83PRE-STROKE FRAILTY IS INDEPENDENTLY ASSOCIATED WITH ACUTE POST-STROKE COGNITION: A CROSS-SECTIONAL STUDY

Abstract

Introduction: Post-stroke cognitive impairment is common but mechanisms and risk factors are poorly understood. Frailty is a condition that may be highly prevalent within the stroke population and could conceivably affect post-stroke cognitive status. We investigated the association between pre-stroke frailty and post-stroke cognition. Method: The Glasgow Royal Infirmary hospital database was utilised. Consecutively admitted acute stroke patients between February 2016 and December 2017 underwent cognitive assessment via the Mini-Montreal Cognitive Assessment, and clinical and demographic information was collected. The Rockwood Frailty Index was employed to generate a frailty score for each patient. Univariate and Multiple linear regression analyses were conducted to investigate the association between pre-stroke frailty and post-stroke cognition. Age, sex, pre-stroke cognitive disorder, delirium & stroke type were included as covariates. Pre-stroke cognition was assessed via medical records (prior history of mild cognitive impairment or dementia), the General Practitioner Assessment of Cognition informant interview, or via a clinical interview using the Clinical Dementia Rating scale; delirium was assessed via the 4 A Test; stroke type was defined according to Bamford classification and was established based upon the clinical impression of the treating stroke consultant. Results: Of 580 patients assessed, sufficient data was available for analysis of 155 patients. Mean age of patients was 68; 60% were male. Frailty prevalence based on Rockwood Frailty index scores of 0.24+ was 38.7%. Pre-stroke cognitive disorder was apparent in 8.4% of patients, and 8.4% of patients screened positive for delirium. Presenting stroke types were 5.3% TACS, 34.8% PACS, 32.2% LACS, 23.2% POCS and 4.5% Unknown. Pre-stroke frailty was significantly associated with post- stroke cognition (p < 0.001) and this effect was independent of covariates (p = 0.023). As pre-stroke frailty scores increased, post-stroke cognition scores declined (B = -0.172). The full model explained a significant amount of variance (p < 0.001; R2 = 0.393). Additional significant variables in the multiple regression model were age (p = 0.001), delirium (p < 0.001), pre-stroke cognitive disorder (p = 0.002). Stroke-type (p = 0.062) and sex (p = 0.546) were not significant. Conclusion: Pre-stroke frailty may be a highly prevalent and important moderator of post-stroke cognition, independent of other well established post-stroke cognitive impairment risk factors. On this basis, recognition of frailty in the stroke population may be important regarding patients' post-stroke cognitive outcomes.