings of Sohmer et
ABER latency during acute hypoxia 943
Acta Otolaryngol (Stockh) 112
Table I. ~~
Subject 1
1
Downloaded by [University of Sydney Library] at 18:55 21 March 2016
9 2 3 4 5 6 8 10
~~~
Hypoxia
0,Sat
Stim Level
Control V,,,
Hypox V ,
dB Shift
50 45 45 50 50 45 45 45 45 45
84.6 84.3 84.2 83.4 83.3 80.4 81 18.6 16.6 14.6
50 55 55 45 55 55 55 55 65 65
6.68 635 6.51 6.11 6.74 6.15 6.26 6.61 6.29 6.39
6.12 6.32 6.64 6.96 6.14 1.17 6.51 6.89 6.41 6.37
-8 -3.5 -3 -6.6 0
-1
- 10 -8 -4 0
al. (6) who reported that reductions in ABER threshold measured in anaesthetised cats stabilized to a constant value after about 20min of severe hypoxia. There was however, considerable inter-subject variation in the blood 0, saturation level produced for a particular P,O, and in the extent of the related shift in latency for similar blood 0, saturations. Such variations in the respiratory response to hypoxia have been documented previously ( 17, 18). A component of the variation in the sensitivity to hypoxia might be attributable to the measurement variation that resulted from a difference in the signal to noise ratio of the ABER between subjects. However, close inspection of individual data indicates that there is no obvious relationship between the signal to noise ratio and the level of sensitivity to hypoxia indicating that these data probably reflect actual differences in sensitivity to hypoxia amongst our subjects. The prolongation of V, noted in this and other studies could be interpreted as either an increase in the central conduction time of the auditory pathway or as a decrease in the peripheral sensitivity. Any change in central conduction time could be estimated by comparing wave V latency with the latency of an earlier wave of the ABER. In other studies, the brainstem conduction time has been reported to be unaffected by hypoxia (7). However, in this study this possibility cannot be ruled out as, due to the noisy recording conditions, it was not possible to reliably measure the response latencies of the earlier ABER waves. A more likely explanation of the prolongation in latency is suggested by the studies of cochlear function under reduced oxygen conditions (e.g. 1,2,3); i.e. that hypoxia produces a reduction in peripheral sensitivity (4,5, 7). As shown in this and other studies, changing the stimulus level under normal breathing conditions produces changes in the response latency (Fig. 1). A hypoxic-induced reduction in peripheral sensitivity would result in an increase in the latency of the response to a constant level stimulus. Sohmer and his colleagues ( 5 ) have also argued that the peripheral nature of this change in sensitivity is also supported by the failure of visual and somatosensory evoked potentials to show the same changes in latency even during intense hypoxia. Interpreted in this light we have used the response latencylstimulus level relation determined for each subject, to relate the prolongation of Vht to a change in auditory sensitivity. Analysed in this way the mean change in sensitivity was found to be 5.1 & 3.4 dB (S.D) for our population of subjects (Table I). This change in sensitivity is consistent with the earlier audiometric measurements of Gellhorn & Spiesman (9). However, as these changes were found with relatively mild hypoxia they seem at variance with the later animal studies (6) and with the human audiometric study by Burkett & Perrin (10). These differences may have
