Physiological Mechanisms of Onset Adaptation and Contralateral Suppression of DPOAEs in the Rat
- First Online:
- Cite this article as:
- Relkin, E.M., Sterns, A., Azeredo, W. et al. JARO (2005) 6: 119. doi:10.1007/s10162-004-5047-9
An investigation was undertaken to measure medial olivocochlear (MOC) reflexes in anesthetized rats before and after sectioning of the middle-ear muscles. Distortion product otoacoustic emission (DPOAE) magnitude and phase temporal responses were measured ipsilaterally to study MOC-mediated “DPOAE onset adaptation” and in the presence of a contralateral noise to study MOC-mediated contralateral “suppression” (terms as used by previous researchers). Distortion product otoacoustic emission onset adaptation and contralateral suppression had predictable changes in direction of magnitude and phase that were dependent on the input–output function. After sectioning of the middle-ear muscles (MEMs), DPOAE onset adaptation and contralateral suppression were greatly reduced, and there were little, if any, changes in phase. These “residual” changes were interpreted as a result of the MOC reflex. The results suggest that what appears to be DPOAE onset adaptation and contralateral suppression can be mediated primarily by MEM reflexes. When studying MOC effects on otoacoustic emissions (OAEs) using acoustic stimulation, it is necessary to make recordings over a span of stimulus levels. In addition, looking at both magnitude and phase of the OAE may help separate what is due to the MOC reflex from MEM reflex.
KeywordsDPOAEsmedial olivocochlear reflexesmiddle-ear muscle reflexesrats
Cell bodies of the medial olivocochlear (MOC) efferent system are located near the medial superior olive and project in two pathways. A proportion of the fibers project to hair cells on the ipsilateral side (which shall be referred to as the “uncrossed” fibers), and a proportion project to hair cells on the contralateral side (referred to as the “crossed” fibers). The influence of the MOC system on auditory processing has been researched in many ways, one of them by observing changes in otoacoustic emissions (OAEs). The first studies of reduction in otoacoustic emissions (OAEs) during stimulation of the contralateral ear (investigating the uncrossed MOC) were published in the early 1990s (e.g., Collet et al. 1990; Veuillet et al. 1991), and the term “contralateral suppression” has been used to identify this phenomenon.
One technique that has been used to study the crossed MOC track is to measure the temporal waveform of the DPOAE. Liberman et al. (1996) first described changes in the amplitude of DPOAEs in the cat at stimulus onset that were largely monotonic with time, had a time constant of approximately 100 ms, and changed by as much as 9 dB from the beginning to the end of the onset response. The authors referred to this phenomenon as “fast onset adaptation,” and it was subsequently reported in mice (Sun and Kim 1999; McGee et al. 2000; Jacobson et al. 2003), gerbils (McGee et al. 2000), guinea pigs (Maison and Liberman 2000; Kujawa and Liberman 2001; Lukashkin and Russell 2002), and humans (Kim et al. 2001). Whereas adaptation is most usually thought of as decreasing as a function of time, onset adaptation of the DPOAE can also be observed as increasing with time (Maison and Liberman 2000; Kujawa and Liberman 2001; Lukashkin and Russell 2002).
Liberman et al. (1996) showed that the direction of DPOAE onset adaptation for a given set of primary stimuli was the same as the direction of the DPOAE change when a noise was presented to the contralateral ear (whether the noise produced an increase or a decrease in the DPOAE) suggesting a common underlying mechanism. Controlled studies in cats and guinea pigs suggest that contralateral suppression and DPOAE onset adaptation in these species are a result of the MOC reflex and not the MEMR (Puel and Rebillard 1990; Kujawa and Liberman 2001; Liberman et al. 1996). In contrast, contralateral suppression and onset adaptation in rabbits are mainly attributed to MEMR effects. Whitehead et al. (1991) indirectly tested DPOAE reduction because of contralateral acoustic stimulation in awake rabbits. In addition, they measured MEMR thresholds using a clinical device. They found that the range of frequencies and levels producing MEMR activation were the same as those for which changes in DPOAEs were seen and concluded that MEMR activation was the cause for the reduction in DPOAEs. Luebke et al. (2002), using changes in the sound pressure level of the primary stimuli to deduce that the MEMR was activated, found that onset adaptation in awake rabbits was primarily mediated by the MEMs.
