Study participants

Patients, aged ≥ 50 years with first-ever ischemic stroke, who were admitted to Korea University Guro Hospital within 7 days of ischemic stroke were screened for enrollment eligibility from June 1, 2021, to April 1, 2022. At baseline, all patients with ischemic stroke underwent a comprehensive stroke evaluation, including neurological examination using the National Institutes of Health Stroke Scale (NIHSS), functional outcome using the modified Rankin scale, and brain magnetic resonance imaging (MRI).

We prospectively enrolled patients with small subcortical infarctions to eliminate the effects of the stroke lesion size and strategic site. We excluded patients with the following conditions: (1) strategic infarcts involving the anterior thalamus and hippocampus; (2) severe aphasia (NIHSS language score > 1), visual impairment, or physical disabilities (Modified Rankin Scale score > 2) due to ischemic stroke; and (3) presence of premorbid cognitive impairment, neurodegenerative diseases, lobar hemorrhage, or psychiatric disorders.

Premorbid cognitive impairment was determined using the Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE) [6], answered by the participants’ spouses or a first-degree relative. A score of ≥ 3.6 on the IQCODE indicated premorbid cognitive impairment, and the participants were excluded [6].

Hypertension was defined as a diagnostic history of hypertension or current use of any antihypertensive medication, and diabetes was defined as a diagnostic history of diabetes or current use of any anti-diabetic medication.

Cognitively unimpaired (CU) participants without stroke

We also enrolled CU participants matched for age and sex with stroke patients. These participants did not have a history of stroke, neurodegenerative disease, or psychiatric disorders and were composed of spouses of patients who visited the neurology clinic, volunteers who applied for comprehensive dementia evaluation advertised in the paper, and participants who had cognitive complaints. They visited the Memory Clinic in the Department of Neurology at Korea University Guro Hospital and underwent a comprehensive dementia evaluation.

All CU participants met the following criteria: (1) no medical history which is likely to affect cognitive function based on Christensen’s health screening criteria [7], (2) no objective cognitive impairment from comprehensive neuropsychological test battery on any cognitive domains (no cognitive test fell more than 1.0 standard deviation [SD] below age-adjusted norms), (3) independent in activities of daily living, and (4) neither structural lesions nor severe white matter hyperintensities (WMH) on brain MRI.

Hypertension was defined as a diagnostic history of hypertension or current use of any antihypertensive medication, and diabetes was defined as a diagnostic history of diabetes or current use of any anti-diabetic medication.

This study was approved by the Institutional Review Board of the Korea University Guro Hospital. Written informed consent was obtained from all participants.

Follow-up assessment

All patients with stroke were followed up at 3 and 12 months after ischemic stroke. At the follow-up visit at 3 months, the patients underwent a neuropsychological battery using the Korean version of the Vascular Cognitive Impairment Harmonization Standards neuropsychological battery (K-VCIHS-NP) [8], Korean version of the Mini-Mental Status Examination (K-MMSE) [9], Clinical Dementia Rating-Sum of Box (CDR-SOB) [10], amyloid PET, and a second MRI. At the follow-up visit at 12 months, patients underwent the K-MMSE and CDR-SOB to evaluate the trajectory of global cognition.

CU participants without stroke were also followed up at 12 months after the initial comprehensive evaluation. At the follow-up visit at 12 months, CU participants also underwent the K-MMSE and CDR-SOB.

Comprehensive neuropsychological battery

Patients underwent neuropsychological testing using the K-VCIHS-NP [8]. Seven cognitive measures were included in the battery, which were representative and important neuropsychological tests to evaluate the cognitive function in five cognitive domains as follows: (1) memory: the Seoul Verbal Learning Test (SVLT) delayed recall (verbal memory); (2) language: Korean version of the Boston Naming Test (K-BNT); (3) visuospatial function: the Rey Complex Figure Test (RCFT) Copying Test; and (4) frontal-executive function: the Digit Span Test Backward, animal component of the Controlled Oral Word Association Test (COWAT), and phonemic component of the COWAT and the Stroop Test (color reading). Results with continuous numeric values were converted to z-scores using the age, sex, and education criteria presented in the K-VCIHS-NP, and the z-scores were used in the analysis.

Diagnosis of vascular cognitive impairment and vascular dementia

PSCI was defined according to the modified Peterson criteria and results of the K-VCIHS-NP. In the K-VCIHS-NP, cognitive functions were classified as impaired when objective cognitive impairment was − 1.5 SD on at least two different cognitive domains. The frontal domain consists of four neuropsychological tests. Cognitive impairment in the frontal domain was classified as objective cognitive impairment − 1.5 SD on two or more tests.

Amyloid PET acquisition and visual reading

The patients underwent 18F-florbetaben PET using a discovery MI PET/computed tomography (CT) scanner (GE Medical Systems, Milwaukee, WI, USA). A 20-min emission PET scan in dynamic mode (comprising 4 × 5 min frames) was performed 90 min after the injection of a mean dose of 296 MBq 18F-florbetaben. Three-dimensional PET images were reconstructed in a 384 × 384 matrix with 0.65 × 0.65 × 2·79 mm voxel size using the ordered-subsets expectation maximization algorithm (iteration = 8 and subset = 34).

