Distortion product otoacoustic emissions in normal and hearing-impaired ears

Abstract

Otoacoustic emissions of distortion products (DPOAE's) were recorded in normal and hearing-impaired human ears using relatively straightforward methods. Two pure-tone stimuli at fixed frequency levels of 73 dB HL for f 1 and of 67 dB HL for f 2 were used. The frequencies of the two primaries were chosen so that their geometric mean represented standard audiometric frequencies. Measurements of the emission amplitudes at 2f 1−f 2 and the adjacent noise floor were achieved by spectral averaging. A total of 101 subjects (199 ears) were tested. Seventy-seven ears in 46 subjects had normal hearing (hearing levels ⩽20 dB at standard audiometric frequencies; average hearing levels, ⩽10 dB). Thirty-six ears in 25 subjects had near-normal hearing (no hearing complaints, hearing levels ⩽40 dB; average hearing levels, ⩽20 dB). No significant differences in mean DPOAE values were apparent between these two groups of ears. All but two of these 113 ears (98%) showed emissions at three or more of the six frequencies tested between 1 and 6 kHz. Emissions were detected in more than 75% at each frequency between 1 and 6 kHz and in more than 85% between 1 and 4 kHz. A further 86 ears in 44 subjects exhibited varying degrees of sensorineural hearing loss caused by different pathologies. In general, emission amplitudes approximated the shapes of the audiograms, and a highly significant correlation between hearing thresholds and emission amplitudes was demonstrated in the frequency range of 1 to 4 kHz. We conclude that the testing of distortion product emissions is feasible as a routine clinical procedure, that asymptomatic and mild hearing loss do not change the properties of emissions measured at approximately 70 dB HL, and that there is a highly significant correlation between hearing threshold and emission amplitude in the middle-frequency range.