Predictors for cognitive impairment and dementia after stroke

Filler, Jule (2025): Predictors for cognitive impairment and dementia after stroke. Dissertation, LMU München: Graduate School of Systemic Neurosciences (GSN)

Abstract

Stroke is one of the leading causes of death and disability worldwide. A large proportion of stroke survivors suffer from subsequent cognitive decline, with dementia representing its most severe form. In the absence of widely available disease-modifying therapies, it is essential to understand the risk factors that predispose patients to post-stroke cognitive impairment (PSCI) and dementia (PSD) to improve individual risk prediction and to develop preventive strategies. However, these risk factors remain insufficiently understood, especially over the long term. Existing observational and clinical studies are constrained by methodological heterogeneity, short follow-up periods, and a limited focus on milder outcomes, such as mild cognitive impairment (MCI), which is included in PSCI. While factors such as older age and greater stroke severity have consistently been linked to PSCI and PSD, the role of other potential predictors is less certain. Moreover, robust evidence for modifiable risk factors remains scarce. Thus, the overarching aim of this thesis was to improve the understanding of the associations between factors present before or at the time of stroke with PSCI and PSD, thereby aiding tailored risk stratification and secondary prevention efforts. Given the heterogeneity of previous studies, the first study synthesized existing evidence in a systematic review and meta-analysis. Eligible studies were required to use validated cognitive assessment tools, have a longitudinal design, and adjust effect estimates for the well-established risk factors age and stroke severity. Pooling data from 89 studies including more than 160,000 stroke patients, we found that acute-phase cognitive impairment, assessed with brief screening tools shortly after stroke, was the strongest predictor of both PSCI and PSD. Among modifiable risk factors, diabetes mellitus and atrial fibrillation (AF) were the most consistent cardiovascular risk factors, although the role of AF remained less clear for the PSD endpoint. Additional robust predictors included lower educational attainment, prior stroke, markers of cerebral small vessel disease (SVD) – particularly white matter hyperintensities and lacunes – cerebral atrophy, left- hemispheric lesions, and acute-phase functional deficits. Notably, the strength of associations for some risk factors, such as stroke severity and AF, appeared weaker in more recent studies, possibly reflecting improvements in acute stroke care and vascular risk factor management. Second, building on these findings, I analyzed longitudinal data from the prospective multicenter DEMDAS (DZNE mechanisms of dementia after stroke) study, which enrolled 736 patients (33% female, mean age 68 years) at the time of stroke. Over a median follow-up of 5.0 years (2899 person- years), 55 patients developed dementia. We identified metabolic syndrome (MetS) – particularly its components hyperglycemia (prediabetes or diabetes) and reduced high-density lipoprotein cholesterol (HDL-C) – as a novel risk factor for PSD and a potential target for secondary prevention. Consistent with the first study, additional key predictors included older age, greater stroke severity, lower educational attainment, acute-phase cognitive and functional impairment, AF, and imaging markers of SVD. PSD risk was also elevated in patients with recurrent stroke, whereas acute reperfusion therapy was associated with a substantially reduced risk. Risk factor profiles for PSCI largely mirrored those for PSD. Importantly, our analyses revealed distinct risk profiles for early-onset PSD (dementia diagnosed within 3–6 months post-stroke) versus delayed-onset PSD (>6 months): early-onset PSD was more strongly linked to factors related to the acute stroke and pre-stroke brain health, while MetS and stroke recurrence were more prominent predictors for delayed-onset PSD. The association between cardiometabolic risk factors and PSD persisted even after adjusting for recurrent stroke, suggesting additional underlying mechanisms beyond recurrent cerebrovascular events. In conclusion, this thesis strengthens the evidence for the multifactorial nature of PSCI and PSD by combining a systematic synthesis of prior studies with longitudinal data from a deeply phenotyped stroke cohort. The most important risk factor categories identified in this thesis include i) pre-stroke brain health and reserve, ii) acute stroke-related severity and deficits, and iii) cardiovascular and metabolic risk factors. MetS and reduced HDL-C emerged as novel cardiometabolic risk factors and potential targets for secondary prevention. These findings can inform individualized risk prediction and guide the design of clinical trials aimed at preventing cognitive decline after stroke. More broadly, they underscore the need for future observational and interventional studies to prioritize modifiable risk factors and to establish cognitive endpoints as central outcomes in secondary prevention trials.

• Dokumententyp: Dissertationen (Dissertation, LMU München)
• Keywords: Neuroscience, Neuroepidemiology, Stroke, Cerebrovascular Disease, Dementia, Vascular Dementia, Cognitive Disorders, Risk Factors
• Fakultäten: Graduate School of Systemic Neurosciences (GSN)
• Sprache der Hochschulschrift: Englisch
• Datum der mündlichen Prüfung: 4. Dezember 2025
• 1. Bericht­erstatter:in: Dichgans, Martin