Downloaded by [University of Sydney Library] at 18:55 21 March 2016
944 S. Carlile et al.
Acta Otolaryngol
resulted from two different effects. Firstly, in Burkett & Perrin’s study (lo), human subjects were pre-exposed to 100% O2 and then allowed only 5 min at the simulated test “altitude” before audiometric sensitivity was determined. They report no effect of the simulated altitude on the measured threshold. The importance of the duration of the exposure to hypoxia in demonstrating a physiological depression in auditory sensitivity [(6)and Fig. 3 of the present study] and the relatively short period of hypoxic exposure used in their study may explain this apparently negative finding. Secondly, in the previous animal studies, changes in the ABER amplitude and latency have been reported to occur only in response to severe hypoxia, where 0,saturation was reduced to or below 50% (4,5,6). However, the failure of these studies to detect any change in the ABER latency with less hypoxia may have resulted from selection of a test stimulus that optimized response amplitude conditions but corresponded to the saturating portion of the stimulus level/response latency relation. Under these conditions large changes in auditory system sensitivity would be required to produce significant changes in the ABER latency. There is no indication in any of these animal studies of the stimulus level in terms of the stimulus-level/response-latencyrelation for these animals. Sohmer et al. (6) also reported that changes in the threshold of the ABER of the cat were observed only when blood 0, saturations fell below 45%. Even very small changes in auditory system sensitivity at higher blood O2 saturation levels should have been evident using this method. However, in that study the ABER thresholds were determined using 5 dB steps which, given the magnitude of the variations in auditory sensitivity found in the present study, may have been too large to detect such changes. The reductions in auditory sensitivity suggested by the prolongation in the latency of the ABER with mild hypoxia are only relatively small when compared to what is considered a clinical impairment in hearing. However, whatever the mechanism producing the prolongation in V,,, it is clear from this study that the human auditory system is sensitive to much lower levels of hypoxia than previously thought. ACKNOWLEDGEMENTS This work was supported by the Wellcome Trust and the Medical Research Council (program grant number PG7900491). SC was an MRC Alexander Werhner Pigott Post Doctoral Fellow during these experiments. We would like to thank Dr Peter Robbins for the use of the computer driven g a s mixing system and Dr David Moore for use of recording equipment. We would also like to thank Drs Moore, Hutchings and Morey for comments on an earlier version of this manuscript. The levels of hypoxia used in this study were approved by the Ethics Committee of the University of Oxford.
REFERENCES 1. Russell IJ, Cowley EM. The influence of transient asphyxia on the receptor potentials in inner hair
cells of the guinea pig cochlea. Hear Res 1983; I I: 373-84. 2. Brown MC, Nuttall AL, Masta R1, Lawrence M. Cochlear inner hair cells: effects of transient asphyxia on intracellular potentials. Hear Res 1983; 9: 131-44. 3. Gafni M, Sohmer H. Intermediate endocochlear potential levels induced by hypoxia. Acta Otolaryngo1 (Stockh) 1976; 8 2 354-8. 4. Sohmer H, Freeman S, Gafni M, Goitein K. The depression of the auditory nerve-brainstem evoked response in hypoxaemia - mechanism and site of effect. Electroencephalogr. Clin Neurophysiol 1986; 64: 334-8. 5. Sohmer H, Freeman S, Malachi S. Multi-modal evoked potentials in hypoxaemia. Electroencaphalogr Clin Neurophysiol 1986; 64: 328-33. 6. Sohmer H, Freeman S, Schmuel M. ABR threshold is a function of blood oxygen level. Hear Res 1989; 40: 87-92.
(Stockh) 112
Downloaded by [University of Sydney Library] at 18:55 21 March 2016
Acta Otolaryngol (Stockh) I12
ABER latency during acute hypoxia 945
7. Attias J, Sohmer H, Gold S, Haran I, Shahar A. Noise and hypoxia induce temporary threshold shifts in rats studied by ABR. Hear Res 1990; 45: 247-52. 8. Deecke L, Goode RC, Whitehead G, Johnson WH, Bryce DP. Hearing under respiratory stress: latency changes of the human auditory evoked potential during hyperventilation, hypoxia, asyphyxia and hypercapnia. Aerospace Med 1973; 1106-11. 9. Gellhorn E, Spiesman IG. The influence of hyperpnea and of variations of 0, and CO, tensions in the inspired air upon hearing. Am J Physiol 1935; 112 519-28. 10. Burkett Pr, Perrin WF. Hypoxia and auditory thresholds. Aviat Space Environ Med 1976; June: 649-51. 11. Carlile S, Bascom DA, Paterson DJ. The effects of mild hypoxia on the latency of the human auditory brainstem response. J Physiol 1990; 435 62P. 12. Morey AL, Carlile S. Auditory brainstem of the ferret: maturation of the brainstem auditory evoked potential. Dev Brain Res 1990; 5 2 279-88. 13. Picton TW, Hillyard SA, Krausz HI, Galambos R. Human auditory evoked potentials. I. Evaluation of components. Electroencephalogr Clin Neurophysiol 1974; 3 6 179-90. 14. Khamnei S, Robbins PA. Hypoxic depression of ventilation in humans? Alternative models for the chemoreflexes. Respir Physiol 1990; 81: 117-34. 15. Winer BJ. Statistical principals in experimental design. New York: McGraw-Hill, 1971. 16. Sokal RR, Rohlf FJ. Biometry. San Francisco: W.H. Freeman, t969. 11. Roughton FJ. Respiratory function of the blood. In: Boothby WM, ed. Handbook of respiratory physiology. USAF School of Aviation Medicine, 1954 51-102. 18. Cunningham DJC. Studies on arterial chemoreceptors in man. J Physiol 1987; 384: 1-26. Manuscript received March 2, 1992; accepted April 7, 1992.
