3.1 Effect of LF Sound Exposure on DPOAEs
The effect of the 80 dBA LF exposure on quadratic (QDP) and cubic (CDP) OAEs is shown in Fig. 1a and b, respectively. In 14 out of 20 tested subjects, the LF exposure induced a subsequent increase of the QDP level lasting for about 60 to 90 s (see Fig. 1a). QDP levels increased with a median of 3.4 dB. In most cases, this QDP increase was followed by a similar QDP decrease (median: ‑2.4 dB), at about 120–150 s post-exposure. This decrease slowly recovered to pre-exposure QDP levels. The median duration of the overall oscillatory change of the QDP level was 214 s.
In many cases it was also possible to extract CDP levels from the same recording (albeit f2/f1 ratios were optimized to achieve maximum QDP levels). Typically, we observed no significant changes of CDP level after LF sound exposure (Fig. 1b).
3.2 Effect of LF Sound Exposure on SOAEs
We recorded 80 SOAEs from 27 ears of 16 young, normal-hearing subjects. The median SOAE sound levels were 0.6 dB SPL (first and third quartiles, ‑4.5 dB SPL; 4.0 dB SPL) with a signal-to-noise ratio of 16.6 dB (11.6 dB, 23.5 dB).
After LF sound stimulation, 56 of these 80 SOAEs increased in both sound level and frequency. This increase was followed by a decrease of both level and frequency relative to pre-exposure (see Fig. 2a). In 10 of the 80 pre-exposure SOAEs, we observed an inverted pattern with an initial level and frequency decrease, followed by a level and frequency increase.
SOAE level- and frequency oscillations were fitted with an (inverted-phase) underdamped sinusoidal oscillation. The period of the fitted sinusoid was 257 s (202 s, 294 s) for the level time course and 252 s (215 s, 367 s) for the frequency time course. The time constant of the damped sinusoid for level changes was 120 s (76 s, 157 s) and for frequency changes 94 s (58 s, 141 s). SOAE frequency changes amounted to 5 Cent (4 Cent, 9 Cent) with peak values of 25 Cent. Relative to the SOAE frequency in the control condition, the frequency showed initial maximum increases of 4 Cent (3 Cent, 7 Cent), followed by maximum decreases of 1 Cent (0 Cent, 2 Cent).
17 of 21 tested subjects revealed an overall of 56 new SOAEs, which had not been measurable before LF stimulation (see Fig. 2b). These new SOAEs were characterized by an initial level and frequency increase, qualitatively similar to the pre-existing SOAEs. Comparable to the enhancing half cycle of Bouncing SOAEs, their level and frequency oscillated before they disappeared into the noise floor. The duration of the level and frequency changes was 67.5 s (47.5 s, 90 s). New SOAEs started to arise within 12.5 s (5 s, 25 s) after LF sound offset and reached a level maximum at 50 s (35 s, 62.5 s) after LF offset. The maximum SOAE level was ‑0.3 dB SPL (‑4.1 dB SPL, 4.9 dB SPL) with a signal to noise ratio of 13.8 dB (11.9 dB, 17.6 dB). The difference between the new SOAE frequency maximum and minimum was 4 Cent (1 Cent, 6 Cent). The time course of level and frequency changes was almost identical and maximum level and frequency changes coincided.
3.3 SOAE Bouncing in the free Sound Field
Although the observed pattern of synchronized SOAE frequency- and amplitude changes is incompatible with the SOAE bouncing being elicited by changes in middle ear impedance, it is conceivable that bouncing may be only seen in the closed sound field where the auditory meatus is blocked by the OAE probe. Here, we recorded SOAEs in the open meatus using an ER10C probe microphone fitted to the meatus via an about 8 cm silicon tube (2.8 mm outer diameter) which did not block the meatus. The tip of the tube was positioned about 10 mm in front of the tympanum. LF exposure was provided by two powerful custom-made subwoofers. Subjects lay on a deck chair in a long basement corridor at a point where standing waves in the corridor maximised the sound level at 30 Hz. LF exposure was 118 dB SPL for 90 s. Photos of LF stimulation apparatus and probe-tube placement are shown in Fig. 3a, b.
Both ears of 45 young, normal-hearing subjects were screened for SOAEs. 33 subjects showed at least one SOAE in one ear. Overall we could record in the open meatus about 52 % of those SOAEs detectable in a sound-attenuated room and a closed-ear recording technique. The remaining 48 % were not significantly above the much higher noise floor of the free-field, open-meatus measurement.
Exemplary measurements of both permanent and transient SOAEs are shown in Fig. 3c and d in the same format as Fig. 2. Indeed many of those 48 % SOAEs that had been initially identified in the closed-meatus measurements, but could no longer be detected in the open meatus, appeared directly after the LF exposure for a short time period, before falling again below the noise floor (Fig. 3d). These data clearly show that Bouncing of SOAEs can indeed be elicited by free-field exposure to LF sound sources of natural or anthropogenic origin.
3.4 Effect of Contralateral Acoustic Stimulation (CAS)
Patuzzi (2011) suggested that large receptor-potentials elicited by low-frequency stimulation produce a net Ca2+ influx. The Bounce presumably reflects an underdamped, homeostatic readjustment of increased Ca2+ concentrations and related gain changes after low-frequency sound offset. Here, we tested this hypothesis by activating the medial olivocochlear (MOC) efferent system during presentation of the Bounce-evoking LF sound. The MOC system is known to modulate OHC Ca2+ concentrations (Sridhar et al. 1997) and receptor potentials (Fex 1967) and therefore it should modulate the characteristics of the Bounce. Here, CAS was provided simultaneously to the (ipsilateral to the observed SOAE) LF exposure. The CAS consisted of a 90 s, bandpass-filtered Gaussian noise (100 Hz—8 kHz) presented at 65 or 70 dB SPL.
CAS is well known to suppress ipsilaterally recorded SOAEs during presentation and SOAEs quickly recover after CAS offset within less than 1 s (Zhao and Dhar 2010, 2011). Due to the duration of our analysis segments (5 s), the SOAEs already fully recovered from the CAS exposure within the first analysis segment. Consequently, we found no SOAEs fulfilling our criteria for the Bounce or indeed any other significant changes in the CAS control recordings. When the CAS was presented simultaneously with the ipsilateral LF tone, however, Bouncing of permanent SOAEs after LF offset changed significantly. Exemplary time courses of a preexisting and a transient SOAE are shown in Fig. 4a and b, respectively. While in the reference recording (red) the preexisting SOAE showed a significant biphasic Bounce, presentation of a 65 dB SPL (blue) or 70 dB SPL (green) CAS together with the ipsilateral LF tone clearly affected the magnitude of the Bounce.
Overall, temporary level reductions of preexisting SOAEs were less pronounced with CAS than without (Wilcoxon signed rank test, p = 0.085 and 0.007 for CAS levels of 65 and 70 dB, respectively), while the SOAE level increases remained fairly unchanged. This resulted in a less symmetrical shape of the SOAE Bounce compared to the reference recording. Consequently, the time constants of the fitted function with CAS shortened significantly.