Most NIHL is caused by damage to the inner ear (cochlea), specifically to the sensory cells (hair cells), which convert sound-induced mechanical vibrations in the cochlear fluids into electrical signals that are carried to the brain by fibers of the cochlear nerve (Fig. 1). Temporary NIHL is caused by sub-microscopic changes to the hair cells machinery that reversibly compromise its function. Permanent NIHL is caused by hair cell death, which can occur within hours of exposure: once hair cells die, they are never replaced (unless youre a bird, but thats another story!). Degeneration of the cochlear nerve is much slower, with neural loss continuing for months to years post exposure. This time delay has suggested that hair cell damage is the primary effect of noise, and that nerve degeneration occurs only when the hair cells are destroyed first.

Recent work in our laboratory has shown that significant degeneration of the cochlear nerve occurs after noise exposure, even when there is no hair cell loss, and even when thresholds have returned to normal. We study acoustic injury in mice and guinea pigs. Since the inner ears of all mammals are very similar, findings in these rodents almost certainly apply to humans. We expose animals to continuous noise for 2 hours at levels from 100 - 105 dB SPL, well below the threshold of pain, and roughly equivalent to the noise produced by a belt sander or circular saw.

Before and after exposure, we measure thresholds by two non-invasive techniques (also used in the clinics to measure hearing in infants): 1) auditory brainstem responses (ABRs), which are electrical potentials measured from scalp electrodes that represent the summed activity of cochlear nerve fibers evoked by short tone bursts; and 2) distortion product otoacoustic emissions (DPOAEs), which are sounds created and amplified by normal hair cells in response to a two-tone input sound that are propagated back out to the ear canal where they can be measured with a sensitive microphone. Immediately after the noise exposure, our animals show a moderate NIHL of 30 40 dB, by both ABRs and DPOAEs. Two weeks later, thresholds have returned to normal, however, ABR amplitudes recover only to < 50% of pre-exposure values, suggesting degeneration of >50% of the cochlear nerve.

At several post-exposure times, we look at the microscopic structure of the inner ear. Using antibodies to stain specific cellular components, we see that > 50% of the synaptic connections between hair cells and neurons disappear within 1 day after exposure. Although functionally disconnected from the hair cell, these nerve fibers survive for many months. However, by two years post-exposure, the nerve loss (~50%) matches the degree of ABR amplitude reduction, even though all the hair cells remain intact. The slow neurodegeneration is caused by disruption of the normal trophic support these neurons receive from their cellular neighbors in the hair cell area.

Clearly, our work challenges the long-held view that reversibility of noise-induced threshold shifts indicates complete recovery of cochlear structures. It also suggests that current damage risk criteria for human noise exposure may be inadequate, because they are based on the assumption that reversible threshold shifts are benign.

Is it paradoxical that thresholds return to normal despite loss of > 50% of the nerve fibers connecting hair cells to the brain? No - research from the 1950s in behaviorally trained animals showed that partial lesions of the cochlear nerve do not affect thresholds for detection of tones in quiet, as long as the hair cells are functioning normally. What, then, are the functional consequences of this primary neural degeneration? We believe that the neuronal loss will affect hearing in a noisy environment and may explain why difficulties with hearing in noise increase so dramatically in the aging ear.