Address for correspondence: Simon Carlile, University Laboratory of Physiology, Parks Rd., Oxford OX1 3PT, U.K.
ISSN: 0001-6489 (Print) 1651-2251 (Online) Journal homepage: http://www.tandfonline.com/loi/ioto20
The Effect of Acute Hypoxia on the Latency of the Human Auditory Brainstem Evoked Response Simon Carlile, Daphne A. Bascom & David J. Paterson To cite this article: Simon Carlile, Daphne A. Bascom & David J. Paterson (1992) The Effect of Acute Hypoxia on the Latency of the Human Auditory Brainstem Evoked Response, Acta OtoLaryngologica, 112:6, 939-945, DOI: 10.3109/00016489209137494 To link to this article: http://dx.doi.org/10.3109/00016489209137494
Published online: 08 Jul 2009.
Submit your article to this journal
Article views: 9
View related articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ioto20 Download by: [University of Sydney Library]
Date: 21 March 2016, At: 18:55
Acta Otolaryngol (Stockh) 1992; 112: 939-945
The Effect of Acute Hypoxia on the Latency of the Human Auditory Brainstem Evoked Response SIMON CARLILE, DAPHNE A. BASCOM and DAVID J. PATERSON From the University Laboratory of Physiology, Oxford, England
Downloaded by [University of Sydney Library] at 18:55 21 March 2016
Carlile S, Bascom DA, Paterson DJ. The effect of acute hypoxia on the latency of the human auditory evoked response. Acta Otolaryngol (Stockh) 1992; 112 939-945. Recent studies have shown a decrease in the amplitude and an increase in the threshold of the cat’s auditory brainstem evoked response (ABER) during severe hypoxia (P.0, of 20 to 30 Torr). In this study we have examined the effects of euoxia (end tidal PO, 100 Torr) and mild hypoxia (end tidal PO, of 45 to 50Torr) on the latency of the ABER in 6 human subjects. Hypoxia resulted in a blood 0, saturation of between 75 to 85% and caused a significant prolongation of the latency of wave V of the ABER by 0.185 & 0.045 ms (Mean f S.D; p c 0.01). The prolongation of the ABER during severe hypoxia has previously been attributed to a change in peripheral sensitivity. Using the stimulus level/response latency relationship obtained for each subject under normal breathing conditions, the change in latency produced by mild hypoxia can be interpreted as a mean shift in auditory sensitivity of 5.1 f 3.4 dB. These results suggest that the auditory system is sensitive to much smaller changes in blood 0, saturation than previously thought. Keywords: mild hypoxia, auditory sensitivity.
INTRODUCTION Recent studies of cochlear function have shown that hypoxia reduces a number of cochlear electro-chemical potentials [e.g. (1,2, 3)] thought to be important in maintaining normal cochlear sensitivity. Other studies of auditory system function during hypoxia have used the auditory brainstem evoked response (ABER) and shown that severe hypoxia (blood 0, saturations of 25-50%) produces a reduction in the amplitude (4, 5) and threshold sensitivity (6) of the response in the anaesthetized cat. A recent study in the rat has shown decreases in the threshold sensitivity of the ABER and increases in the latencies of waves I and V of the ABER following a 2-h exposure to 5% 0, in N, (7). In humans, Deecke et al. (8) have reported significant prolongation of the latencies of the middle and long latency components of the auditory evoked potential during hypoxia (alveolar PO, (PA02)of 55Torr). Furthermore, an early audiometric study of threshold sensitivity (9) showed that exposure to hypoxia (9% to 12% 0,in N,) for periods of up to 30 min produced a significant depression in auditory sensitivity. In contrast, Burkett & Perrin (10) failed to find any change in audiometric sensitivity when subjects were exposed for a short time to simulated altitudes of up to 6,100 m. In this study we found that 20 minutes’ exposure to mild hypoxia of between 45 to 50 Torr produced a blood 0, saturation of 75 to 85% resulting in a significant prolongation of the latency of wave V of the ABER. If this change in latency is interpreted as a reduction in auditory sensitivity at the periphery, this prolongation corresponds to an average reduction in auditory sensitivity of 5.1 dB. A preliminary report of these results have been published previously (11).
