Speech in noise perception and sound localization, relationship with pure tone audiometry in unilateral hearing loss

Binaural hearing yields different cues necessary for sound localization and speech recognition performances. These cues are altered in the case of monaural hearing due to the alteration of neural mechanisms supporting interaural level differences and interaural time differences. This alteration may vary according to the severity of deafness. In this study, we investigated sound localization and speech in noise performance for 21 unilateral hearing loss (UHL) patients compared to a population of 20 normal hearing subjects (NHS) subjects in binaural condition and fitted with an earplug for the simulation of monaural condition. UHL patients were assigned into two groups according to their pure tone averages (PTA; mild<72.5 dB and severe >72.5 dB). Our findings demonstrated a deficit in speech in noise and in spatial performances after unilateral hearing loss compared to NHS. The results of speech in noise evaluation reported better sppech reception thresholds (SRT) for moderate hearing loss compared to the sever population only in dichotic condition, suggesting a possible link between PTA and SRT. Furthermore, moderate UHL showed better localization accuracies than severe UHL in addition to a strong correlation between PTA and root mean square errors (RMS). However, we did not find a correlation between RMS errors and SRT levels. In addition, when compared to NHS, speech in noise deficit was present for moderate and severe UHL; however, moderate UHL demonstrated higher localization accuracies compared to NHS with monaural plug. These results suggest that pure tone audiometry can be a good predictor of sound localization and possibly, in speech in noise perception in the dichotic condition; however, spatial abilities and speech in noise recognition are two different processes that are not directly related at the behavioral level. UHL patients seem to rely more on monaural spectral cues for spatial location rather that speech in noise segregation.