White matter network structure as a substrate of cognitive brain reserve in cerebral small‐vessel disease: The Maastricht Study

Abstract

Method The Maastricht Study is a population‐based cohort study with extensive phenotyping, enriched for type‐2 diabetes. Cognitive test scores, structural connectivity data, and 3T MRI were available for n=4798 participants (mean age(±SD)=59.2(±8.7), 50.0% male). Images were assessed by Fazekas score (£1 or ≥2), presence of microbleeds and lacunar infarcts in terms of dichotomous measures, and combined into a cSVD score (range 0‐3). Node degree (ND), a measure of the mean interconnectedness of nodes in the structural connectome, was used as connectivity score. A composite cognition score (CS) was taken as the mean score across three cognitive domains (Memory, Information Processing Speed and Executive Function). Multivariable linear regression analyses were used to investigate the interaction effect between ND and cSVD score on CS, adjusted for age, sex, education, and diabetes status. Result Higher cSVD score was associated with lower CS. ND modified the association between cSVD score and CS in the adjusted model (p for interaction p<0.004). While there were clear cognitive differences across cSVD levels if ND was low, the association between cSVD and cognition was attenuated in a dose‐response fashion with increasing ND (Figure 1). Conclusion Structural ND provides evidence of an individual’s CR based on white matter organization in the presence of damage from cSVD. Even in participants with the highest cSVD burden, sufficiently high ND score was associated with normal CS.