940 S. Carlile et al.
Acta Otolaryngol (Stockh) 112
Downloaded by [University of Sydney Library] at 18:55 21 March 2016
MATERIAL AND METHODS Under normal (euoxic) conditions there is a systematic reduction in the latency of the ABER with increasing stimulus level (12). The relationship is essentially sigmoidal with the latency changing most rapidly over the middle portion of the dynamic range. The selection of a test stimulus level corresponding to the steepest portion of this relation would provide the most sensitive stimulus for detecting any changes in ABER latency, particularly if these changes result from change in the sensitivity of the auditory system. In this study our strategy was to first determine the stimulus-leve1,kesponse-latencyrelation for each subject, and then to select a test stimulus level based on the criteria outlined above. Using a constant intensity stimulus, ABERs were then determined during euoxia and mild hypoxia. Ten experiments were performed on 6 healthy adult volunteers (2 females and 4 males) ranging in age from 23 to 50 years (mean 36, s.d. 10 years). Subjects reclined comfortably on a couch and were required to remain still with their eyes closed to reduce muscle artifacts. The skin was cleaned to reduce electrical resistance (Omniprep) and the ABER to binaural click stimuli was recorded using three EEG cup electrodes filled with conductive jel (Camjel); one positioned at the vertex and one over each mastoid (left ear active, vertex reference and right ear ground). The signal was amplified (preamp: Iselworth Type A102, high-gain amp: Neurolog NL104), band pass filtered (20 Hz to 5 kHz: Neurolog NL12.5) and delivered to a computer interface (Cambridge Electronic Design: CED1401) for analog to digital conversion. A conventional signal averaging software package (SigAvg: Cambridge Electronic Design) running on a 80286 PC was used to average the responses to between 500 and 2,000 repetitions of the click stimulus presented at 15 pulses per second. The number of repetitions were determined by the amount of averaging required to obtain a clear wave V component in the response (see below). Auditory stimuli were produced using a 50ps transient positive voltage pulse generated by a function generator (Wavetek Model 182A) driving calibrated circumaural headphones (Bayer; DT202). Because of the requirements of the breathing control apparatus, these recordings could not be achieved in an electrically isolated environment. This unfortunately resulted in a relatively noisy EEG signal. In addition, as hypoxia results in increased ventilation, the increased inspiratory effort produced an increase in the respiratory muscle artifact which also contributed to the relatively noisy recordings. Under these recording conditions the latency of the ABER could only be assessed reliably using the latency of wave V (V,J which proved to be the most robust feature of the human ABER [See (13) for wave nomenclature]. The gas delivery system used for producing hypoxia has been described in detail elsewhere (14). In brief, hypoxia was produced by varying the concentration of inspired N2 in air. Each subject had his or her nose occluded and breathed the gas mixture through a mouthpiece. By measuring end tidal O2 partial pressure this mixture was adjusted to result in a PA02of either 50 Torr or 45 Torr. Blood O2 saturations were estimated using a pulse oximeter (PhysiolControl Lifestat 1600) with a finger probe. A P A 0 2 of 50 Torr produced average blood O2 saturation levels of 83.2 f 1.4% (SEM), while reducing PA02to 45 Torr resulted in an average blood 0,saturation of 80.9 5 1.15%. PAO, and PAC02 were held constant throughout the experiment by a computer driven gas mixing system (14). After the stimulus-level/response-latencyrelation was determined, the subject was placed on the breathing apparatus and allowed to breathe normal room air (euoxic condition). Once breathing and blood 0,saturation stabilized, a number of control ABER recordings were taken. This was followed by a period of hypoxia which lasted 30min. During the hypoxic period the ABER was obtained as many times as possible (12 to 22 measures); the precise
ABER latency during acute hypoxia 941
Acta Otolaryngol (Stockh) 112
number of samples were determined by the number of stimulus repetitions required to obtain each sample. For the data analysis, latencies obtained during the period of hypoxia were pooled over three 10-min epochs and compared with the pooled control records.
