Recognition of low-pass-filtered consonants in noise with normal and impaired high-frequency hearing.

Abstract

People with high-frequency hearing loss often complain of difficulty understanding speech, particularly in noisy environments. The reduction in audible high-frequency speech information provides one explanation. In addition, high-frequency hearing loss may reduce the contribution from the "tails" of high-frequency auditory nerve fibers, resulting in diminished availability of lower frequency speech cues. This study was designed to determine if high-frequency hearing loss results in speech-understanding deficits beyond those accounted for by reduced high-frequency speech information. Recognition of speech, both low-pass filtered and unfiltered, was measured for subjects with normal hearing and those with hearing loss limited to high frequencies. Nonsense syllables were presented in three levels of noise that was spectrally shaped to match the long-term spectrum of the speech. Scores for subjects with impaired high-frequency hearing were significantly poorer than scores for subjects with normal hearing. In the case of the low-pass-filtered speech, performance differences between groups could not be attributed to differences in speech audibility, as high-frequency speech cues were absent for all subjects. These results are consistent with the hypothesis that high-frequency fibers encode useful low-frequency speech information.