Tinnitus, masked speech perception, and auditory event-related potentials in clinically normal-hearing adults.

Tinnitus can increase stress and anxiety and reduce quality of life. Compared to listeners with normal hearing and no tinnitus, listeners with tinnitus (with or without hearing loss) often have greater difficulty segregating competing speech. Previous studies have investigated neural correlates of tinnitus using the P3 component of event-related potentials (ERPs). Speech reception thresholds (SRTs) and ERPs were measured in 29 listeners with tinnitus and 25 listeners without tinnitus; all were clinically normal hearing, with pure-tone average (PTA) thresholds <25 dB HL across 500, 1000, 2000, and 4000 Hz. SRTs for a male target talker were measured in speech-shaped noise (SSN; energetic masking) or with a single female speech masker (informational masking). Auditory ERPs were measured for a 1 kHz vs. 2 kHz contrast. SRTs in competing speech were significantly lower for the non-tinnitus than for the tinnitus group (p = 0.042); there was no significant group difference for SRTs in SSN. Latency was significantly longer for the tinnitus than for the non-tinnitus group for N1 (p = 0.034) and P3 (p < 0.001), but not for P2. There were no significant group differences in terms of P2-N1 or P3 amplitude. Multilinear regression analysis showed that age at testing (p = 0.005), N1 latency (p = 0.031), P3 latency (p = 0.020), and P3 amplitude (p = 0.031) were significant predictors of SRTs in competing speech. The present results suggest that among adults with clinically normal PTAs, susceptibility to informational masking may be greater for listeners with than without tinnitus.