Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Changes in EEG beta functional connectivity during narrative listening reflect linguistic-pragmatic impairment in schizophrenia
Poster Session E, Sunday, September 14, 11:00 am - 12:30 pm, Field House
Federico Frau1, Riccardo Venturini1, Paolo Canal1, Romain Aubonnet2, Giulia Agostoni3, Giorgio Arcara4,5, Roberto Cavallaro3,6, Giorgio Di Lorenzo2,7, Marta Bosia3,6, Valentina Bambini1; 1Laboratory of Neurolinguistics and Experimental Pragmatics (NEPLab), University School for Advanced Studies IUSS Pavia, Pavia, Italy, 2Laboratory of Psychophysiology and Cognitive Neuroscience Department of Systems Medicine, University of Rome Tor Vergata, Rome, Italy, 3School of Medicine, Vita-Salute San Raffaele University, Milan, Italy, 4Department of General Psychology, University of Padua, Padua, Italy, 5IRCCS San Camillo Hospital, Venice, Italy, 6Schizophrenia Research and Clinical Unit, IRCCS San Raffaele Scientific Institute, Milan, Italy, 7IRCCS Fondazione Santa Lucia, Rome, Italy
INTRODUCTION. Disorganization in language is a hallmark feature of schizophrenia, which has received attention both at behavioral and neural levels, with studies reporting alterations within the language network[1,2]. However, little is known about functional connectivity (FC) mechanisms underlying language difficulties[3], including higher-level pragmatic abilities, which are especially vulnerable in schizophrenia[4]. Here, we aimed to investigate FC within the language network in a group of participants with schizophrenia undergoing an ecological linguistic task, i.e., narrative listening, linking it to participants’ linguistic-pragmatic profile. METHODS. We recruited 49 participants with schizophrenia based on DSM-5 criteria (Age: 38.34±12.75, Education: 11.50±2.22, Illness Duration: 16.51±11.33, Chlorpromazine-equivalent dose: 492.91±500.32), clinically stabilized and treated with the same antipsychotic therapy for at least two months, assessed for pragmatic abilities using the APACS Test[5]. Participants were engaged in a passive-listening task of a 5-minute-long narrative during EEG recording. EEG signal was acquired using a 128-channel system and underwent an automated pre-processing pipeline, including a 0.1Hz high-pass filter, 50Hz notch filter, bad channels removal and interpolation, and artifacts correction with Independent-Component-Analysis and Artifact-Subspace-Reconstruction. From source-level oscillations (computed with eLORETA), we extracted static connectivity matrices with Phase Locking Value (PLV), measuring the consistent alignment in time of oscillations between two ROIs. We computed PLV values in five frequency bands (alpha, lower-beta, higher-beta, lower-gamma, higher-gamma) across 39 ROIs from Destrieux’s cortical atlas, reflecting the Extended Language Network of narrative comprehension[6]. A representation of the pipeline is available in Figure1 (https://tinyurl.com/4k5vdxfc). Statistical analysis focused on the comparison between participants above the APACS total score cut-off (Unimpaired, n=16) and participants below the cut-off and with different degrees of pragmatic impairment, defined via median split as Mildly-Impaired (n=16) and Severely-Impaired (n=17). Differences in PLV values were analyzed using permutation tests (50,000 permutations) across ROIs and frequency bands (FDR-corrected). RESULTS. The Severely-Impaired group, compared to Unimpaired participants, showed a significantly reduced FC in lower-beta oscillations between the left angular and the right middle frontal gyri (p=.023). They also exhibited a significantly increased FC, compared to Unimpaired participants, in lower-beta oscillations between the left transverse temporal sulcus and the right posterior superior temporal gyrus (p=.023). See Figure2 (https://tinyurl.com/33hsy48z). Between-group differences remained significant after controlling for Age, Illness Duration, and Chlorpromazine-equivalent dose via linear regression analysis. CONCLUSIONS. Here, we found differences in beta FC within the language network during narrative listening, reflecting different pragmatic profiles. Our results confirm the importance of fronto-temporo-parietal connections and beta oscillations in discourse comprehension[7]. A possible explanation for changes in beta FC might be related to altered predictive mechanisms involved in language[8]. Our data suggest a link between these changes in the extended language network and pragmatic impairment in schizophrenia, supporting the hypothesis that altered predictive mechanisms at the oscillatory level underlie language disorganization in this population[3]. Our findings have implications for the neurophysiological characterization of linguistic difficulties in schizophrenia, as well as for understanding the neural underpinnings of linguistic-pragmatic skills. REFERENCES. [1]Kircher et al.LancetPsychiatry.2018;5(6):515-526. [2]Palaniyappan et al.SchizophrBull.2023;49(2):486-497. [3]Palaniyappan.SeminCellDevBiol.2022;129:47-60. [4]Bambini et al.ComprPsychiatry.2016;71:106-120. [5]Arcara & Bambini.FrontPsychol.2016;7:70. [6]Ferstl et al.HumBrainMapp.2008;29(5):581-593. [7]Yang et al.Neuroimage.2019;203:116200. [8]Weiss & Mueller.FrontPsychol.2012;3:201.
Topic Areas: Meaning: Discourse and Pragmatics, Disorders: Developmental