Effects of conductive hearing loss on temporal aspects of sound transmission through the ear

Abstract

Effects of conductive hearing loss on level and spectrum are well known. However, little is known about possible additional effects on temporal aspects of sound transmission. This study investigated effects of earplugs and middle ear effusions on amplitude and timing of cochlear microphonic (CM) responses in gerbils. Bilateral CM responses to pure tones (1–16 kHz) were monitored before and after (i) unilateral earplug insertion or (ii) injection of silicone oil, of various viscosities, into one middle ear. Earplugs produced flat hearing losses (mean 13 dB) and delayed CMs more at lower (mean 80 μs, 1–6 kHz) than at higher (20 μs, 8–16 kHz) frequencies. Effusions also produced flat hearing loss. On average, high viscosity effusions produced larger hearing losses (36 dB) than medium (25 dB) or low (20 dB) viscosity effusions. Low and medium viscosity effusions delayed responses to lower (mean 82 and 65 μs respectively, 1–6 kHz) more than to higher (mean 20 and 10 μs respectively, 8–16 kHz) frequencies. High viscosity effusions produced smaller delays across all frequencies (mean 31 μs, 1–16 kHz). In normal animals, CM responses were not delayed over a wide range of stimulus levels. Therefore, in addition to attenuation, conductive loss distorts acoustic temporal cues important for hearing.

Introduction

Conductive hearing loss is caused by an obstruction to the sound transmission mechanisms of the external or middle ear. Common causes of conductive hearing loss in adults include cerumen (wax) in the outer ear and otosclerosis (stapes fixation). The most common cause of hearing loss in children is otitis media with effusion (OME; Hogan et al., 1997), which is characterized by inflammation and fluid in the middle ear (Bluestone and Klein, 1995). The dynamic viscosity of an effusion varies considerably, from serous to ‘glue-like’ in nature (0.16–460 Poise; Takeuchi et al., 1989), leading, in Britain, to the colloquial term for OME of ‘glue ear’. The hearing loss associated with an effusion is also variable. OME commonly causes an equal (flat), mild to moderate hearing loss across all frequencies (mean 26 dB, range 0–50 dB, S.D. 9.9; Bluestone et al., 1973). However, other spectral patterns of hearing loss can occur (Gravel and Ruben, 1996). Therefore, some children are likely to receive a variable auditory input for the first few years of life.

While it has long been known that a conductive loss causes a reduction in the amplitude of sounds transmitted to the inner ear (Wever and Lawrence, 1954), the effects of a conductive loss on temporal aspects of sound transmission have received little attention. This has become an important issue because mounting evidence suggests that a conductive hearing loss early in life can lead to changes in the anatomy (Webster, 1977, Webster, 1979, Webster, 1983, Conlee and Parks, 1981, Evans et al., 1983, Smith et al., 1983, Moore et al., 1989) and physiology (Silverman and Clopton, 1977, Clopton and Silverman, 1978, Moore and Irvine, 1981, Knudsen, 1983, Knudsen, 1985, King et al., 1988, Tucci et al., 1999) of the auditory brainstem. Animal and human studies have also found auditory perceptual changes that can outlive the peripheral impairment, including adaptive shifts in sound localization (Knudsen et al., 1984, Morrongiello, 1989, Slattery and Middlebrooks, 1994, Wilmington et al., 1994, Parsons et al., 1999) and deficits in other binaural processing tasks (Florentine, 1976, Hall et al., 1990, Hall et al., 1995, Moore et al., 1991, Moore et al., 1999, Pillsbury et al., 1991, McPartland et al., 1997, King et al., 2000). In order to understand the possible mechanisms underlying these changes, it is desirable to characterize the effects of conductive hearing loss on sound transmission.

The middle and external ear couple the transfer of sound energy in the air to the fluid-filled cochlea (Wever and Lawrence, 1954, Dallos, 1973). Hall and Derlacki (1986) speculated that middle ear disease might cause an interaural asymmetry in the mechanics of the middle ear. This could introduce, among other effects, interaural phase shifts to stimuli reaching the cochlea. Their hypothesis is supported by indirect evidence from studies of binaural unmasking in children with a history of OME (Hall et al., 1998) and in adults with otosclerosis (Hall et al., 1995). In these studies, thresholds were measured for 500 Hz tones, presented binaurally with masking noise (450–550 Hz). The interaural phase difference of the noise maskers (θ) ranged from −150° to +180°. When the noises had no interaural phase delay (0°), the tone in one ear was presented in counter phase (180°) with respect to the tone in the other ear. For conditions when the noise was phase-shifted in one ear, the tone in that ear was inverted with respect to the phase-shifted noise (θ+180°). For example, when the masker led in the right ear by 30°, the tone led in the right ear by 210°. Tone thresholds for normal hearing listeners were lowest when the masking noises had no interaural phase delay (0°). As the interaural phase delay of the masker increased, tone thresholds increased. In contrast, most subjects with conductive hearing losses had minimum tone thresholds when the maskers were presented with an interaural phase delay other than zero (Hall et al., 1995, Hall et al., 1998). These results suggested that conductive hearing loss may introduce interaural phase shifts to the auditory stimulus.

Other studies have suggested that an earplug inserted in the external ear may also produce a change in the timing of a signal passing around and through it (Knudsen et al., 1984, Moore et al., 1999). In the barn owl, Knudsen et al. (1984) measured the effect of a foam plug, inserted in the external ear, on the timing of sound reaching the inner ear. They measured bilateral CM responses to pure tone stimuli, before and after unilateral plug insertion. Before plug insertion, they showed that the time properties of the CM response were effectively independent of stimulus level. For each tone frequency tested, the timing of the CM was constant (±5 μs) with changes in stimulus level, over a 60 dB range. After plug insertion, the CM in the plugged ear (relative to the unplugged ear) was delayed by about 50 μs in response to 3–6 kHz tones. The CM was delayed less in response to higher frequency tones (mean ∼20 μs, 8–10 kHz). These data suggest that earplugs modify the mechanics of the external ear, and cause frequency-dependent time delays in transmission of sound to the inner ear. Hogan et al. (1995) showed that foam earplugs produced phase shifts of between 97 and 149° (about 500 μs at 500 Hz), from recordings in a model ear canal.

In this study, using similar methods to Knudsen et al. (1984), we measured the effects of (i) earplugs and (ii) silicone oil injected into the middle ear, on the amplitude and phase of the CM in the gerbil. Silicone oil proved useful for the purposes of this experiment, as it is nonconductive, biologically inert, and comes in a variety of quantitatively documented viscosities. The specific hypotheses of the experiment were (i) that a conductive hearing loss, produced by ear-plugging or middle ear effusion, causes frequency-dependent time delays in sound transmission through the ear, and (ii) that variations in the viscosity of a middle ear effusion systematically change the time delays produced by the effusion.