Action language processing in Parkinson’s disease: Characterization of neuro-oscillatory dynamics and linguistic performance
Human language capacity is based on temporally coordinated neural activity across distributed brain regions. Although the left hemispheric perisylvian cortex constitutes the core region of language processing, a network of additional sites is further involved. For example, in the healthy brain, sema...
Saved in:
Main Author: | |
---|---|
Contributors: | |
Format: | Doctoral Thesis |
Language: | English |
Published: |
Philipps-Universität Marburg
2022
|
Subjects: | |
Online Access: | PDF Full Text |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | Human language capacity is based on temporally coordinated neural activity across distributed brain regions. Although the left hemispheric perisylvian cortex constitutes the core region of language processing, a network of additional sites is further involved. For example, in the healthy brain, semantic access to action concepts has been associated with increased neural activity within frontal motor areas. These findings are complemented by studies demonstrating impaired action language processing in patients with Parkinson's disease, a condition leading to impaired motor control. Therefore, both lines of inquiry suggest an involvement of sensorimotor brain regions in the semantic access to action concepts. However, as the neural underpinnings of the putative action language deficit in Parkinson's disease are unknown, the contribution of motor areas to this phenomenon remains unresolved.
This study therefore aimed at resolving this question by characterizing neurophysiological and behavioral correlates of action language processing in patients with Parkinson's disease. For this purpose, two experiments were carried out. The goal of Experiment 1 was to compile and validate a data set of action pictures for the German language. This part of the study aimed at identifying psycholinguistic variables affecting naming latency in a picture naming task, allowing the selection of matched sets of stimuli in prospective studies. Experiment 2 built upon these data and employed an action naming task and high-density electroencephalography to characterize oscillatory patterns during action language production in both healthy participants and patients with Parkinson's disease. Specifically, this part of the study examined whether action language processing is accompanied with aberrant oscillatory patterns in the mu and beta frequency range over motor cortical areas in the parkinsonian state. Furthermore, the influence of dopaminergic medication on these patterns was assessed.
In Experiment 1, a total of 283 freely available action pictures could be assembled and characterized. The principal variables affecting naming latency describe the agreement in responses across subjects: Less homogeneous response distributions were associated with longer reaction times. Furthermore, word frequency as well as the motor content of the pictures and responses were significant predictors of naming latency. Experiment 2 could not replicate the behavioral action naming deficit in patients with Parkinson's disease when compared to healthy participants. However, differential neurophysiological correlates of action naming were observed. In contrast to healthy subjects, a transient episode of beta hypersynchronization was present over central to frontal electrodes in Parkinson's disease patients off medication within 300 to 700 ms after stimulus presentation. Cluster-based permutation tests confirmed this difference in oscillatory power and by reconstructing the sources of neural activity it could be localized to the left pre- and postcentral cortex and to the right anterior temporal lobe. Furthermore, subsequent mu power suppression (from 800 ms onwards) was stronger in patients with Parkinson's disease than in healthy controls.
The associations between psycholinguistic variables and naming latency found in Experiment 1 were largely consistent with action naming normative studies carried out in other languages. The data set of 283 action pictures may therefore constitute a valuable resource for future psycholinguistic investigations of action language processing. In Experiment 2, behavioral results were not in keeping with a specific action language deficit in patients with Parkinson's disease, which stands in contrast to prior studies. However, patients included in this study attained a higher level of education as those examined in earlier reports, potentially compensating the hypothesized deficit. On the neurophysiological level though, exaggerated beta power in Parkinson's disease patients showed a spatiotemporal pattern which may reflect aberrant semantic access to action concepts grounded in the motor system: Differential neural activity was partly observed during a previously established time frame for semantic processing and located to brain regions that have been associated with access to action concepts, including the sensorimotor cortex.
In conclusion, this study established a methodological basis for further psycholinguistic studies on action language processing by validating a normative action picture data set for the German language. By applying this data set in an action naming task and recording high density electroencephalography in Parkinson's disease patients and healthy controls, neurophysiological correlates of action language processing were examined. While behavioral results were not in keeping with a hypothesized action naming deficit, differential oscillatory activity in the beta frequency range suggests a contribution of the motor system to altered semantic processing of action concepts in patients with Parkinson's disease. |
---|---|
DOI: | 10.17192/z2022.0349 |