Post-stroke memory impairment among patients with vascular mild cognitive impairment

Six hundred twenty acute ischemic stroke patients from 12 hospitals were enrolled in the study within 7 days of symptom onset from October 2007 to August 2008. Exclusion criteria and the exact number of patients for each stage of the study are shown in Figure 1. A detailed description of the methodology has been published elsewhere [17]. Briefly, we conducted a baseline evaluation within two weeks of stroke onset to determine demographic characteristics, vascular risk factors, stroke subtype [18], and functional status. Among these 620 patients, 353 completed the 60-min K-VCIHS-NP evaluation at three months post-stroke.

The 60-min K-VCIHS-NP includes 7 tests that assess 4 cognitive domains: the Hopkins Verbal Learning Test for memory; Semantic (animal) and phonemic fluency tests, the Digit Symbol Coding, and the Trail Making Test for executive function/activation; the Rey Complex Figure Test-Copy for visuospatial skills; and the Boston Naming Test for language [19].

We also administered the Korean version of the Informant Questionnaire on Cognitive Decline in the Elderly (K-IQCODE) to examine pre-morbid history of cognitive impairment; the Instrumental Activities of Daily Living scale (IADL) to assess instrumental complex daily functioning; and the Korean version of the Mini-Mental Status Examination (K-MMSE) and Clinical Dementia Rating (CDR) as supplementary tests of cognitive function at 3 months post-stroke. All tests and scales were previously validated for the Korean language and standardized in Korean subjects by age and educational levels [20],[21].

The diagnostic procedure included a clinical interview and the administration of the K-VCIHS-NP, K-IQCODE, and IADL. Impaired cognitive function was defined as impairment in at least one cognitive domain on the K-VCIHS-NP, as indicated by a score below the 10th percentile using normative corrections for an individual domain. For executive function/activation domains, which consisted of 4 tests, patients who scored below the 10^th percentile on more than 2 tests were identified as having impaired functions. Abnormal functional status was indicated by a score greater than 0.43 on the Korean IADL [20]. Pre-stroke cognitive impairment was considered present if the K-IQCODE score exceeded 3.6 [21]. A vascular cause was presumed when cognitive impairment emerged at 3 months after the stroke in the absence of pre-stroke impairment. VaMCI was defined as impairment in at least 1 of 4 cognitive domains, combined with normal or mild impairments in daily living activities and temporal evidence of vascular causation [1].

We compared the demographic features and characteristics of index infarcts of VaMCI patients based on the presence of MI, using a Mann-Whiteny test for continuous variables and chi-square tests for categorical variables. Univariate and multivariate logistic analyses examined the influence of baseline demographic features, infarct locations, and classification of stroke subtypes on memory impairment in VaMCI at 3-month after stroke. We entered age, sex, duration of education, all locations and subtypes of ischemic stroke in the logistic model for detailed description. Analyses were performed with SPSS (Windows version 18.0).

This study was approved by the Institutional Review Board/Ethics Committee of each participating hospital (Hallym University Sacred Heart Hospital IRB 2007–137 http://​hallym.​hallym.​or.​kr/​irb/​). All subjects provided informed consents. This study followed Good Clinical Practice guidelines and was consistent with the International Conference on Harmonization of ethical principles for medical research involving human subjects.