Change in Cognition Following Ischaemic Stroke.

Objective: Cognitive decline can occur following ischaemic stroke. How cognition changes over time and associations with cognitive change are poorly understood. This study aimed to explore these issues over 2 years following ischaemic stroke.

Methods: This analysis used data from the XILO-FIST study, a clinical trial of allopurinol versus placebo in people with ischaemic stroke according to Tissue-Based Definition. Participants underwent clinical assessment, brain MRI at baseline, and Montreal Cognitive Assessment (MoCA) at baseline, year 1 and year 2. We defined cognitive impairment as a MoCA score < 26 and cognitive change as a difference in MoCA score of 2 points or more at year 1 or year 2 after randomisation. Associations with cognitive impairment and cognitive change were assessed by univariable analysis and multiple logistic regression.

Results: Three hundred and sixty participants with complete MoCA data were included. Mean age was 65.4 (SD 8.36) years, and mean baseline MoCA score was 26.4 (SD 2.7). Seventy-seven participants had second-year cognitive improvement. Eighty-four had second-year cognitive decline. After adjustment for age and education year, second-year cognitive improvement was associated with smaller brain volume, lower albumin level, smoking and greater white-matter hyperintensity, and second-year cognitive decline was associated with peripheral arterial disease, higher cholesterol level, small-vessel stroke and greater white-matter hyperintensity.

Interpretation: Cognition is dynamic following stroke, with different patterns of change. Brain reserve and vascular risk factors relate to later post-stroke cognitive change. This complex nature of cognitive trajectory has implications for cognitive rehabilitation provision and cognitive impairment detection after stroke.