Hearing loss affects about 28 million Americans. The two most common causes of sensorineural hearing loss are aging and exposure to noise. Noise exposure can injure the ear and produce loss of hearing. The injury and hearing loss can be temporary—that is, fully recoverable after the noise exposure is terminated—or permanent. When hearing loss is measured at a postexposure time of 2 weeks, it is considered to be permanent, inasmuch as very little additional recovery occurs at postexposure times in excess of 2 weeks (Miller, Watson, and Covell, 1963; Mills, 1973). The presence and severity of a temporary or permanent hearing loss depend on several factors and the susceptibility of the individual. Acoustically, the level (intensity), spectrum (frequency), and temporal properties (duration, intermittency, number) of the exposures are the most pertinent properties. Nonacoustic factors such as interactions with various medicines or drugs, eye color, smoking, sex, and other personal characteristics of an individual (except aging and hearing loss) are second-order effects, or are not consistently observed (Ward, 1995). Interactions with other factors such as chemicals, solvents, and toxic substances such as carbon monoxide can be significant and are an active area of investigation (Morata and Dunn, 1995).

The database available on noise-induced hearing loss caused by exposure to steady-state noise is substantial. Some of the data are from laboratory studies of temporary effects in human subjects (Davis et al., 1950; Mills et al., 1970; Melnick and Maves, 1974) and both temporary and permanent effects in laboratory animals (Miller, Watson, and Covell, 1963; Mills, 1973). Other data are from studies of permanent hearing loss in humans in occupational settings (Taylor et al., 1965; Burns and Robinson, 1970; Johnson, 1991). These and other field studies of noise-induced permanent hearing loss (see Johnson, 1991) are the scientific bases of international (International Organization for Standardization [ISO], 1990) and American standards (American National Standards Institute [ANSI], 1996), which present methods to estimate noise-induced permanent threshold shifts as a function of A-weighted sound level and years of exposure time. The development and acceptance of these standards represents many years of work and intense debate. Regulations for industry are given by the U.S. Department of Labor (Occupational Safety and Health Administration [OSHA], 1983).

Data from humans and animals exposed to a wide variety of steady-state noises suggest that the range of human audibility can be categorized with respect to risk of acoustic injury of the ear and noise-induced hearing loss. This categorization is shown in Figure 1, where the range of human audibility is bounded by the threshold of audibility at one extreme and the threshold of pain (Yost and Nielsen, 1985) at the other. Of course, sounds below the threshold of audibility are inaudible and present no risk of noise-induced hearing loss. Sounds in excess of the threshold of pain present a risk of acoustic injury of the ear and noise-induced hearing loss even with one, short exposure (see Mills et al., 1993). Between the extremes of pain and audibility are acoustic injury thresholds (open triangles in Fig. 1). These thresholds define the highest levels of noise that will not produce a noise-induced threshold shift regardless of the duration of exposure, the number of exposures, or the temporal properties of the exposure. The data points in Figure 1 are from temporary threshold shift experiments for octave bands of noise from 63 Hz to 4 kHz in octave steps. Data at lower and higher frequencies are extrapolations. Thus, between the extremes of the threshold of pain and audibility are two categories: no risk and risk. The no-risk category can also be described as “effective quiet” (Ward, Cushing, and Burns, 1976). The region bounded by safe levels on the low side and threshold of pain on the high side is the area where the risk of hearing loss and acoustic injury of the ear depends on the parameters of the noise exposure as well as on the susceptibility of the individual. In qualitative terms, risk increases with noise level, duration, number of exposures, and susceptibility of the individual. Although individual differences can be substantial, no method has been developed that allows the a priori identification of those individuals who are most susceptible to noise-induced hearing loss. Quantitative relations between noise-induced hearing loss and exposure parameters are given in ANSI S3.44-1996.

Figure 1..  

Categorization of the range of human audibility with respect to acoustic injury of the ear and noise-induced hearing loss. (From Mills, J. H., et al., 1993, Hazardous Exposure to Steady-State and Intermittent Noise. Working Group Report, Committee on Hearing, Bioacoustics and Biomechanics, National Research Council. Washington, DC: National Academy Press. Reproduced with permission.)

Whereas the database is massive for noise-induced hearing losses produced by exposure to continuous noise (see Johnson, 1991), it is unimpressive for intermittent (quiet periods of a few seconds to a few hours) and time-varying (level fluctuations greater than 10 dB) noises. On a qualitative basis, there is agreement that intermittent and fluctuating noises are less hazardous than continuous noises, presumably because the “quiet” periods allow time for the ear to recover. Because of regulatory efforts in noise control and a perceived need for simplicity, several single-number correction factors have evolved. That is, as an exposure is increased from 4 hours to 8 hours, what change in noise level is needed to maintain an equal risk of hearing loss? The equal-energy rule specifies a 3-dB reduction in noise level for a doubling of exposure duration. This rule is incorporated into the ISO 1999 standard. Other standards and regulations use 4-dB, 5-dB, and 6-dB rules, as well as other more complicated schemes (see Ward, 1991). It is likely that each of these single-number rules may apply only to a restricted set of exposure conditions. With the continued absence of needed data, the effects of intermittence on noise-induced hearing loss may always be a contentious issue.

