Investigating how narrator acoustics influence continuous speech-evoked multiband auditory brainstem responses

Frequency-specific auditory brainstem responses (ABRs) are used to assess hearing function and can now be measured to continuous “peaky” speech audiobook stimuli. However, these speech-evoked multiband ABRs are small. The amplitudes can be boosted if using narrators with lower fundamental frequencies (f0s) and peaky speech composed of chirp-phase profiles. But even if set to the same optimal parameters, different narrators and stories evoke differently sized multiband ABRs, thereby impacting the length of testing required to ensure all participants have visible responses. This may limit what audiobooks can feasibly be used. Thus, this study aimed to determine what factors of the remaining narrator/story differences contribute to the variability of ABR amplitudes. Data comprised both measured ABRs across three previous studies in adults with normal hearing and modeled ABRs of the same stimuli. Stimuli comprised stories from different narrators and two stories from the same narrator, all with a chirp-phase profile and mean f0 of 90 Hz. Response amplitude differences were compared and correlated with stimulus f0 distribution statistics, such as variation, kurtosis, and skewness. Understanding which stimulus factors most influence ABR responses will expand the diversity of audiobooks that can be used for feasible and engaging multiband ABR measurement.