Objectives Based on data from the National Health and Nutrition Examination Survey (NHANES), this work aimed to probe into the linkage between estimated pulse wave velocity (ePWV) and the risk and mortality of chronic kidney disease (CKD). Methods Data from five cycles (2009–2018) of the NHANES database were analyzed, with ePWV as the independent variable and CKD as the dependent variable. Weighted logistic regression analysis was undertaken. First, a stratified analysis of unadjusted confounding variables was conducted, and the P-value of the interaction term after adjusting for all confounders was evaluated using the chi-square test. Subsequently, ePWV was stratified into quartiles, and regression models of ePWV and CKD were created with adjustments for different confounding factors. The association threshold between ePWV and CKD was studied using restricted cubic spline (RCS) and threshold effect analysis. Kaplan–Meier (K–M) curves were plotted to study the relationship between ePWV and all-cause mortality and mortality induced by cardiovascular disease (CVD) in CKD patients. The association between ePWV and mortality risk was assessed using multivariable-adjusted Cox regression and Fine-Gray competing risk models, with model predictive performance compared via Harrell’s C-index. Triple sensitivity analyses were conducted to verify the robustness of the results. Results This study included a total of 22,438 participants. Stratified analyses demonstrated that elevated ePWV was significantly associated with an increased risk of CKD across all populations (P < 0.001). Interaction terms adjusted for all covariates indicated that gender, BMI, and T2D significantly influenced this association (P < 0.05). Elevated ePWV was consistently associated with a higher CKD risk in the unadjusted model (Crude), partially adjusted model (Model I), and fully adjusted model (Model II) (OR > 1, P < 0.05). RCS and threshold effect analyses revealed a J-shaped relationship between ePWV and CKD. When ePWV exceeded 6.31 m/s, each increment in ePWV was significantly associated with an increased CKD risk (OR = 1.38, 95% CI 1.32–1.44, P < 0.001). K-M survival curves demonstrated that over time, CKD patients in the highest ePWV quartile had significantly higher all-cause and CVD mortality rates compared to those in the lowest quartile (P < 0.05). Multivariable-adjusted Cox regression showed that, compared to the Q1 group, the Q3 (OR = 2.53, 95% CI 1.55–4.15) and Q4 (OR = 4.12, 95% CI 2.38–7.13) groups had significantly elevated all-cause mortality risk (P < 0.001). The competing risk model further confirmed that Q4 was associated with a significantly higher CVD mortality risk than Q1 (OR = 2.71, 95% CI 1.24–5.95, P = 0.013). Harrell’s C-index indicated that ePWV provided statistically significant incremental predictive value for all-cause mortality risk (Δ = + 0.02, P < 0.001). Triple sensitivity analyses consistently validated the robust association between ePWV and CKD (P < 0.001). Conclusion The ePWV is an independent risk factor for CKD, exhibiting a non-linear relationship with CKD risk. Gender, BMI, and T2D have a substantial impact on this association. Furthermore, higher ePWV levels were associated with increased mortality risk in CKD patients. In the future, the research on how to apply ePWV more effectively in clinical practice to identify the risk of CKD should be advanced, especially for populations with ePWV > 6.68 m/s.

Analyzing the association of estimated pulse wave velocity with the risk and mortality of chronic kidney disease: a study based on the NHANES database

期刊论文
发表日期：

2025年7月14日

查看全文

Hua Zhang,
Shiyi Pan,
Ding Li

摘要

Objectives

Based on data from the National Health and Nutrition Examination Survey (NHANES), this work aimed to probe into the linkage between estimated pulse wave velocity (ePWV) and the risk and mortality of chronic kidney disease (CKD).

Methods

Data from five cycles (2009–2018) of the NHANES database were analyzed, with ePWV as the independent variable and CKD as the dependent variable. Weighted logistic regression analysis was undertaken. First, a stratified analysis of unadjusted confounding variables was conducted, and the P-value of the interaction term after adjusting for all confounders was evaluated using the chi-square test. Subsequently, ePWV was stratified into quartiles, and regression models of ePWV and CKD were created with adjustments for different confounding factors. The association threshold between ePWV and CKD was studied using restricted cubic spline (RCS) and threshold effect analysis. Kaplan–Meier (K–M) curves were plotted to study the relationship between ePWV and all-cause mortality and mortality induced by cardiovascular disease (CVD) in CKD patients. The association between ePWV and mortality risk was assessed using multivariable-adjusted Cox regression and Fine-Gray competing risk models, with model predictive performance compared via Harrell’s C-index. Triple sensitivity analyses were conducted to verify the robustness of the results.

Results

This study included a total of 22,438 participants. Stratified analyses demonstrated that elevated ePWV was significantly associated with an increased risk of CKD across all populations (P < 0.001). Interaction terms adjusted for all covariates indicated that gender, BMI, and T2D significantly influenced this association (P < 0.05). Elevated ePWV was consistently associated with a higher CKD risk in the unadjusted model (Crude), partially adjusted model (Model I), and fully adjusted model (Model II) (OR > 1, P < 0.05). RCS and threshold effect analyses revealed a J-shaped relationship between ePWV and CKD. When ePWV exceeded 6.31 m/s, each increment in ePWV was significantly associated with an increased CKD risk (OR = 1.38, 95% CI 1.32–1.44, P < 0.001). K-M survival curves demonstrated that over time, CKD patients in the highest ePWV quartile had significantly higher all-cause and CVD mortality rates compared to those in the lowest quartile (P < 0.05). Multivariable-adjusted Cox regression showed that, compared to the Q1 group, the Q3 (OR = 2.53, 95% CI 1.55–4.15) and Q4 (OR = 4.12, 95% CI 2.38–7.13) groups had significantly elevated all-cause mortality risk (P < 0.001). The competing risk model further confirmed that Q4 was associated with a significantly higher CVD mortality risk than Q1 (OR = 2.71, 95% CI 1.24–5.95, P = 0.013). Harrell’s C-index indicated that ePWV provided statistically significant incremental predictive value for all-cause mortality risk (Δ = + 0.02, P < 0.001). Triple sensitivity analyses consistently validated the robust association between ePWV and CKD (P < 0.001).

Conclusion

The ePWV is an independent risk factor for CKD, exhibiting a non-linear relationship with CKD risk. Gender, BMI, and T2D have a substantial impact on this association. Furthermore, higher ePWV levels were associated with increased mortality risk in CKD patients. In the future, the research on how to apply ePWV more effectively in clinical practice to identify the risk of CKD should be advanced, especially for populations with ePWV > 6.68 m/s.