Impact of White Matter Hyperintensity Changes on Cognition One Year After Mild Ischemic Stroke

Abstract

AbstractBackgroundWhite matter hyperintensities (WMH) are a main MRI feature of cerebral small vessel disease, and are associated with increased risk of stroke and dementia. Progression can lead to cognitive decline, but some evidence suggests that WMH can also regress. Regression might relate to better clinical outcomes, but this is not well examined yet. We explored if net WMH volume change accounts for changes in cognition 1‐year after mild ischemic stroke. We also explore tissue signatures in areas of white matter change using brain diffusion imaging.MethodIn a prospective study, we assessed vascular risk factors (VRF), structural and diffusion‐tensor brain MRI and cognition (Montreal Cognitive Assessment; MoCA) at baseline (≤3 months post‐stroke) and at 1‐year. We assessed change in cognition over 1‐year using a linear mixed model. Predictors were age, sex, stroke severity (National Institutes of Health Stroke Scale; NIHSS), combined VRFs, disability (modified Rankin Score), occurrence of incidental infarct in‐between visits and WMH volume (%intracranial volume). We studied tissue areas of stable normal‐appearing white matter (NAWM), stable WMH, regressing and progressing WMH (Figure1) and measured diffusion‐tensor fractional anisotropy (FA) and mean diffusivity (MD), to assess microstructure in each areaResultAt baseline, mean age = 67.5 (SD = 11.1), 66% male (N = 229), mean MoCA score = 25.0 (SD = 3.5). Between baseline and 1‐year post‐stroke, MoCA score decreased with older age (standardized B[95%CI]: ‐0.271[‐0.383; ‐0.159]) and increasing NIHSS (‐0.193[‐0.287; ‐0.099]). Areas of NAWM that progress to WMH already have lower FA and higher MD at baseline than NAWM that stays NAWM (N = 191; Figure2). WMH that regress to NAWM have higher FA and lower MD than stable WMH. At 1‐year, MD values seem higher and FA values seem lower in progressing WMH than at baseline.ConclusionCognition deteriorates 1 year post‐stroke with older age and higher stroke severity, but not with WMH changes, when considering net WMH volume. Areas of WMH changes exist within individuals, these areas show different underlying tissue structures before becoming apparent on conventional MRI, indicating that WMH regression is real. Separate areas of WMH progression and regression might be more sensitive to cognitive changes and potentially be intervention targets.