Aging Is Associated with Reduced Topographic Differentiation Between Speaking and Silent Reading

Abstract

Aging affects how neural activity is organized and specialized, but its impact on neural differentiation across language tasks remains unclear. In this study, we investigated whether aging influences neural differentiation during speaking and silent reading. Fifteen young adults and 15 older adults completed speaking and silent-reading tasks in a delayed-response paradigm while we collected electroencephalography (EEG) data. We examined EEG signal strength using time-resolved global field power (GFP), spatial configuration using topographic analysis of variance (TANOVA), and task-specific neural differentiation using the global dissimilarity index (DISS). We also examined the relationship between neural and behavioral measures, including digit-span scores and speech onset time. We found that speaking was associated with higher GFP than silent reading in the 388-800 ms interval, with no significant effects of group or group x task interaction, indicating similar EEG strength across groups. TANOVA revealed age-related differences in scalp topography in the 229-334 ms and 377-538 ms intervals, a task effect in the 618-723 ms interval, and a group x task interaction in the 562-636 ms interval. Notably, young adults had greater DISS than older adults during the 508-629 ms and 660-800 ms intervals, reflecting more distinct neural configurations for speaking and reading. We did not find significant correlations between DISS or GFP and behavioral measures, and the two groups performed similarly on behavioral measures. Overall, aging was associated with reduced neural differentiation between speaking and silent reading, especially during later processing stages (e.g., motor command preparation), while preserving overall EEG activity and behavioral performance. These findings may suggest that aging reduces neural network specificity (as previously suggested) and alters spatial organization without necessarily impairing behavior.