Because it is uncertain in some species or experimental protocols whether observed effects are a result of the MOC reflex, MEMR, or a possible combination of both, it seems appropriate to define the observed effects without reference to their underlying physiological mechanism. In this paper, the abbreviation “ΔLI” stands for changes in level of the ipsilaterally recorded DPOAE over the first few hundred milliseconds of the response, the same time frame over which DPOAE onset adaptation is observed. The abbreviation “ΔPI” refers to phase changes during the same time frame. The abbreviation “ΔLC” stands for changes in DPOAEs during acoustic stimulation of the contralateral ear, and “ΔPC” stands for changes in phase during the same time frame. The previously used terms “DPOAE onset adaptation” and “contralateral suppression” will be reserved for those situations where there is reasonable evidence that the phenomena are a result of the MOC reflex. The goal of the present study was to compare ΔLI and ΔLC in rats, before and after cutting of middle-ear muscles, to characterize the underlying physiological mechanisms responsible for the changes.
Subjects were nine Sprague–Dawley male rats obtained from a licensed breeder. Rats were chosen as the experimental animal because a stereotaxic atlas of the brain is available for this species, and one component of our research is to compare changes of OAEs with electric and acoustic stimulation of the MOC reflex. The rats were retired breeders weighing 450–650 g. The rats were anesthetized with a mixture of acepromazine (0.48 mg/ml) and ketamine (95 mg/ml) administered i.m. at a starting dose of 2 ml/kG. Atropine (0.4 mg/kG) was also administered at the start of the experiment to control for respiratory secretions. Supplemental doses of anesthesia of one fifth of the original bolus were administered approximately once per hour or when the rat responded to a toe pinch. Throughout surgery and data collection, rectal body temperature was maintained at 38°C by a feedback-controlled heating pad.
The animal facilities have full accreditation from Association for Assessment and Accreditation of Laboratory Animal Care. All experimental procedures were approved by the local Institutional Animal Care and Use Committee and met the National Institutes of Health guidelines for animal care.
Primaries (f1 and f2) were sinusoidal waveforms generated by two, identical, digital oscillators. One oscillator generated the lower-frequency primary (f1), and a second oscillator generated the higher-frequency primary (f2). The ratio of the primaries (f2/f1) was 1.25. A third, identical digital oscillator was used to tune the lock-in amplifier to the DPOAE frequency. All three oscillators were driven by a common signal derived from an 8-MHz crystal-controlled clock, and the oscillators were reset simultaneously to zero phase at the beginning of each stimulus presentation, thus insuring near-perfect synchrony with zero phase drift. The frequencies of the oscillators could be set with millihertz resolution.
The primary stimuli were 500 ms in duration, and the onsets and offsets were shaped by cosine-squared functions with a rise/fall of 5 ms. The rise/fall was chosen as the lowest value that allowed adequate rejection of any “artifacts” associated with the onset and offset of gated primary stimuli for the range of DPOAE frequencies and the f1/f2 ratio used in this study. The primary stimuli were then attenuated, amplified, and delivered to the subjects through ER-2 earphones.
To study ΔLC, a broadband noise was presented to the opposite ear also using an Etymotic ER-2 speaker. The noise level was calibrated in a 1-cm3 coupler using a Bruel and Kjaer, 0.5-in. microphone but was not otherwise calibrated for each individual animal. The noise was 200 ms in duration and began 50 ms after the onset of the f2 primary stimulus. The level of the broadband noise was 60 dB SPL, except in one animal, in which the noise was 70 dB SPL. Early in the experimentation, we tried several levels of noise and ultimately settled on 60 dB SPL because we saw adequate changes, but we were hoping to minimize MEM involvement using the lower noise level.
To measure DPOAEs, sound pressure in the ear canal was recorded with the microphone of an Etymotic ER-10B+ probe. The sound in the ear canal was amplified by 40 dB and then filtered. To measure the time course of the DPOAE, it is necessary to separate the sound pressure produced by the DPOAE from the sound pressure produced by the primary stimuli. This signal processing can be difficult because of the relative closeness of the frequency of the DPOAE to the primary frequencies and because the intensity of the DPOAE is typically 40 dB or lower than the intensities of the primary stimuli. The following example, using conditions encountered in our study, illustrates the quality of filtering that is required to isolate the DPOAE. For a DPOAE frequency of 3.6 kHz, using primary stimuli of frequency 4.8 and 6.0 kHz (frequency ratio of 1.25), if the DPOAE is 20 dB SPL when the lower frequency primary is at 80 dB SPL, attenuating that primary to a level 20 dB below that of the DPOAE requires a rejection slope of −240 dB/octave. The problem of separating the signals in the frequency domain is compounded by any frequency splatter that occurs at onset and offset when the primary stimuli are gated. The f2/f1 ratio of 1.25 and rise/fall time of the cosine-squared gating allowed for nearly complete separation of the DPOAE from the nearest primary stimulus, f1, including any frequency splatter associated with the onset of f1. Because f2 was further away from the DPOAE frequency, it was of less concern in this regard.