Downloaded by [University of Sydney Library] at 18:55 21 March 2016
RESULTS The stimulus-level/response-latencyrelation was determined from ABER threshold to between 65 and 75 dBHL in 5 dB steps. Fig. 1 shows the changes in V,, associated with the variation in stimulus level obtained for 3 subjects. V,,, was found to be a systematic function of the stimulus level, showing a decrease of around 1.5 ms over a dynamic range of between 30 dB and 40 dB in stimulus level. Vlat was only measured where wave V could be clearly identified as a prominent peak preceding the large negative component around 9 to 10 ms latency commonly referred to as SN,,[see Fig. 2a and (1311. The change in response latency was most rapid for the middle 20 dB to 30 dB of the dynamic range of the response. The test stimulus for each subject was then selected so as to be centred in this range. In Fig. 2, four ABERs obtained from one subject during a euoxic control period (Fig. 2u) have been compared with 4 ABERs obtained from the same subject during the last 10 min of hypoxia (mean blood 0, saturation: 81.4%; Fig. 2b). Although some aspects of these ABERs are quite variable there is virtually no variation in the V,, within each condition. Fig. 2 shows a significant prolongation in V,,, obtained during hypoxia when compared with the latency of the response obtained during the control period. The latencies obtained during the 30 min of hypoxia were pooled into three 10-min epochs to examine the time course of the hypoxic effect. As the actual latencies recorded were dependent on the test stimulus level, which was varied from subject to subject, the hypoxic latencies were normalized with respect to the control latencies for each subject before pooling for analysis. The hypoxic stimulus produced a significant prolongation of V,,, by an average of 0.184 ms, s.d. 0.045 ms [Fig. 3; p < 0.01: One way ANOVA with repeated measures (1511. There was no significant difference between the control mean V,,, and the mean V , , obtained for the first two 10-min epochs of
7.5-
7.0-
6.5-
6.0 -
5.58
.
I
.
I
. '
.
'
'
Fig. I . Wave V latency (Vlat) is plotted against stimulus level for 3 subjects. Note that the relationship is essentially sigmoidal, changing most rapidly over the 40 to 60 dB range in stimulus level.
' . 0
Downloaded by [University of Sydney Library] at 18:55 21 March 2016
942 S. Carlile et al.
Acta Otolarmeol (Stockhl 112
Hypoxia ms
Fig. 2. (a) Four ABERs obtained during the euoxic control period have been overlaid and the mean latency of V , for all the control records for this subject is indicated by the dotted line at wave V. Note that despite the variation in other components of the waveform the wave V component is clearly identifiable and the latency is relatively constant. (b) Four ABERs obtained during the last 10min of a 30-min hypoxic period (blood 0, saturation of 81.4%) for the same subject as in (a). Note the delay in the latency of the wave V component when compared to the control condition. Stimulus level for both sets of records is 55 dBHL.
** p < 0.01
0.0 0-10 mln
10-20
min
Fig. 3. The mean prolongation of V , compared to the control V,, in each subject is plotted for the three 10-min epochs of hypoxia. Data are pooled from the 10 experiments in this study and the error bars indicate one standard error of the mean. The latency prolongation during the third 10min epoch of hypoxia was found to be significantly different to the control (see text).
20-30min
Hypoxia
hypoxia (p < 0.05). However, the difference in the mean V , obtained during the third epoch compared to the control and the first two epochs was highly significant (p < 0.01). The hypoxia used in this study resulted in blood 0,saturations of between 75% and 85%. Surprisingly, no significant correlation was found between blood O2 saturation and V,,, in our population [Spearman correlation: p = 0.513 (16)]. However, inspection of the individual results revealed that there was considerable individual variation in both the blood 0, saturation that produced the prolongation of V,, and the extent of the prolongation at any , at a PAOzof 50 Tom one level of saturation (Table I). One subject showed no change in V (mean blood 0,saturation of 83.3%) but showed a significant change in latency a t a P,02 of 45Torr (mean blood 0,saturation of 78.6%). Furthermore, one subject showed no change in latency for a blood O2 saturation of 74.6% (PAO, 45 Torr).
