Abstract
<jats:sec> <jats:title>Background</jats:title> <jats:p>Ischemic stroke is a heterogeneous disease influenced by inflammation, coagulation dysfunction, and metabolic disturbances. However, integrated analysis incorporating these biological domains for patient stratification remain limited.</jats:p> </jats:sec> <jats:sec> <jats:title>Methods</jats:title> <jats:p> A retrospective study of 132 ischemic stroke patients was conducted. Clinical, coagulation, inflammatory, and metabolic parameters were collected. Principal component analysis (PCA) was applied for dimensionality reduction and visualization. <jats:italic>K</jats:italic> -means clustering was then used to identify subtypes with the optimal cluster number validated by elbow plot and silhouette analysis. Differences among cluster’s groups were assessed using ANOVA or Kruskal–Wallis tests for continuous variables and Chi-square tests for categorical variables. </jats:p> </jats:sec> <jats:sec> <jats:title>Results</jats:title> <jats:p> PCA revealed underlying heterogeneity among patients. Validated <jats:italic>K</jats:italic> -means clustering identified three distinct subtypes. Cluster 1 represented a low inflammatory subtype with reduced inflammatory markers. Cluster 2 was a high inflammatory and hypercoagulable subtype, characterized by elevated WBC, NEU, hsCRP, FIB, D-dimer, PT, INR along with a higher prevalence of coronary heart disease and carotid plaque, smoking, and drinking. Cluster 3 was a metabolic risk subtype, characterized by relatively younger age, elevated TG, CHOL, HDL-C, LDL-C, APOB, APOA-1 and APOB/APOA1 ratio, and intermediate inflammatory activity. </jats:p> </jats:sec> <jats:sec> <jats:title>Conclusion</jats:title> <jats:p>Data driven clustering identified biologically distinct ischemic stroke subtypes based on inflammation, coagulation, and metabolic profiles. This stratification highlights the heterogeneity of ischemic stroke and may inform future personalized approaches to risk assessment and management.</jats:p> </jats:sec>