The technique for separating the DPOAE from the primaries consisted of a combination of analog and lock-in filtering followed by signal averaging. After amplification of the recording microphone, the output was band-pass filtered at the DPOAE frequency that included one stage of high-pass filtering and three stages of low-pass filtering. The high-pass rejection slope was 48 dB/octave, and the total low-pass rejection slope was 228 dB/octave (measured at twice the center frequency). The filtered output was sent to a lock-in amplifier. In-phase and quadrature components were averaged independently with the time constant on the lock-in amplifier set to 3 ms. Stimuli were repeated every 2 s, and the average consisted of 50 stimulus repetitions. The in-phase and quadrature outputs of the lock-in amplifier were low-pass filtered to provide anti-aliasing and then were sampled by a computer-controlled A/D converter at 10,000 samples/s. At the end of each stimulus condition, magnitude, i.e., the temporal envelope, and phase were calculated from the averaged components and stored for off-line analysis.
As indicated in the schematic in Figure 2a, f1 and f2 were switched on simultaneously during the switched f1 condition. All onset and offset phenomena could be observed, including those associated with ipsilateral activation of either the MOC or MEM reflex. Figure 2a shows the DPOAE magnitude as a function of time along with a schematic of the timing of the stimulus presentations. In panel c, the corresponding phase is shown. Using a 2-s interstimulus interval, each averaged response was largely independent of the previous response. Preliminary data showed that as much as 10 s was needed to achieve absolute independence, but 2 s was found to be an adequate compromise that provided more practical data collection times.
The continuous f1 condition (Figure 2b) was used to observe DPOAE changes to the gated, contralateral broadband noise. Because the continuous f1 primary was typically 10 dB greater than f2, time-dependent phenomena associated with ipsilateral responses to the primaries, such as activation of the MEM or MOC reflex, should be in a “steady-state” condition when the f2 stimulus is turned on to evoke the DPOAE. Thus the temporal envelope of the DPOAE was observed to be similar to the envelope of the gated f2 primary, with the possible exception of very rapid onset and offset transients (duration of approximately 10 ms) similar to those described by Talmadge et al. (1999). Those authors attributed the transients to the interactions of multiple DPOAE sources with differing magnitudes, phases, and latencies. These phenomena were not of primary interest in this study. Presentation of the contralateral noise during the “flat” portion of the temporal envelope of the DPOAE made it easier to observe and quantify the time course of the response to the noise. Panel d shows the corresponding phase plotted as a function of time.
Overlaid on the magnitude functions (Figure 2a and b) are hatched areas that indicate the temporal windows that were used to quantify respective changes in the magnitude and phase of the DPOAE. The ΔLI (panel a) was quantified as the difference between the average of the DPOAE magnitude, in dB SPL, in a 10-ms window 65–75 ms after stimulus onset and the average of the magnitude, also in dB SPL, for a window of equal duration from 540 to 550 ms after stimulus onset. Similarly, ΔLC (panel b) was quantified as the difference between the average magnitude measured in a 20-ms window near stimulus onset (80–100 ms) and the average of the magnitude measured in a window of equal duration near the end of the offset of the contralateral noise (from 280 to 300 ms poststimulus onset). These same measurement windows were used to measure phase from the phase temporal waveforms. Justification for this procedure was based on the observation that the average value over 540–550 ms of the switched f1 condition is similar to the average value recorded 80–100 ms for the continuous condition. The average of the envelope near the end of the DPOAE for the switched f1 condition is an approximation of the response that would be recorded with quasi-continuous primaries as is more typically the case in the literature.
Distortion product otoacoustic emission stimulus frequencies and levels
Primary frequencies used and the corresponding 2f1–f2 DPOAE frequency.
