Estimated pulse wave velocity and incident cognitive impairment risk in stroke patients: a longitudinal cohort study

Arterial stiffness is a significant risk factor for cardiovascular and cerebrovascular diseases. Estimated pulse wave velocity (ePWV), as a surrogate marker of arterial stiffness, may be associated with cognitive decline through vascular pathways. Yet studies examining the relationship between ePWV and cognitive impairment in stroke patients remain limited. Using data from the China Health and Retirement Longitudinal Study (CHARLS) collected between 2011 and 2020, we included 2,003 stroke patients. ePWV was calculated using age and blood pressure, and cognitive impairment was assessed using the Mini-Mental State Examination (MMSE) and word recall tests. Cox proportional hazards models were used to analyze the association between ePWV and cognitive impairment risk, with nonlinear relationships evaluated through restricted cubic spline analysis. Model performance was assessed using receiver operating characteristic curves, decision curve analysis, and calibration plots, along with subgroup analyses. After adjusting for confounders, each standard deviation increase in ePWV was associated with a 26.3% increased risk of cognitive impairment (HR = 1.263, 95% CI: 1.154–1.383). Restricted cubic spline analysis revealed a significant nonlinear association (P nonlinear = 0.010). The highest ePWV quartile (Q4: 12.1–17.7 m/s) showed significantly elevated risk compared to the lowest quartile (Q1: 5.7–9.3 m/s) (HR = 2.828, 95% CI: 1.909–4.189), demonstrating a clear dose-response relationship. The predictive model showed good discrimination (AUC = 0.812) and positive net benefit at threshold probabilities 0.1–0.8. Subgroup analyses revealed stronger associations in males (HR = 1.333, 95% CI: 1.168–1.522) and patients with diabetes (HR = 1.426, 95% CI: 1.038–1.959). ePWV demonstrates a significant nonlinear positive association with cognitive impairment risk in stroke patients, suggesting its potential as a marker for cognitive impairment risk assessment and offering new perspectives for early prediction and intervention.