Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Grounding discourse-wide semantics: Source-level EEG connectivity signatures of action and social texts in Parkinson's disease and behavioral variant frontotemporal dementia
Poster Session E, Friday, October 2, 11:00 am - 1:00 pm, Wangari Maathai
Emma Rico1,2, Agustina Birba3, Lucía Amoruso3,4,5, Miquel Martorell3, Paula Arellano6, Andrea Slachevsky7,8,9, María Isabel Behrens7,10,11,12, Martín A. Bruno6,13, Agustín Ibañez3,14,15,16,17, Adolfo M. García3,14,18; 1Instituto Universitario de Neurociencia, Universidad de La Laguna, Tenerife, Spain, 2Departamento de Psicología Cognitiva, Social y Organizacional. Universidad de La Laguna, Tenerife, Spain, 3Cognitive Neuroscience Center (CNC), Department of Life and Behavioral Sciences, University of San Andres, Buenos Aires, Argentina, 4Basque Center on Cognition, Brain and Language (BCBL), San Sebastian, Spain, 5Ikerbasque, Basque Foundation for Science, Bilbao, Spain, 6Instituto Ciencias Biomedicas (ICBM), Facultad Ciencias Médicas, Universidad Catolica de Cuyo, San Juan, Argentina, 7Departamento de Neurología y Psiquiatría, Clínica Alemana - Universidad del Desarrollo, Santiago, Chile, 8Laboratorio de Neuropsicología y Neurociencias Clínicas (LANNEC) Programa de Fisiopatología − ICBM, Departamento de Ciencias Neurológicas Oriente y Neurociencias, Facultad de Medicina, Universidad de Chile, Santiago, Chile, 9Centro de Memoria y Neuropsiquiatría (CMYN), Servicio de Neurología, Hospital del Salvador y Facultad de Medicina, Universidad de Chile, Santiago, Chile, 10Centro de Investigación Clínica Avanzada (CICA), Hospital Clínico Universidad de Chile, Santiago, Chile, 11Departamento de Neurología y Neurocirugía, Hospital Clínico Universidad de Chile, Santiago, Chile, 12Departamento de Neurociencia, Facultad de Medicina, Universidad de Chile, Santiago, Chile, 13National Council on Scientific and Technical Research (CONICET), Buenos Aires, Argentina, 14Global Brain Health Institute (GBHI), University of California San Francisco, United States; Trinity College Dublin (TCD), Dublin, Ireland, 15Latin American Brain Health Institute (BrainLat), Universidad Adolfo Ibañez, Santiago de Chile, Chile, 16Department of Biophysics, School of Medicine, Istanbul Medipol University, Istanbul, Türkiye, 17Barcelonaβeta Brain Research Center (BBRC), Pasqual Maragall Foundation, Barcelona, Spain, 18Departamento de Lingüística y Literatura, Facultad de Humanidades, Universidad de Santiago de Chile, Santiago, Chile
Embodied cognition theories posit that action and social concepts recruit neural circuits subserving physical movement and socio-cognitive skills, respectively. However, most evidence comes from artificial tasks involving decontextualized linguistic units and thus limiting ecological validity. Here we bridge these gaps by capturing source-level spectral connectivity signatures of action and social text processing in persons with Parkinson’s disease (PD, typified by motor symptoms), behavioral variant frontotemporal dementia (bvFTD, characterized by sociobehavioral disruptions), and Alzheimer’s disease (AD, as a disease control group), all relative to healthy controls (HCs). Eighty participants (20 per group) listened to four validated, carefully matched naturalistic stories: an action text (rich in motor events), a social text (with abundant socio-interactive scenes), and two control texts (non-action, non-social). Electroencephalography data was acquired using a 128 Biosemi Active-two system. Recordings were segmented into two-second epochs, enabling the capture of low- to high-frequency modulations. Following preprocessing, a global forward model based on FreeSurfer’s FsAverage template was computed (source space spacing = oct6; BEM = 3 layers). Data were mapped to source space using individualized eLORETA inverse operators and then collapsed into 68 cortical regions of the Desikan-Killiany atlas. Functional connectivity was computed using the weighted Phase Lag Index across theta (4-8 Hz), alpha (8-13 Hz), and beta (13-30 Hz) frequency bands. To identify significant sub-network differences between clinical groups and HCs, we utilized Network-Based Statistics (edge threshold of T > 3.0, 2000 permutations; cluster alpha of p < 0.01), employing a spatial adjacency matrix with interhemispheric bridges between homologous regions. To rule out spurious or marginal connections, only clusters comprising four or more nodes were included. Relative to HCs, PD patients exhibited hyperconnectivity during action text processing along fronto-cingular hubs in the theta band (31 edges, p = 0.001) and fronto-posterior hubs in the beta band (16 edges, p < 0.009), with no differences for other texts. BvFTD patients exhibited fronto-insulo-temporal theta hyperconnectivity (17 edges, p < 0.007) for the social text and beta hyperconnectivity for the action text (23 edges, p < 0.006), alongside null effects for both control texts. AD patients showed a non-specific pattern, involving fronto-temporal alpha (5 edges, p = 0.001) and theta (9 edges, p < 0.001) hypoconnectivity for the social text, fronto-cingulate theta hypoconnectivity for the non-social text (2 clusters: 4 edges, p < 0.003, 4 edges, p < 0.003), and both frontotemporal alpha hypoconnectivity (4 edges, p < 0.001) and alpha hyperconnectivity (12 edges, p = 0.009) for the non-action text. Processing action-laden and socially rich texts elicited distinct anomalies in PD and bvFTD, respectively. These alterations mapped directly onto the domain-specific neural networks intrinsically degraded in each condition. The specificity of these patterns is reinforced by the broad pattern observed in AD. Taken together, our results indicate that embodied motor and social grounding extends beyond isolated words and sentences, being dynamically driven during naturalistic text processing by domain-preferential anatomo-frequential networks. These findings illustrate the potential of discourse-level paradigms and strategic neuropathologies to advance ecologically valid models in the embodied cognition framework.
Topic Areas: Speech Perception, Meaning: Lexical Semantics