Principal Component‐Based Comprehensive Physical Health Score Associates With Long‐Term Death in Patients With Ischemic Stroke

BACKGROUND: Principal component analysis (PCA) integrates multiple clinical indicators into a single score, providing a holistic assessment. Existing clinical indicators often fail to fully reflect health conditions. To address this gap, our study constructs a comprehensive physical health score (CPHS) based on PCA, incorporating 42 clinical and laboratory indicators, and explores its association with long-term death after acute ischemic stroke. METHODS: We collected 42 indicators across 9 categories from 1017 patients with thrombolysis-treated acute ischemic stroke in the training cohort and 316 in the testing cohort. The outcome was 1-year death. CPHS was constructed using PCA and Cox regression in the training cohort and validated in the testing cohort. Hierarchical clustering was performed on the basis of PCA results to define subgroups and assess the predictive value of CPHS. RESULTS: <0.001). To better define the characteristics, hierarchical clustering on the 8 most important principal components identified 5 clusters: healthy, potential kidney dysfunction, liver and biliary dysfunction, metabolic abnormalities with high inflammation, and poor health with high disease risk. Cox regression showed that, compared with cluster 1, the other 4 clusters were significantly associated with 1-year mortality, with mortality risk decreasing from cluster 5 to cluster 1. CONCLUSIONS: CPHS reliably predicted 1-year death after acute ischemic stroke. Additionally, CPHS was able to identify markers of liver and kidney dysfunction, cardiovascular abnormalities, metabolic dysfunction, and high inflammation levels, all of which were associated with increased death after stroke.