After completing the entire array of measurements, the rat was transported to the surgical room and the bulla was opened. Tendons of the MEMs were cut using microscissors and electrocautery. The muscle was laid over the open bulla, or in some animals, bone wax was applied to the open bulla. The rat was then repositioned in the booth. After recalibrating the sound pressure in the ear canal, the measurements described above were repeated with the exception that DPOAEs evoked by the f2 frequencies of 12.5 and 16 kHz were recorded last. We chose to record DPOAEs evoked by the three lower f2 frequencies, as DPOAEs could be recorded over a wider stimulus range than those evoked by the two highest f2 frequencies. During the second half of the recording period, fluid slowly accumulated in the middle-ear cavity in almost every animal. If DPOAE levels continually decreased, the animal was taken back to the surgery room to suction the fluid. After repositioning and recalibration in the recording booth, DPOAE levels returned to baseline levels. In five rats, I/O functions for all five DPOAE frequencies were recorded both before and after sectioning of the middle-ear muscles. In two rats, only partial data were obtained after sectioning the middle-ear muscles because fluid accumulation was too rapid. For an additional two rats, no data were recorded after sectioning the middle-ear muscles because DPOAE amplitude had decreased by more than 15 dB or were completely absent. In these rats, the stapedial artery had been nicked during middle-ear muscle section, and it was felt that possible cochlear damage occurred.
Except for the results shown in Figure 12, the level of f2 was always 10 dB below the level of f1. An experiment was undertaken to investigate whether the effects of varying the level of f2 relative to f1 in 1-dB increment in the rat varied in a similar manner to that reported in guinea pigs (Maison and Liberman 2000). Responses in two rats were studied for this purpose.
The “constant voltage” method was used to set the intensity of the primary stimuli. Calibrations of both primary channels were performed at 1.0 kHz in the ear of the subject, and intensity was assumed to remain constant with stimulus frequency. This method has been shown to be more accurate than attempting to calibrate the stimuli using the recording microphone (Siegel and Hirohata 1994).
Careful examination of responses in a coupler and computer simulations were used to insure that no responses reported here were “artifacts” resulting from the gating of the primary stimuli. For the stimulus conditions reported here, all onset and offset transients associated with the primary stimuli were at least 80 dB below the level of the f1 stimulus. Additionally, the noise floor measured before stimulus onset was typically −5 dB SPL or lower.
Temporal envelopes for phase also change with time and, in a manner, linked with magnitude changes. Near the onset of the stimulus, the phase changes rapidly by small amount, then there are small, slow changes during the later duration of the stimuli. The direction of phase change is linked with the direction of the magnitude change. In the quasi-steady-state part of the response, the phase increases slightly and slowly with time when the corresponding DPOAE magnitude decreases with time. When the DPOAE onset magnitude increased with time (L2 = 70 dB SPL), the corresponding phase decreased with time. At the highest intensity, phase remained as decreasing with time.
The relationship between amount of ΔLI/ΔLC and the I/O function is illustrated in Figure 10. Input–output functions for a DPOAE measured at f2= 6 kHz before MEM sectioning were measured near stimulus onset for the switched f1 condition (black line) and the continuous f1 condition (dashed line). The vertical bars show the amount of ΔLI (black bars) and ΔLC (striped bars) at each intensity. The I/O functions are nonmonotonic, each with a pronounced notch at a mid-to-high intensity. The data plotted here are representative of the entire data set. The direction and extent of ΔLI and ΔLC are related to the slope of the DPOAE I/O function. The direction of the ΔLI, decreasing or increasing, generally depends on whether the slope of the input/output function is positive or negative, respectively. From the maximum peak of the I/O function to the minimum of the notch, where the slope is negative, is where apparently paradoxical increasing ΔLI (and increasing responses to contralateral noise) is observed. At the minimum, notch response shape can be multiphasic because of the discontinuity in the slope.
Whereas it has been established convincingly in some species that ΔLI and ΔLC result from activation of the MOC reflex by either cutting the MEMs (Liberman et al. 1996; Puel and Rebillard 1990) or cutting the OCB (Kujawa and Liberman 2001), data from the rat suggest that ΔLI and ΔLC can be attributed largely to MEM reflexes. Across species, the most common range and/or mean of ΔLI and ΔLC tends to be between 0.5 and 4 dB. However, some very large effects, as much as 20 dB, have been reported (McGee et al. 2000; Kujawa and Liberman 2001; Maison and Liberman 2000). In the rat, ΔLI ranged from −14 dB to as high as 26 dB, similar to the large effects reported in guinea pigs, in which it has been demonstrated that the changes are due to the MOC (e.g., Kujawa and Liberman 2001). Data from one rabbit, in which it has been interpreted that ΔLI is influenced by MEMRs, demonstrated a ΔLI of approximately 8 dB. One reason why it is difficult to compare absolute values of ΔLI and ΔLC across species, and even across animals, is because the size and direction of change are heavily dependent on the I/O function. Moreover, in the notch region, ΔLI can be extremely large with differences in stimulus level (Maison and Liberman 2000; McGee et al. 2000; Kujawa and Liberman 2001). Without knowing the underlying I/O function, it is difficult to compare the magnitudes of ΔLI and ΔLC across animals and species. Furthermore, looking at DPOAE magnitude only, especially at one primary level, will not allow us to adequately quantify the effects. In sum, one cannot determine from size of ΔLI or ΔLC whether the effect is due to MOC or MEMRs. In other words, it is the case that ΔLI and ΔLC across species appear to be similar phenomena, yet these phenomena result from different neural and effector pathways across species.
