Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Neurophysiology of the aging language network: An individual-level investigation of topography and timing
Poster Session B, Wednesday, September 30, 4:30 - 6:30 pm, Wangari Maathai
Stéphanie De Keulenaer1, Rose Bruffaerts1,2; 1UAntwerp, 2University Hospital Antwerp (UZA)
Although language remains largely preserved with age (Shafto & Tyler, 2014), neurophysiological studies report age-related neural changes in the timing of sentence processing (Pagán et al, 2025). Furthermore, aging is associated with higher interindividual variability, and differences in neurophysiological responsivity relate to individual cognitive performance (Bruffaerts et al., 2019), requiring an individualized approach. The functional language identification approach has been extensively validated in young adults using fMRI (Fedorenko et al., 2010; 2024) and MEG (Huybrechts & Bruffaerts, preprint), and recent fMRI work shows the language network's responsiveness and topography are preserved in aging (Billot et al., preprint). Here, we characterize this well-known paradigm using EEG to study age-related changes, reporting the regions most responsive to sentences, as well as their stability and variability at the individual level. We collected hd-EEG (64 channels) in 28 native Dutch-speaking young controls (YHC; mean age: 24.2, education: 16.9 years) and 31 older controls (OHC; mean age: 69.4, education: 14.8 years) matched for gender. The language task contained 80 sentences of 12 words, and 20 control sequences of 12 nonwords, matched for phonological properties and syllable length (adapted to Dutch from Fedorenko et al., 2010). Stimuli were presented sequentially for 500ms. Each sequence was followed by a probe judgement to assess attention and behavioural performance. A collapsed localizer was used to define a group-level ROI and time window, from which mean amplitude and latency of the sentence condition were extracted. To test whether EEG allows for robust identification of subject-specific language-responsive sensors, we evaluated the stability of the sensors within individuals across trials. Specifically, we examined the Spearman correlation of the sentence effect (signal change for the sentence condition vs control condition) across all sensors within each participant across the first and second half of trials and compared this to between-participant correlations (Huybrechts & Bruffaerts, preprint). Behaviourally, OHC responded slower than YHC (1364ms vs. 1123 ms, p=.005), while no significant group differences were found for accuracy. The language task elicited the strongest ERP between 166–266 ms over frontocentral regions, with the same region emerging when YHC and OHC were analyzed separately. Within the frontocentral ROI, P200 amplitude was significantly decreased in OHC (2.61 µV) compared to YHC (3.65 µV, p=0.04) and latency was significantly prolonged in OHC (222.6ms) compared to YHC (213.38ms, p=0.009). Topographical response patterns across sensors within this timewindow were stable within participants and distinguishable between individuals in both groups, with young and older controls showing similar within-participant correlations (within ρ = 0.61–0.62 vs. between ρ = 0.02–0.05, both p < 0.001). In line with prior fMRI work, we demonstrate that the neurophysiological correlates of the aging language network are largely preserved. Although language performance remains stable with age, both ERP and behavioural measures revealed slower neurophysiological processing. The stability of the individualized framework in both young and older controls enables further research into individual differences in neurophysiological correlates of language. A potential application is the use of task-based EEG to discriminate between normal and abnormal cognitive aging (De Keulenaer et al., 2025).
Topic Areas: Methods, Computational Approaches