Interplay of White Matter Hyperintensities, Cerebral Networks, and Cognitive Function in an Adult Population: Diffusion-Tensor Imaging in the Maastricht Study
Laura W M Vergoossen,Jacobus F A Jansen,Thomas T van Sloten,Coen D A Stehouwer,Nicolaas C Schaper,Anke Wesselius,Pieter C Dagnelie,Sebastiaan Köhler,Martin P J van Boxtel,Abraham A Kroon,Joost J A de Jong,Miranda T Schram,Walter H Backes
DOI: https://doi.org/10.1148/radiol.2021202634
IF: 19.7
Radiology
Abstract:Background Lesions of cerebral small vessel disease, such as white matter hyperintensities (WMHs) in individuals with cardiometabolic risk factors, interfere with the trajectories of the white matter and eventually contribute to cognitive decline. However, there is no consensus yet about the precise underlying topological mechanism. Purpose To examine whether WMH and cognitive function are associated and whether any such association is mediated or explained by structural connectivity measures in an adult population. In addition, to investigate underlying local abnormalities in white matter by assessing the tract-specific WMH volumes and their tract-specific association with cognitive function. Materials and Methods In the prospective type 2 diabetes-enriched population-based Maastricht Study, structural and diffusion-tensor MRI was performed (December 2013 to February 2017). Total and tract-specific WMH volumes; network measures; cognition scores; and demographic, cardiovascular, and lifestyle characteristics were determined. Multivariable linear regression and mediation analyses were used to investigate the association of WMH volume, tract-specific WMH volumes, and network measures with cognitive function. Associations were adjusted for age, sex, education, diabetes status, and cardiovascular risk factors. Results A total of 5083 participants (mean age, 59 years ± 9 [standard deviation]; 2592 men; 1027 with diabetes) were evaluated. Larger WMH volumes were associated with stronger local (standardized β coefficient, 0.065; P < .001), but not global, network efficiency and lower information processing speed (standardized β coefficient, -0.073; P < .001). Moreover, lower local efficiency (standardized β coefficient, -0.084; P < .001) was associated with lower information processing speed. In particular, the relationship between WMHs and information processing speed was mediated (percentage mediated, 7.2% [95% CI: 3.5, 10.9]; P < .05) by the local network efficiency. Finally, WMH load was larger in the white matter tracts important for information processing speed. Conclusion White matter hyperintensity volume, local network efficiency, and information processing speed scores are interrelated, and local network properties explain lower cognitive performance due to white matter network alterations. © RSNA, 2020 Online supplemental material is available for this article.