Effects of Signal Type and Noise Background on Auditory Evoked Potential N1, P2, and P3 Measurements in Blast-Exposed Veterans

Abstract

Objectives: Veterans who have been exposed to high-intensity blast waves frequently report persistent auditory difficulties such as problems with speech-in-noise (SIN) understanding, even when hearing sensitivity remains normal. However, these subjective reports have proven challenging to corroborate objectively. Here, we sought to determine whether use of complex stimuli and challenging signal contrasts in auditory evoked potential (AEP) paradigms rather than traditional use of simple stimuli and easy signal contrasts improved the ability of these measures to (1) distinguish between blast-exposed Veterans with auditory complaints and neurologically normal control participants, and (2) predict behavioral measures of SIN perception.

Design: A total of 33 adults (aged 19-56 years) took part in this study, including 17 Veterans exposed to high-intensity blast waves within the past 10 years and 16 neurologically normal control participants matched for age and hearing status with the Veteran participants. All participants completed the following test measures: (1) a questionnaire probing perceived hearing abilities; (2) behavioral measures of SIN understanding including the BKB-SIN, the AzBio presented in 0 and +5 dB signal to noise ratios (SNRs), and a word-level consonant-vowel-consonant test presented at +5 dB SNR; and (3) electrophysiological tasks involving oddball paradigms in response to simple tones (500 Hz standard, 1000 Hz deviant) and complex speech syllables (/ba/ standard, /da/ deviant) presented in quiet and in four-talker speech babble at a SNR of +5 dB.

Results: Blast-exposed Veterans reported significantly greater auditory difficulties compared to control participants. Behavioral performance on tests of SIN perception was generally, but not significantly, poorer among the groups. Latencies of P3 responses to tone signals were significantly longer among blast-exposed participants compared to control participants regardless of background condition, though responses to speech signals were similar across groups. For cortical AEPs, no significant interactions were found between group membership and either stimulus type or background. P3 amplitudes measured in response to signals in background babble accounted for 30.9% of the variance in subjective auditory reports. Behavioral SIN performance was best predicted by a combination of N1 and P2 responses to signals in quiet which accounted for 69.6% and 57.4% of the variance on the AzBio at 0 dB SNR and the BKB-SIN, respectively.

Conclusions: Although blast-exposed participants reported far more auditory difficulties compared to controls, use of complex stimuli and challenging signal contrasts in cortical and cognitive AEP measures failed to reveal larger group differences than responses to simple stimuli and easy signal contrasts. Despite this, only P3 responses to signals presented in background babble were predictive of subjective auditory complaints. In contrast, cortical N1 and P2 responses were predictive of behavioral SIN performance but not subjective auditory complaints, and use of challenging background babble generally did not improve performance predictions. These results suggest that challenging stimulus protocols are more likely to tap into perceived auditory deficits, but may not be beneficial for predicting performance on clinical measures of SIN understanding. Finally, these results should be interpreted with caution since blast-exposed participants did not perform significantly poorer on tests of SIN perception.

Conflict of interest statement

The authors have no conflicts of interest to disclose.

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.

Figures

Fig. 1.

Mean audiometric thresholds for each participant group averaged across right and left ears. Error bars = ±1 SD.

Fig. 2.

Grand averaged waveforms obtained for blast-exposed (solid black) and control (broken gray) participants in response to the standard signal for each of the four stimulus conditions tested measured at Cz. N1 and P2 peaks are labeled for each condition.

Fig. 3.

Mean N1 latencies measured in blast-exposed (solid black) and control (broken gray) participants in response to each of the four stimulus conditions. Overall, control participants demonstrated larger changes in N1 latency response to speech signals with and without background babble compared to responses obtained in blast-exposed participants. Error bars = ±1 SD.

Fig. 4.

Grand averaged P3 waveforms obtained for blast-exposed (solid black) and control (broken gray) participants in response to the target signals for each of the four stimulus conditions tested measured at Pz. P3 peaks are labeled in response to each condition.

Fig. 5.

Variability in P3 responses among individuals in the blast-exposed (left; solid black) and control groups (right; broken gray). Potentials were measured at Pz in response to target tones presented in quiet. Individual traces were selected to represent maximum variability within each participant group. Grand averaged waveforms for each group are presented at the bottom for reference.

Fig. 6.

P3 latencies measured in response to tone and speech signals averaged across background condition for blast-exposed (solid black) and control (broken gray) participant groups. Error bars = ±1 SD.

Fig. 7.

Group average responses to each question on the FHQ as well as composite score. Higher scores indicate greater levels of perceived difficulty. Error bars = ±1 SD. *p < 0.05. Only those measures that remained significant after correction for false discovery rate are indicated. FHQ, Functional Hearing Questionnaire.

Fig. 8.

Model predictions of performance on the AzBio at 0 dB SNR (panel A) and the BKB-SIN (panel B). Factors predictive of AzBio performance included N1 latency and amplitude in response to tones in quiet, N1 amplitudes in response to speech in quiet, P2 latency in response to speech in quiet, and P3 amplitude in response to speech in background babble. Factors predictive of BKB-SIN performance included N1 and P2 latencies in response to tones in quiet and N1 latencies in response to tones in noise. Full equations detailing the relationships between predictive factors and outcome measures are presented in the text. SNR, signal to noise ratio.