DISCUSSION In this study there was a significant increase in the ABER wave V latency following 20 min of mild hypoxia. The time course of this effect is consistent with the findings of Sohmer et
ABER latency during acute hypoxia 943
Acta Otolaryngol (Stockh) 112
Table I. ~~
Subject 1
1
Downloaded by [University of Sydney Library] at 18:55 21 March 2016
9 2 3 4 5 6 8 10
~~~
Hypoxia
0,Sat
Stim Level
Control V,,,
Hypox V ,
dB Shift
50 45 45 50 50 45 45 45 45 45
84.6 84.3 84.2 83.4 83.3 80.4 81 18.6 16.6 14.6
50 55 55 45 55 55 55 55 65 65
6.68 635 6.51 6.11 6.74 6.15 6.26 6.61 6.29 6.39
6.12 6.32 6.64 6.96 6.14 1.17 6.51 6.89 6.41 6.37
-8 -3.5 -3 -6.6 0
-1
- 10 -8 -4 0
al. (6) who reported that reductions in ABER threshold measured in anaesthetised cats stabilized to a constant value after about 20min of severe hypoxia. There was however, considerable inter-subject variation in the blood 0, saturation level produced for a particular P,O, and in the extent of the related shift in latency for similar blood 0, saturations. Such variations in the respiratory response to hypoxia have been documented previously ( 17, 18). A component of the variation in the sensitivity to hypoxia might be attributable to the measurement variation that resulted from a difference in the signal to noise ratio of the ABER between subjects. However, close inspection of individual data indicates that there is no obvious relationship between the signal to noise ratio and the level of sensitivity to hypoxia indicating that these data probably reflect actual differences in sensitivity to hypoxia amongst our subjects. The prolongation of V, noted in this and other studies could be interpreted as either an increase in the central conduction time of the auditory pathway or as a decrease in the peripheral sensitivity. Any change in central conduction time could be estimated by comparing wave V latency with the latency of an earlier wave of the ABER. In other studies, the brainstem conduction time has been reported to be unaffected by hypoxia (7). However, in this study this possibility cannot be ruled out as, due to the noisy recording conditions, it was not possible to reliably measure the response latencies of the earlier ABER waves. A more likely explanation of the prolongation in latency is suggested by the studies of cochlear function under reduced oxygen conditions (e.g. 1,2,3); i.e. that hypoxia produces a reduction in peripheral sensitivity (4,5, 7). As shown in this and other studies, changing the stimulus level under normal breathing conditions produces changes in the response latency (Fig. 1). A hypoxic-induced reduction in peripheral sensitivity would result in an increase in the latency of the response to a constant level stimulus. Sohmer and his colleagues ( 5 ) have also argued that the peripheral nature of this change in sensitivity is also supported by the failure of visual and somatosensory evoked potentials to show the same changes in latency even during intense hypoxia. Interpreted in this light we have used the response latencylstimulus level relation determined for each subject, to relate the prolongation of Vht to a change in auditory sensitivity. Analysed in this way the mean change in sensitivity was found to be 5.1 & 3.4 dB (S.D) for our population of subjects (Table I). This change in sensitivity is consistent with the earlier audiometric measurements of Gellhorn & Spiesman (9). However, as these changes were found with relatively mild hypoxia they seem at variance with the later animal studies (6) and with the human audiometric study by Burkett & Perrin (10). These differences may have
Downloaded by [University of Sydney Library] at 18:55 21 March 2016
944 S. Carlile et al.
Acta Otolaryngol
resulted from two different effects. Firstly, in Burkett & Perrin’s study (lo), human subjects were pre-exposed to 100% O2 and then allowed only 5 min at the simulated test “altitude” before audiometric sensitivity was determined. They report no effect of the simulated altitude on the measured threshold. The importance of the duration of the exposure to hypoxia in demonstrating a physiological depression in auditory sensitivity [(6)and Fig. 3 of the present study] and the relatively short period of hypoxic exposure used in their study may explain this apparently negative finding. Secondly, in the previous animal studies, changes in the ABER amplitude and latency have been reported to occur only in response to severe hypoxia, where 0,saturation was reduced to or below 50% (4,5,6). However, the failure of these studies to detect any change in the ABER latency with less hypoxia may have resulted from selection of a test stimulus that optimized response amplitude conditions but corresponded to the saturating portion of the stimulus level/response latency relation. Under these conditions large changes in auditory system sensitivity would be required to produce significant changes in the ABER latency. There is no indication in any of these animal studies of the stimulus level in terms of the stimulus-level/response-latencyrelation for these animals. Sohmer et al. (6) also reported that changes in the threshold of the ABER of the cat were observed only when blood 0, saturations fell below 45%. Even very small changes in auditory system sensitivity at higher blood O2 saturation levels should have been evident using this method. However, in that study the ABER thresholds were determined using 5 dB steps which, given the magnitude of the variations in auditory sensitivity found in the present study, may have been too large to detect such changes. The reductions in auditory sensitivity suggested by the prolongation in the latency of the ABER with mild hypoxia are only relatively small when compared to what is considered a clinical impairment in hearing. However, whatever the mechanism producing the prolongation in V,,, it is clear from this study that the human auditory system is sensitive to much lower levels of hypoxia than previously thought. ACKNOWLEDGEMENTS This work was supported by the Wellcome Trust and the Medical Research Council (program grant number PG7900491). SC was an MRC Alexander Werhner Pigott Post Doctoral Fellow during these experiments. We would like to thank Dr Peter Robbins for the use of the computer driven g a s mixing system and Dr David Moore for use of recording equipment. We would also like to thank Drs Moore, Hutchings and Morey for comments on an earlier version of this manuscript. The levels of hypoxia used in this study were approved by the Ethics Committee of the University of Oxford.
