Characterization of EEG-based functional brain networks in myotonic dystrophy type 1. Issue 8 (August 2020)

• Record Type: Journal Article
• Authors: Biere, Joost; Okkersen, Kees; van Alfen, Nens; Kessels, Roy P.C.; Gouw, Alida A.; van Dorst, Maud; van Engelen, Baziel; Stam, Cornelis J.; Raaphorst, Joost
• Is Part Of: Clinical neurophysiology. Volume 131:Issue 8(2020:Aug.)
• Page Start: 1886
• Page End: 1895
• Publication Date: 2020-08
• Subjects: Muscle disease -- Myotonic dystrophy -- Trinucleotide repeat diseases -- EEG -- Cognition; DM1 myotonic dystrophy type 1 -- EEG electro-encephalography -- FDR false-discovery rate -- PLI phase-lag index; Neurophysiology -- Periodicals; Electroencephalography -- Periodicals; Electromyography -- Periodicals; Neurology -- Periodicals; 612.8
• DOI: 10.1016/j.clinph.2020.05.014
• Languages: English
• ISSNs: 1388-2457

Abstract

Highlights: Resting-state EEG can be used to study functional brain networks. In myotonic dystrophy type 1 (DM1) patients, functional network organization differs from healthy controls. The clinical implications of functional network alterations remain to be further studied. Abstract: Objective: In the autosomal dominant, multisystem, chronic progressive disease myotonic dystrophy type 1 (DM1), cognitive deficits may originate from disrupted functional brain networks. We aimed to use network analysis of resting-state electro-encephalography (EEG) recordings of patients with DM1 and matched unaffected controls to investigate changes in network organization in large-scale functional brain networks and correlations with cognitive deficits. Methods: In this cross-sectional study, 28 adult patients with genetically confirmed DM1 and 26 age-, sex- and education-matched unaffected controls underwent resting-state EEG and neuropsychological assessment. We calculated the Phase Lag Index (PLI) to determine EEG frequency-dependent functional connectivity between brain regions. Functional brain networks were characterized by applying concepts from graph theory and compared between-groups. Network topology was evaluated using the minimum spanning tree (MST). We evaluated correlations between network metrics and neuropsychological tests that showed statistically significant between-group differences. Results: Functional connectivity estimated as whole-brain median PLI for DM1 patients versus healthy controls was higher in theta band (0.141 [0.050] versus 0.125 [0.018], p = 0.029), and lower in the upper alpha band (0.154 [0.048] versus 0.182 [0.073], p = 0.038), respectively. Functional MST-constructed networks in DM1 patients were significantly dissimilar from healthy controls in the delta, (p = 0.009); theta, (p = 0.009); lower alpha, (p = 0.036); and upper alpha, (p = 0.008) bands. In evaluation of local MST network measures, trends toward networks with higher global integration in the theta band and lower global integration in the upper alpha band were observed. Compared to unaffected controls, DM1 patients performed worse on tests of attention, motor function, executive function and visuospatial memory. Visuospatial memory correlated with the global median PLI in the upper alpha band; the Stroop interference test correlated with betweenness centrality in this band. Conclusion: This study supports the hypothesis that brain changes in DM1 give rise to disrupted functional network organization, as modelled with EEG-based networks. Further study may help unravel the relations with clinical brain-related DM1 symptoms. Significance: EEG network analysis has potential to help understand brain related DM1 phenotypes. Funding: This work was supported by the European Community's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 305697 (OPTIMISTIC) and the Marigold Foundation.