PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
May 8, 2026European Stroke Journal0 citations

Abstract Number: Esoc2026a2135 Identification of Ischemic Stroke Subtypes by Cluster Analysis and Their Long-Term Vascular Event Risk

View Full Paper
MSMine SezginOAOzan AlpEBElif Begik

Key Points

  • The aim was to classify ischemic stroke patients into distinct clusters based on clinical characteristics and long-term outcomes.
  • Patients diagnosed with ischemic stroke between 2012 and 2018 were analyzed.
  • Hierarchical cluster analysis was employed to identify groups using Gower’s distance.
  • Incidences of stroke recurrence, mortality, and hospital admissions were compared across clusters.
  • Three clusters were identified based on clinical and laboratory variables.
  • Cluster A had younger patients with higher hematocrit and LDL levels, Cluster B included older patients with higher NIHSS scores and more frequent atrial fibrillation.
  • Deaths were observed only in Cluster B, while recurrent stroke rates were similar across all clusters.

Abstract

Abstract Background and aims Ischemic stroke (IS) is a heterogeneous disease with substantially varying etiologies and outcomes, and long-term subtype-specific follow-up is often lacking. We aimed to determine whether patients with IS can be classified into distinct clusters with differing clinical characteristics and long-term outcomes. Methods Demographic and clinical characteristics of patients with IS diagnosed between 2012 and 2018 were recorded. Hierarchical cluster analysis (HCA) was used to identify the optimal number of clusters, applying Gower’s distance to generate distinct patient groups. Clusters were compared with respect to clinical characteristics. Stroke recurrence, mortality, and hospital admissions were recorded and analyzed across clusters using survival analysis to evaluate differences in real-world clinical outcomes. Results Of 1,276 IS patients, 799 with ≥3 months of follow-up were included. According to HCA, three clusters were identified as the optimal solution. Using prespecified clinical and laboratory variables, Gower’s distance was calculated, and three distinct clusters were formed (Figure-1A). Cluster A comprised younger patients (55±14 years) with higher hematocrit and LDL levels, whereas Cluster B included older patients (70±14 years) with higher NIHSS scores. Atrial fibrillation and hemorrhagic transformation were most frequent in Cluster B. Cluster C (65±12 years) had the highest burden of cardiovascular comorbidities, including hypertension, coronary artery disease and diabetes. The median follow-up duration was 26 months. Recurrent stroke rates were similar across clusters (Figure-1B); however, deaths were observed only in Cluster B. Conclusions Cluster analysis may identify clinically meaningful stroke subtypes, and future studies incorporating richer datasets could inform subtype-specific clinical management. Conflict of interest Mine Sezgin : Nothing to disclose, Muhammed Yusuf Cansever : Nothing to disclose, Elif Begik: Nothing to disclose, Cemal Kerem Karaduman: Nothing to disclose, Mehmet Sait Bingöl: Nothing to disclose, Ozan Alp: Nothing to disclose, Esme Ekizoğlu: Nothing to disclose, Nilüfer Yeşilot: Nothing to disclose, Mehmet Güven Günver: Nothing to disclose Figure 1 - belongs to Conclusions

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Sezgin et al. (2026) studied this question.

synapsesocial.com/papers/69fd7f4fbfa21ec5bbf07c85https://doi.org/10.1093/esj/aakag023.1737
Ask AI
Helpful
Bookmark
Share
View Full Paper