Geriatric nutritional risk index is associated with mortality in peritoneal dialysis patients

Background Malnutrition is one of the most common complications among dialysis patients. The Geriatric Nutritional Risk Index (GNRI) is rarely used in dialysis patients, especially peritoneal dialysis (PD). Aim To use the GNRI to evaluate the initial nutritional state of PD patients and to examine the association between the GNRI and mortality in chronic PD patients. Methods We retrospectively examined the medical records at our centre to identify all adults (>= 18 years) who had undergone PD for over 3 months before recruitment from January 2005 to December 2017. The correlation between the GNRI and mortality was examined by Kaplan-Meier and Cox proportional hazards analyses. Results A total of 1804 patients was enrolled in the study. Significant correlations were noted between the initial GNRI and Charlson index, uric acid, blood calcium, potassium, triglycerides, low-density lipoprotein cholesterol, haemoglobin and so on. Multivariate Cox proportional hazards analyses demonstrated that the GNRI was associated with all-cause mortality (hazard ratio = 0.96, P < 0.001, 95% confidence interval: 0.95-0.98) after adjustment. Compared with the lowest GNRI group, all-cause mortality decreased significantly for each level of GNRI after adjusting for various influencing factors, and the mortality risk of the highest GNRI grade decreased by 66%. The Kaplan-Meier analysis survival rate was significantly different among the four groups in terms of all-cause mortality and cardiovascular and cerebrovascular mortality (log-rank test, P < 0.05). Conclusions These results demonstrated that the GNRI is significantly associated with mortality and can be a simple, clinically useful marker for the assessment of nutritional status in PD patients.