Associations of prognostic nutritional index with cardiovascular all-cause mortality among CVD patients with diabetes or prediabetes: evidence from the NHANES 2005-2018.

Background: Immunonutritional status is linked to the prognosis of cardiovascular disease (CVD) and diabetes, but the relationship between immunonutritional disorders and clinical outcomes in CVD patients with diabetes is unclear. This study aims to investigate the association of the novel immunonutritional indicator of prognostic nutritional index (PNI) with all-cause and CVD mortality in diabetic and prediabetic CVD patients. Method: This is an open-cohort study involving 1,509 CVD patients with diabetes or prediabetes collected from The National Health and Nutrition Examination Survey (NHANES) and initially interviewed between 2005 and 2018. Subjects were followed up until on December 31, 2019. Mortality outcomes and causes of death were obtained from National Death Index (NDI) records. We used restricted cubic spline (RCS) and maximally selected rank statistics method (MSRSM) to assess the nonlinearity of the PNI-mortality association and determine the optimal PNI cutoff for survival outcomes. Additionally, weighted multivariable Cox regression models, subgroup analyses, and interaction tests were employed to examine the relationship between PNI and all-cause and CVD mortality. The predictive accuracy of PNI for survival outcomes was evaluated using time-dependent receiver operating characteristic curve (ROC) analysis. Results: During a median follow-up of 61 months (interquartile range, 33-103 months), 507 of the 1509 (33.60%) diabetic or prediabetic CVD patients died. A negative and nonlinear association between PNI and all-cause/CVD mortality was identified by RCS analysis in all patients. In the fully-adjusted Cox regression model, in the entire cohort, higher PNI (≥46.5) was significantly associated with reduced risks for all-cause and CVD mortality. A consistent association between PNI and all-cause/CVD mortality was observed in diabetic CVD patients, but not in prediabetic CVD patients. No significant interaction between PNI and other covariates was observed (all P interaction >0.05). Time-dependent ROC curve revealed that the areas under the curve (AUC) of PNI for 1-, 3-, 5-, and 10-year survival rates were 0.66, 0.66, 0.66, and 0.67 for all-cause mortality, and 0.72, 0.70, 0.72, and 0.69 for CVD mortality, respectively. Conclusion: Increased PNI is significantly associated with reduced risks for all-cause and CVD mortality in diabetic or prediabetic CVD patients, especially for diabetic CVD patients.