Association between PCSK9 and post-stroke cognitive impairment: a retrospective cohort study and predictive models establishment based on machine learning
Related Experiment Videos
Ling Wang1, Xin Zhang2, Xinyu Fan1
1Department of Neurology, the First Affiliated Hospital (Yijishan Hospital) of Wannan Medical College, Wuhu City, Anhui Province, China.
Background And Purpose:
Proprotein convertase subtilisin/kexin type 9 (PCSK9) has gained increasing attention due to its involvement in lipid metabolism and neuroinflammation regulation. High PCSK9 levels are linked to an elevated risk of cerebrovascular events and small vessel diseases. However, systematic research exploring the association between PCSK9 and post-stroke cognitive impairment (PSCI) remains scarce. This study investigated the link between PCSK9 and PSCI.
Methods:
A cohort of 354 patients with PSCI was enrolled in this investigation. The link between PCSK9 and PSCI occurrence was evaluated through multivariate logistic regression analysis. Restricted cubic spline methodology was employed to examine the non-linear association between PCSK9 levels and PSCI risk. Prediction models were developed using machine learning algorithms.
Results:
After adjusting for confounders, elevated PCSK9 levels were identified as an independent predictor of PSCI. Including PCSK9 in prediction models alongside traditional risk factors significantly improved the accuracy of PSCI predictions. Feature importance rankings from the SVM model highlighted the significance of PCSK9, with the SVM-based model yielding the highest performance (AUC = 0.895).
Conclusion:
PCSK9 blood levels in individuals after acute ischemic stroke function as an independent predictor for PSCI and improve the precision of machine learning-based predictive models.