Subcortical responses to continuous speech under bimodal divided attention.

Abstract

Everyday speech perception often occurs in multimodal environments, requiring listeners to divide attention across sensory modalities to prioritize relevant information. Although this division of attention modestly reduces cortical encoding of natural continuous speech, its impact on subcortical processing remains unclear. To investigate this, we used an audiovisual dual-task paradigm to manipulate bimodal divided attention. Participants completed a primary visual memory task (low or high cognitive load) while simultaneously performing a secondary task of listening to audiobook segments. Sixteen young adults with normal hearing completed these tasks while their EEG signals were recorded. In a third condition, participants performed only the listening task. Subcortical responses to the audiobook segments were analyzed using temporal response functions (TRFs), which predicted EEG responses from speech predictors derived from auditory nerve models. Across all conditions, TRFs displayed a prominent peak at ∼8 ms, resembling the wave V peak of auditory brainstem responses, indicating subcortical origins. No significant differences in latencies or amplitudes of this peak, nor in TRF prediction correlations, were observed between conditions. These findings provide no evidence that bimodal divided attention affects the subcortical processing of continuous speech, indicating that its effects may be restricted to cortical levels.NEW & NOTEWORTHY This study shows that auditory subcortical processing of natural continuous speech remains unaffected when attention is divided across auditory and visual modalities. These findings indicate that the influence of crossmodal attention on the processing of natural continuous speech may be restricted to cortical levels.