Fine structure of distortion product otoacoustic emissions: its dependence on age and hearing threshold and clinical implications

Distortion product otoacoustic emissions (DPOAE) are routinely used in audiological diagnostics. When the stimulus frequencies f1 and f2 are varied in small steps, distinct non-monotonicities (peaks and valleys) in DPOAE level versus frequency functions can be observed. This so-called DPOAE fine structure (FS) is supposed to be the result of physiological interferences between two different cochlear sources which generate the DPOAE signal. Although FS can complicate interpretations with respect to cochlear functional status at the primary source near f2, its presence might also be relevant in clinical audiological diagnostics. It is therefore of interest to determine FS prevalence and its dependence on age, frequency and hearing threshold. First, it was screened for FS using two tone stimulation (L1/L2 = 55/45 dB SPL, f2/f1 = 1.22) and frequency steps of 40 Hz in the frequency range of 1.8-4.2 kHz. DPOAE (2f1 - f2) were then recorded in 1/3 octave-bands centered around f2 = 2, 3 and 4 kHz with a frequency resolution of 12.5, 20 and 25 Hz, respectively, both with and without a third stimulus (L3 = 45 dB SPL, f3 = 2f1 - f2 + 25 Hz) which was supposed to act as a suppressor of FS. Results of measurements in 102 human individuals from a mixed patient population are reported. Prevalence of DPOAE and FS in a specific frequency range, (i.e. 2, 3, or 4 kHz) was classified into five categories: I) distinct FS within the respective frequency range, II) "single dip" in DP-gram, III) "flat" DP-gram well above noise floor but no distinct FS, IV) DPOAE near noise floor with "irregular responses", and V) no DPOAE measurable. The prevalence of the categories was set in relation to the subject's age and the audiometric threshold at the corresponding center frequency. The estimated probability for a FS (category I and II) was 50-80% if hearing threshold was better than 10 dB HL at the corresponding center frequency. FS prevalence strongly decreased with increasing hearing loss (P < 0.0001). There was no statistically significant age effect (P = 0.088). In more than 50% of the subjects with a behavioral hearing threshold of 10 dB HL or better, a distinct FS near the according frequency was detected, given the presented measurement conditions. While further research is directed at optimal suppression of the second cochlear source of DPOAE and thereby of FS in order to obtain information about the cochlear status near f2 only, the evaluation of FS itself may be clinically useful for monitoring subtle cochlear changes, e.g. during exposure to ototoxic substances or noise.