Our conclusions that MEM reflexes are the main contributor to ΔLI and ΔLC in the rat are similar to those reached by researchers studying awake rabbits (Luebke et al. 2002; Whitehead et al. 1991). These researchers did not cut MEMs or cut MOC at midline, but rather used acoustic middle-ear impedance and changes in primary stimulus dB SPL to determine whether MEM reflexes were being activated at same levels used to measure MOC effects. They could not separate the MEM reflex and MOC reflex effects, nor show conclusively that an MOC effect was present. They concluded that ΔLI and ΔLC were combined MEM/MOC reflexes. Our data, along with theirs, suggest that there may be a species component to thecontribution of MOC and MEMR on the reduction of DPOAEs.
One striking aspect of rat data from the current study is that there is MEMR influence at lower levels and over a broader frequency range in that rat than is typical for other species. In this respect, our data are similar to those found in albino Wistar rats. Murata et al. (1986) found that the most sensitive mean frequency that elicited the MEMR was at 3.3 kHz (SD = 1 kHz), and that the mean threshold measured through EMG recordings in either the tensor tympani or stapedius muscle was 57 dB SPL (SD = 8 dB). They presented data from one rat using visual inspection of muscle contraction that indicated that the threshold was approximately 47 dB SPL (their Figure 3b). Their group also found that thresholds were similar between stapedius and tensor tympani reflexes and between ipsilateral and contralateral reflexes. In the Wistar rat, middle-ear transmission was affected from 1 to 20 kHz.
Examples of what we interpret as MEMR influence can be seen throughout our data and are consistent with observations reported by Murata et al. (1986). For example, the I/O functions for the lowest frequency in the left panel of Figure 7 suggest that the MEMR is activated with an f1 frequency of 4.8 kHz at 45 dB SPL (f2 level and frequency are displayed in the figure). Whereas this seems low for MEMR activation in most species, f1 is close to the most sensitive MEMR frequency, and the level is within two standard deviations of the mean threshold for MEMR activation based on data from Murata et al. (1986). For higher frequencies, the I/O for the end of the trace starts to diverge from the onset I/O between 40 and 80 dB SPL for f1 frequencies of 6.4–12.8 kHz with increasingly higher levels at increasingly higher frequencies, which would also be expected. If one follows the I/O growth pattern for the beginning andend of the trace, including the nonmonotonicities, it is as if there is a suppressive effect that affects the end of the trace. If one were to shift the solid line to the right to meet the dashed line, the two I/O functions would be similar. One would predict that if the above thinking was true, there should be little difference between the beginning and end of the traces after MEM section, which is the case, as shown in the second column of Figure 7.
With MEMR activation, one would expect to see DPOAE magnitude and phase changes. Observations and measurements in cats indicate that the stapedius-muscle contraction increases the stapes impedance through strain on the annular ligament (Pang and Peak 1986). Other researchers have used increased stapedial impedance in the middle-ear model by Zwislocki (1962) to model how the middle ear changes with stapedius muscle contraction. The middle ear becomes stiffer, which increases middle-ear impedance below the resonant frequency with a corresponding phase lead (Pascal et al. 1998; Buki et al. 2000). Before MEM section, we see large ΔLC and ΔPC over a broad range of frequencies as would be predicted from previous work on MEMR in the rat (Murata et al. 1986). We would predict that the MEMR contraction has reduced the effective level of the primaries transmitted to the cochlea and has also reduced the DPOAE traveling through the middle ear. From Figure 11, it can be seen that the phase of the DPOAE during contralateral stimulation is relatively constant across L2 level. Below the I/O notch, we observe a phase lead between the onset of the DPOAE trace and the DPOAE phase during contralateral stimulation as would be predicted if the noise was activating the MEMR. Additional phase changes are often observed for the highest primary levels, and it is hypothesized that the primaries are evoking a MEMR, and when the contralateral noise is applied, the MEMR strength increases.
