Chathuri Yatawara1, Kok Pin Ng1, Russell Chander1, Nagaendran Kandiah2. 1. From the Department of Neurology (C.Y., K.P.N., R.C., N.K.), National Neuroscience Institute; and Duke-NUS (N.K.), Singapore. 2. From the Department of Neurology (C.Y., K.P.N., R.C., N.K.), National Neuroscience Institute; and Duke-NUS (N.K.), Singapore. Nagaendran.Kandiah@singhealth.com.sg.
Abstract
OBJECTIVE: To investigate whether the effect of prestroke and stroke-related lesions on incident poststroke dementia (PSD) is mediated by a unique pattern of domain-specific cognitive impairment, and the relative strength of these anatomical-cognitive associations in predicting incident PSD. METHODS: In this incident case-control study (n = 150), we defined incident cases as acute stroke patients who developed PSD and controls as acute stroke patients who remained free from dementia at a 6 month follow-up, matched on age, prestroke cognitive status, and number of stroke-related lesions. MRI was performed at initial clinical presentation; neuropsychological assessments and clinical diagnosis of PSD was performed 6 months poststroke. Moderated mediation analysis evaluated the interactions among PSD, anatomical lesions, cognitive domains, and individual demographic and medical characteristics. RESULTS: Compared to stroke-related lesions, prestroke lesions were associated with the widest range of cognitive domain impairments and had stronger clinical utility in predicting incident PSD. Specifically, global cortical atrophy (GCA) and deep white matter hyperintensities (WMH) were indirectly associated with PSD by disrupting executive functions, memory, and language. Acute infarcts were indirectly associated with PSD by disrupting executive functions and language. The strongest mediator was executive dysfunction, increasing risk of PSD in patients with deep WMH, GCA, and large infarcts by more than 9 times, with sex and educational attainment moderating the magnitude of association. Periventricular WMH were directly associated with incident PSD but not mediated by deficits in cognitive domains. CONCLUSION: We provide an anatomical-cognitive framework that can be applied to stratify patients at highest risk of PSD and to guide personalized interventions.
OBJECTIVE: To investigate whether the effect of prestroke and stroke-related lesions on incident poststroke dementia (PSD) is mediated by a unique pattern of domain-specific cognitive impairment, and the relative strength of these anatomical-cognitive associations in predicting incident PSD. METHODS: In this incident case-control study (n = 150), we defined incident cases as acute strokepatients who developed PSD and controls as acute strokepatients who remained free from dementia at a 6 month follow-up, matched on age, prestroke cognitive status, and number of stroke-related lesions. MRI was performed at initial clinical presentation; neuropsychological assessments and clinical diagnosis of PSD was performed 6 months poststroke. Moderated mediation analysis evaluated the interactions among PSD, anatomical lesions, cognitive domains, and individual demographic and medical characteristics. RESULTS: Compared to stroke-related lesions, prestroke lesions were associated with the widest range of cognitive domain impairments and had stronger clinical utility in predicting incident PSD. Specifically, global cortical atrophy (GCA) and deep white matter hyperintensities (WMH) were indirectly associated with PSD by disrupting executive functions, memory, and language. Acute infarcts were indirectly associated with PSD by disrupting executive functions and language. The strongest mediator was executive dysfunction, increasing risk of PSD in patients with deep WMH, GCA, and large infarcts by more than 9 times, with sex and educational attainment moderating the magnitude of association. Periventricular WMH were directly associated with incident PSD but not mediated by deficits in cognitive domains. CONCLUSION: We provide an anatomical-cognitive framework that can be applied to stratify patients at highest risk of PSD and to guide personalized interventions.
Authors: Austyn D Roseborough; Berk Rasheed; Youngkyung Jung; Kevin Nishimura; William Pinsky; Kristopher D Langdon; Robert Hammond; Stephen H Pasternak; Ali R Khan; Shawn N Whitehead Journal: Brain Pathol Date: 2021-09-19 Impact factor: 6.508