Association of estimated pulse wave velocity with all-cause mortality in critically ill patients with atrial fibrillation: a retrospective study from MIMIC IV3.0

Estimated pulse wave velocity (ePWV) is a reliable method for evaluating arterial stiffness. While the prognostic significance of ePWV has been examined in a range of cardiovascular diseases, its particular importance in atrial fibrillation (AF) has not been thoroughly investigated. This study seeks to explore the relationship between ePWV and all-cause mortality (ACM) in individuals with AF. Patient information was collected from the Medical Information Mart for Intensive Care (MIMIC)-IV database, resulting in the selection of 14,704 individuals with AF for this study. The dominant end measure in this research was ACM at 30 days and 1 year, while the secondary end measure was ACM at 60 and 90 days post-admission to the ICU. Patients were divided into quartiles based on ePWV levels for analysis, and Kaplan-Meier curves were used to compare outcomes across groups. To explore the association between ePWV and patient prognosis, Cox proportional hazards models and restricted cubic splines (RCS) were employed. Additionally, subgroup and sensitivity analyses were conducted to verify the robustness of the results. The study included 14,704 participants, 8,816 (59.96%) of whom identified as male. Survival analysis demonstrated that elevated ePWV levels were associated with an increased risk of 30-day, 60-day, and 90-day ACM, as well as 1-year ACM. Cox proportional hazards regression analysis showed that, compared with the quartile Q1, the quartile Q2 (HR, 1.20; 95% CI: 1.06-1.37), Q3 (HR, 1.59; 95% CI: 1.40-1.81), and Q4 (HR, 2.38; 95% CI: 2.09-2.71) groups had a significantly increased risk of 30-day ACM. Similar results were observed for 1-year ACM and secondary outcomes. Additionally, RCS analysis revealed a J-shaped relationship between ePWV and ACM, with an inflection point of 9.8 m/s for both 30-day and 1-year ACM. When ePWV was ≥ 9.8 m/s, each 1-unit increase in ePWV was associated with a 16% increase in the risk of 30-day ACM (HR = 1.16, 95% CI = 1.14-1.19, p < 0.001) and a 16% increase in the risk of 1-year ACM (HR = 1.16, 95% CI = 1.14-1.18, p < 0.001). However, this association was not observed when ePWV levels were below 9.8 m/s. Sensitivity analyses confirmed the robustness of the results. The ePWV and SOFA composite score models both outperform their respective standalone models in predicting mortality. Elevated ePWV is linked to increased short- and long-term ACM in critically ill AF patients. ePWV may act as an important indicator for evaluating the severity of AF patients in the ICU and guiding therapeutic decisions.