Post-stroke Cognitive Impairment-Impact of Follow-Up Time and Stroke Subtype on Severity and Cognitive Profile: The Nor-COAST Study

Abstract

Background: Post-stroke cognitive impairment (PSCI) is common, but evidence of cognitive symptom profiles, course over time, and pathogenesis is scarce. We investigated the significance of time and etiologic stroke subtype for the probability of PSCI, severity, and cognitive profile. Methods: Stroke survivors (n = 617) underwent cognitive assessments of attention, executive function, memory, language, perceptual-motor function, and the Montreal Cognitive Assessment (MoCA) after 3 and/or 18 months. PSCI was classified according to DSM-5 criteria. Stroke severity was assessed with the National Institutes of Health Stroke Scale (NIHSS). Stroke subtype was categorized as intracerebral hemorrhage (ICH), large artery disease (LAD), cardioembolic stroke (CE), small vessel disease (SVD), or un-/other determined strokes (UD). Mixed-effects logistic or linear regression was applied with PSCI, MoCA, and z-scores of the cognitive domains as dependent variables. Independent variables were time as well as stroke subtype, time, and interaction between these. The analyses were adjusted for age, education, and sex. The effects of time and stroke subtype were analyzed by likelihood ratio tests (LR). Results: Mean age was 72 years (SD 12), 42% were females, and mean NIHSS score at admittance was 3.8 (SD 4.8). Probability (95% CI) for PSCI after 3 and 18 months was 0.59 (0.51-0.66) and 0.51 (0.52-0.60), respectively and remained constant over time. Global measures and most cognitive domains were assessed as impaired for the entire stroke population and for most stroke subtypes. Executive function and language improved for the entire stroke population (LR) = 9.05, p = 0.003, and LR = 10.38, p = 0.001, respectively). After dividing the sample according to stroke subtypes, language improved for ICH patients (LR = 18.02, p = 0.003). No significant differences were found in the severity of impairment between stroke subtypes except for attention, which was impaired for LAD and CE in contrast to no impairment for SVD (LR = 56.58, p < 0.001). Conclusions: In this study including mainly minor strokes, PSCI is common for all subtypes, both early and long-term after stroke, while executive function and language improve over time. The findings might contribute to personalizing follow-up and offer new insights into underlying mechanisms. Further research is needed on underlying mechanisms, PSCI prevention and treatment, and relevance for rehabilitation.

Keywords: cerebrovascular disease; cognitive domains; intracerebral hemorrhag; post-stroke cognitive impairment; prognosis; stroke; stroke subtype; vascular dementia.

Figure 1

Flowchart of participants included in the study.

Figure 2

(A–H) Probability for cognitive impairment according to DSM-5 criteria with 95% CI and mean z-score with 95% CI for the cognitive domains at 3 and 18 months post-stroke in model 1. MoCA, Montreal Cognitive Assessment. †Adjusted for age, education, and sex. ^### LR χ²(1), Likelihood Ratio test model 1 vs a model with only age, education, and sex as confounders, with one degree of freedom; hypothesis test of whether there is an effect of time. ^‡p < 0.01.

Figure 3

(A–H) Probability for cognitive impairment according to DSM-5 criteria with 95% CI and mean z-score with 95% CI for the cognitive domains at 3- and 18-months post-stroke in model 2. MoCA, Montreal Cognitive Assessment. † Adjusted for age, education, and sex. ^#LAD, Large artery disease. **CE, Cardiac emboli. ††SVD, Small vessel disease. ^§§UD, Undetermined and other determined strokes. ^¶¶ICH, Intracerebral hemorrhage. ^###LR χ²(8), Likelihood Ratio test model 1 vs. model 2 with 8 degrees of freedom; hypothesis test of whether there is an effect of stroke subtype. ***LR χ²(5), Likelihood Ratio test model 2 vs. model 3 with 5 degrees of freedom; hypothesis test of whether there is an effect of time for at least one stroke subtype. ^‡p < 0.01.