Whereas our data suggest that major contribution to ΔLC and ΔLI is due to MEMR, we have not performed the control experiment where we left the MEMs intact and cut the MOC fibers. It is extremely difficult in the rat to cut the MOC without inadvertently damaging the stapedius reflex arc as well. Because we have not performed this control, we cannot rule out definitively that the MOC and MEMR act independently from each other. Analysis of the MEMR and MOC systems has led some to conclude that both systems work together to enhance antimasking effects in the cochlea (e.g., Liberman and Guinan 1998). Although remote, it cannot be ruled out that there is a direct signaling between MOC reflexes and MEMRs, such that, by our act of severing the muscles, we have deactivated the MOC reflex. Although we are not aware of any literature describing direct connections between the two reflex arcs, there is evidence indicating that there are direct projections from the cochlear nucleus to the tensor tympani (Ito and Honjo 1988; Itoh et al. 1986). In addition, a body of research by Fritzsch (1999) suggests that auditory efferent fibers derive developmentally from facial nerve motoneurons, and the two tracts lie close to each other in the brainstem.
Mechanisms underlying the Distortion Product Otoacoustic Emission and Medial Olivocochlear Phenomena
The DPOAE I/O functions presented here are typical for rodents; that being, DPOAE level increases with low and moderate primary levels, exhibits a notch, then increases for higher stimulus levels (e.g., Norton and Rubel 1990; Mills 2002; Lukashkin and Russell 2002; Schmeidt and Adams 1981). Notches were generally observed when L2 was 60–70 dB SPL. Sometimes the DPOAE level was low for L2 = 75 dB SPL, and had we been able to present higher-level stimuli, it is likely that the notch would have been at or above 75 dB SPL. For high-frequency stimuli, we appeared to only observe the initial, monotonic section of the DPOAE function and did not observe notches. Perhaps if we had been able to present higher stimulus levels, we would have seen notches in these functions as well. Distortion product otoacoustic emission phase behavior was also typical for rodents (Mills 2002; Lukashkin et al. 2002); that being, there were small changes in DPOAE phase for stimulus levels lower than the notch and an approximately 180° phase change in or close to the notch. A DPOAE generator that can produce this characteristically linked amplitude and phase behavior has been modeled by Lukashkin et al. (2002) as a nonlinear amplifier with a saturating I/O characteristic as a result of the mechanoelectrical transduction of the hair cell bundle (Lukashkin and Russell 1999; Lukashkin et al. 2002). They concluded that theMOC effect of either positive or negative ΔLI could be explained by a single DPOAE generator (Lukashkin and Russell 1999). The current work demonstrates that this behavior is independent of MEMR, as we observed the linked amplitude and phase behavior before and after sectioning the MEMs. Similar to Lukashkin et al. (2002), who changed the input to the hair cells by using furosemide, we changed the input level to the cochlea by changing the middle ear. The result was that the notch was often found at different stimulus levels and had phase characteristics that followed the hypothesized model.
The assumed MOC-mediated DPOAE onset adaptation and contralateral suppression in the rat exhibited small changes in amplitude but minimal changes in phase. Seeing minimal ΔPC and ΔPI, which was mostly limited to the notch region of the I/O, seems to be most harmonious with the explanation that MOC action is to reduce cochlear amplifier gain. Alternatively, one could argue that the MOC altered both primaries equally, resulting in a reduced DPOAE with minimal phase change. However, given the nontypical ΔLC and ΔPC results in one animal at one frequency that exhibited no ΔLC along with a ΔPC phase lead, we cannot rule out that there are two regions generating DPOAEs in the rat, such as that modeled by Shera and Guinan (1999), and that the MOC reflex affected one region more than the other. Although we have not tested for two DPOAE-emitting regions in Sprague–Dawley rats by measuring microstructure, other rodents, such as guinea pigs (Withnell et al. 2003) and kangaroo rats (Long et al. 1999), do have some DPOAE microstructure.
Although the behavior of ΔLI and ΔLC in the rat is similar with and without intact MEMs, except for a difference in size, ΔPI and ΔPC are dramatically different under these two conditions. Based on these data, it brings up the question whether observing phase changes along with ΔLI and ΔLC will help parse out what changes are due to MOC and what may be due to MEMR. Our data lend support to work in humans where phase changes are being used to help determine middle-ear muscle contamination of OAE responses (Buki et al. 2000; Guinan et al. 2003).
This work was supported by grant no. 2P01 DC00380 from the National Institute of Deafness and other Communication Disorders.