Abstract
OBJECTIVE: This study evaluates the predictive value of prognostic nutritional index (PNI) for all-cause and cardiovascular mortality in chronic kidney disease (CKD) patients based on data from the National Health and Nutrition Examination Survey (NHANES), and to explore its variability across different CKD stages. METHODS: A total of 4,528 CKD patients from the NHANES database (1999-2018) were included. Cox regression models were used to analyze the association between PNI quartiles (Q1-Q4) and mortality risk. Restricted cubic spline (RCS) analysis was employed to explore non-linear relationships, and subgroup and mediation analyses were conducted. RESULTS: Patients in low PNI group (Q1) exhibited significant metabolic disturbances including elevated blood urea nitrogen and creatinine, reduced albumin and estimated glomerular filtration rate (eGFR). Compared to the Q4 group, the Q1 group had a 67% increased risk of all-cause mortality (HR: 0.598, 95% CI: 0.517-0.692) and a 103% increased risk of cardiovascular mortality (HR: 0.492, 95% CI: 0.374-0.648). RCS analysis revealed a non-linear relationship between PNI and mortality risk (threshold: 52), with significant predictive efficacy in CKD stages 1, 4, and 5 (P < 0.05), but not in stages 2 and 3 (P > 0.05). Mediation analysis indicated that age partially mediated the association between PNI and mortality (indirect effect proportion: 33%), while eGFR showed no mediating effect (P > 0.05). CONCLUSION: PNI is an independent predictor of all-cause and cardiovascular mortality in CKD patients. Future longitudinal studies are warranted to validate its clinical utility and intervention potential.