REFERENCES 1. Russell IJ, Cowley EM. The influence of transient asphyxia on the receptor potentials in inner hair
cells of the guinea pig cochlea. Hear Res 1983; I I: 373-84. 2. Brown MC, Nuttall AL, Masta R1, Lawrence M. Cochlear inner hair cells: effects of transient asphyxia on intracellular potentials. Hear Res 1983; 9: 131-44. 3. Gafni M, Sohmer H. Intermediate endocochlear potential levels induced by hypoxia. Acta Otolaryngo1 (Stockh) 1976; 8 2 354-8. 4. Sohmer H, Freeman S, Gafni M, Goitein K. The depression of the auditory nerve-brainstem evoked response in hypoxaemia - mechanism and site of effect. Electroencephalogr. Clin Neurophysiol 1986; 64: 334-8. 5. Sohmer H, Freeman S, Malachi S. Multi-modal evoked potentials in hypoxaemia. Electroencaphalogr Clin Neurophysiol 1986; 64: 328-33. 6. Sohmer H, Freeman S, Schmuel M. ABR threshold is a function of blood oxygen level. Hear Res 1989; 40: 87-92.
(Stockh) 112
Downloaded by [University of Sydney Library] at 18:55 21 March 2016
Acta Otolaryngol (Stockh) I12
ABER latency during acute hypoxia 945
7. Attias J, Sohmer H, Gold S, Haran I, Shahar A. Noise and hypoxia induce temporary threshold shifts in rats studied by ABR. Hear Res 1990; 45: 247-52. 8. Deecke L, Goode RC, Whitehead G, Johnson WH, Bryce DP. Hearing under respiratory stress: latency changes of the human auditory evoked potential during hyperventilation, hypoxia, asyphyxia and hypercapnia. Aerospace Med 1973; 1106-11. 9. Gellhorn E, Spiesman IG. The influence of hyperpnea and of variations of 0, and CO, tensions in the inspired air upon hearing. Am J Physiol 1935; 112 519-28. 10. Burkett Pr, Perrin WF. Hypoxia and auditory thresholds. Aviat Space Environ Med 1976; June: 649-51. 11. Carlile S, Bascom DA, Paterson DJ. The effects of mild hypoxia on the latency of the human auditory brainstem response. J Physiol 1990; 435 62P. 12. Morey AL, Carlile S. Auditory brainstem of the ferret: maturation of the brainstem auditory evoked potential. Dev Brain Res 1990; 5 2 279-88. 13. Picton TW, Hillyard SA, Krausz HI, Galambos R. Human auditory evoked potentials. I. Evaluation of components. Electroencephalogr Clin Neurophysiol 1974; 3 6 179-90. 14. Khamnei S, Robbins PA. Hypoxic depression of ventilation in humans? Alternative models for the chemoreflexes. Respir Physiol 1990; 81: 117-34. 15. Winer BJ. Statistical principals in experimental design. New York: McGraw-Hill, 1971. 16. Sokal RR, Rohlf FJ. Biometry. San Francisco: W.H. Freeman, t969. 11. Roughton FJ. Respiratory function of the blood. In: Boothby WM, ed. Handbook of respiratory physiology. USAF School of Aviation Medicine, 1954 51-102. 18. Cunningham DJC. Studies on arterial chemoreceptors in man. J Physiol 1987; 384: 1-26. Manuscript received March 2, 1992; accepted April 7, 1992.
Address for correspondence: Simon Carlile, University Laboratory of Physiology, Parks Rd., Oxford OX1 3PT, U.K.
