Neurophysiological Functional Connectivity Changes during Difficult Listening in Older and Younger Adults

Abstract

Older adults often report increased difficulty understanding speech in noisy listening environments. These difficulties are thought to arise from neurophysiological changes associated with aging, including at the level of cortex. Speech comprehension is believed to rely on coordinated activity across distributed cortical regions, mediated by the directional flow of neural signals - their functional connectivity. This connectivity can be analyzed within specific frequency bands, e.g., Delta and Theta, each thought to play a distinct role in speech and language processing. Here, we utilize the Network Localized Granger Causality (NLGC) framework applied to magnetoencephalography (MEG) data to simultaneously estimate cortical source current activities and the their directed functional connectivity. Whole-brain connectivity graphs were analyzed using graph-cluster enhancement and cluster testing. Temporal Response Functions (TRFs) were obtained by relating NLGC-derived neural current estimates to the speech stimulus presented to listeners. The band-specific networks we identify are consistent with the canonical "dual-stream" model of the auditory processing pathway. Delta-band networks primarily involve temporofrontal connections, whereas Theta-band networks showed stronger temporoparietal connectivity. These cortical networks are shown to evolve as listening conditions become increasingly adverse, incorporating frontoparietal areas associated with attention orienting and control. We find that older adults exhibit increased connectivity within areas of the ventral and dorsal auditory streams across all listening conditions, suggesting increased cortical recruitment to support speech comprehension. Although older adults exhibit enhanced time-locked responses to the speech stimulus, these responses did not significantly coincide with their connectivity patterns (and similarly for younger adults), indicating that their connectivity increases are not explained by central gain mechanisms. Together, our findings demonstrate age- and condition-dependent reorganization of speech-related cortical connectivity across neurophysiological frequency bands.