Predictive value of novel inflammatory markers for mortality in patients with chronic kidney disease ‒ Evidence from...

Introduction: this study evaluated the predictive value of inflammatory indices including neutrophil-percentage-to-albumin ratio (NPAR), systemic immune-inflammation index (SII), neutrophil-to-lymphocyte ratio (NLR) and eosinophil-to-lymphocyte ratio (ELR) for mortality in adults with chronic kidney disease (CKD). Methods: this study analyzed adults aged 40-79 years with CKD from the National Health and Nutrition Examination Survey (NHANES) 1999-2018. Cox regression models assessed associations between inflammatory indices and all-cause/cardiovascular disease (CVD) mortality. Receiver operating characteristic curves evaluated their predictive accuracy. Restricted cubic splines and recursive algorithm analyzed threshold effects. Subgroup analyses were performed by demographic and clinical characteristics. Results: among 2496 subjects, elevated NPAR (HR: 2.05, 95 % CI: 1.62-2.59), NLR (HR: 2.28, 95 % CI: 1.78-2.93), ELR (HR: 1.61, 95 % CI: 1.27-2.03) and SII (HR: 1.61, 95 % CI: 1.26-2.06) were independently associated with increased all-cause mortality risk after adjustment. Similar patterns were observed for CVD mortality. NPAR showed the highest predictive value for both all-cause mortality (AUC: 0.610) and CVD mortality (AUC: 0.599). Non-linear relationships and threshold effects were identified between NPAR and mortality. The associations remained consistent across subgroups, with NPAR showing enhanced prediction in diabetic patients (p for interaction = 0.029). Conclusion: NPAR, SII, and NLR independently predicted mortality in CKD patients. NPAR demonstrated the strongest predictive capability, particularly in patients with diabetes, suggesting its potential as a practical tool for mortality risk stratification in CKD.