Although the biological bases of noise-induced hearing loss have been studied extensively, with the greatest emphasis on the cochlea, both the external and middle ear play a prominent role. The external auditory meatus (essentially a tube open at one end with a length of about 25 mm) has a resonant frequency of about 3 kHz and a gain of about 20 dB. Thus, the typical industrial or environmental noise, which may have a flat or slightly downward-sloping spectrum when measured in the field, will have a peak at 3 kHz because of the external ear canal. Thus, the “4-kHz notch” in the audiogram that is characteristic of noise-induced hearing loss (and head injuries) may reflect the acoustic properties of the external ear. In the middle ear, the acoustic reflex, the consensual contraction of the stapedial and tensor tympani muscles, may reduce the level of intense sounds. In addition, the efferent innervation of outer hair cells may have a protective role (Maison and Liberman, 2000).

Although there has been a substantial effort, the anatomical, chemical, and biological bases of noise-induced temporary threshold shift are unknown. The pathological anatomy associated with noise-induced permanent hearing loss involves the organ of Corti, especially the hair cells. Loss of outer hair cells is the most prominent anatomical feature of permanent noise-induced loss, and is almost always greater than the loss of inner hair cells. This greater loss of outer than inner hair cells may occur for several reasons, including the direct shearing forces on the outer hair cell stereocilia, which are embedded in the tectorial membrane. The correlations between loss of hair cells, both inner and outer, and permanent threshold shift are very high, ranging from 0.6 to 0.8, depending on frequency (Hamernik et al., 1989). Even with such high correlations there remains considerable variance between hair cell loss and permanent threshold shift. This variance can be reduced by consideration of the status of the stereocilia (Liberman and Dodds, 1984). With degeneration of inner hair cells following severe exposures, there can be retrograde degeneration of auditory nerve fibers, as indicated by losses of spiral ganglion cells. Neural degeneration is not restricted to the auditory nerve but progresses throughout the ascending auditory system (Morest, 1982). Regeneration of hair cells has been observed after intense exposures to noise. This dramatic effect has been reported only for the cochlea of various species of birds. Regenerating sensory cells have not been observed in the cochlea of mammals.

Reactive oxygen species and oxidative stress have been implicated in the production of noise-induced hearing loss and in age-related hearing loss as well (Ohlemiller et al., 2000). It is believed that acute impairment of antioxidant defenses promotes cochlear injury, and conversely, strengthening antioxidant defenses should provide protection, including possibly rescuing cells that are in the early stages of injury. Efforts at prevention and rescue are in the early stage of development, with some promising initial results (Hu et al., 1997). Additional protective functions can be obtained by conditioning exposures or exposures that protect the ear from subsequent noise (Canlon, Borg, and Flock, 1988). A related phenomenon is improvements (reductions) in threshold shifts observed during the course of an extended sequence of intermittent exposure to noise (Miller, Watson, and Covell, 1963; Clark, Bohne, and Boettcher, 1987). Clearly, noise-induced hearing loss is not related simply to the sound level of the exposure.

Acoustic trauma refers to injury of the ear and permanent hearing loss caused by exposure to an intense, short-duration sound. In contrast to the gradual loss of outer hair cells and stereocilia typically seen from steady-state or intermittent exposures with sound levels less than 100–110 dB SPL, the injury to the organ of Corti is more extensive, involving the tearing of membranes, rupturing of cells, and mixing of cochlear fluids. At extremely high sound levels, the tympanic membrane and middle ear can be injured, with a resultant conductive/mixed hearing loss. The most common form of acoustic trauma is hearing loss associated with the impulsive noises produced by small-arms gunfire (Clark, 1991). Hearing loss from impulses is related to the peak SPL of the impulse, duration, number of impulses, and other variables (Henderson and Hamernik, 1986; Hamernik and Hsueh, 1993; Hamernik et al., 1993).

A longstanding issue is the interaction between noise-induced hearing loss incurred throughout a person's working lifetime and the hearing loss associated with aging. This issue is particularly important for medical reasons, for litigation involving worker's compensation for occupational hearing loss, and in establishing noise standards and damage-risk criteria. Both the current ISO (1990) and ANSI (1996) standards assume that noise-induced permanent threshold shifts add (in decibels) to age-related threshold shifts, i.e., a 20-dB loss from noise and a 20-dB loss from aging results in a loss of 40 dB. Some data from noise-exposed and aged animals do not support additivity in decibels but additivity in intensity: i.e., 20 dB and 20 dB produces a loss of 23 dB (Mills et al., 1997). The issues of additivity and medical-legal aspects of noise-induced hearing loss are discussed by Dobie (2001).