Amyloid PET images were reviewed by three experienced physicians (one neurologist and two nuclear medicine doctors) who were blinded to clinical information and dichotomized as either Aβ positive or negative using visual reads [11]. 18F-florbetaben PET was classified as positive when interpreters scored the visual assessment as 2 or 3 on the brain amyloid-plaque load (BAPL) score [11, 12]. Specifically, the regional cortical tracer uptake (RCTU) score was used for four brain areas (lateral temporal cortex, frontal cortex, posterior cingulate, cortex/precuneus, and parietal cortex). RCTU scores of 1, 2, and 3 indicated no tracer uptake, moderate tracer uptake, and pronounced tracer uptake, respectively.

An RCTU score of 1 in each brain region corresponded to a BAPL score of 1, an RCTU score of 2 in any brain region and no score of 3 corresponded to a BAPL score of 2, and an RCTU score of 3 in any of the four brain regions corresponded to a BAPL score of 3. Inter-rater agreement was excellent (Fleiss, k = 0.89). After the physicians individually rated, we determined the final Aβ positivity based on the majority of the visual reading result.

MRI acquisition and WMH visual rating

We acquired standardized three-dimensional T1 Turbo Field Echo and three-dimensional fluid-attenuated inversion recovery (FLAIR) images using a 3.0-T MRI scanner (Philips 3.0T Ingenia Elition X; Philips Healthcare, Andover, MA, USA) by following imaging parameters: sagittal slice thickness, 0.6 mm; no gap; TR, 4800 ms; TE, 363 ms; flip angle, 90°; and matrix size, 288 × 287 pixels. As described previously [13], the Clinical Research Center for Dementia of South Korea WHM visual rating scale was used to investigate WMH in the deep subcortical and periventricular regions of the FLAIR images.

Briefly, deep WMH (DWMH) were classified as D1 (< 10 mm), D2 (10–25 mm), or D3 (≥ 25 mm) based on the longest lesion diameter. Periventricular WMH (PWMH) were classified as P1 (cap and band < 5 mm), P2 (5–10 mm), or P3 (cap or band ≥ 10 mm) based on the maximum length measured perpendicular (cap) and horizontal (band) to the ventricle. The combination of these D and P ratings yielded nine cells and the overall WMH severity (minimal, moderate, and severe) was defined based on the following combinations of D and P ratings: minimal (D1P1 and D1P2), moderate (D1P3, D2P1, D2P2, D2P3, D3P1, and D3P2), and severe (D3P3).

We also counted the number of microbleeds (MBs), defined as ≤ 10 mm in diameter on 150 axial slices of T2 susceptibility-weighted imaging-MRI by following imaging parameters: sagittal slice thickness, 2.0 mm; no gap; TR, 24 ms; TE, 0 ms; flip angle, 18°; and matrix size, 384 × 383 pixels [14]. Strictly lobar MBs (number of lobar MBs ≥ 1 and deep MBs = 0) and cerebral superficial siderosis (cSS) were considered as cerebral amyloid angiopathy (CAA) markers. Regarding lobar MBs, the lobar regions were according to the criteria proposed by Gregoire et al. [15]. cSS was defined as chronic blood linear residues in superficial layers of the cerebral cortex [16].

Statistical analyses

For comparison between the clinical characteristics of stroke and CU without stroke groups, independent t-tests and chi-squared tests were used. To assess the effect of stroke on longitudinal cognitive changes, we performed a linear mixed-effect regression analysis and included stroke group (stroke vs CU without stroke), time, and stroke group × time as fixed effects, along with age, sex, years of education, hypertension, diabetes, Aβ positivity, and WMH severity. The patients were included as random effects. To identify the association of stroke with cognitive function at 3 months and 12 months, we performed linear regression analyses with the stroke group as a predictor after controlling for with age, sex, years of education, hypertension, diabetes, Aβ positivity, and WMH severity.

In stroke groups, independent t-tests and chi-squared tests were used to compare the clinical characteristics of the post-stroke normal cognition (PSNC) and PSCI groups. To investigate the association between Aβ deposition and the development of PSCI, we performed a logistic regression analysis with Aβ positivity, hypertension, diabetes, and WMH severity as predictors and PSCI development as an outcome after controlling for age, sex, and years of education. To assess the effect of Aβ positivity on longitudinal cognitive changes, we performed a linear mixed-effect regression analysis and included Aβ positivity, time, and Aβ positivity × time as fixed effects, along with age, sex, years of education, hypertension, diabetes, and WMH severity. The patients were included as random effects. To identify the association of Aβ positivity with cognitive function at 3 months and 12 months, we performed linear regression analyses with Aβ positivity as a predictor after controlling for with age, sex, years of education, hypertension, diabetes, and WMH severity.

All reported p-values were two-sided, and the significance level was set at 0.05. All analyses were performed using R version 4.3.0 (Institute for Statistics and Mathematics, Vienna, Austria; http://www.r-project.org, RRID: SCR_001905).