Malnutrition–Inflammation Score is a Useful Tool in Peritoneal Dialysis Patients

Background Malnutrition–Inflammation Score (MIS) is a quantitative assessment tool based on Subjective Global Assessment (SGA) and predicts mortality and morbidity in maintenance hemodialysis patients. However, there are not enough data about the use of MIS in peritoneal dialysis (PD). In this study, relationships between MIS and prospective hospitalization indices, risk of developing peritonitis, anemia indices, and laboratory and anthropometric parameters were analyzed and compared with SGA in PD.Methods 50 PD patients (M/F 26/24, age 45.2 ± 14.9 years, mean PD duration 30.8 ± 23.1 months) were included. The same physician performed the SGA and MIS evaluations. Clinical, laboratory, and anthropometric parameters were measured.Results 18 patients were classified as SGA-A (without malnutrition), 24 as SGA-B (with moderate malnutrition), and 8 as SGA-C (with severe malnutrition). Increment in MIS was concordant with SGA groups A to C ( p < 0.0001). Peritonitis rate, number of hospitalizations, total number of hospitalization days, erythropoietin requirements, C-reactive protein (CRP), and ferritin levels were positively correlated with MIS ( p < 0.0001). Midarm muscle circumference ( p = 0.04), albumin ( p < 0.0001), prealbumin ( p = 0.001), creatinine ( p = 0.04), hemoglobin ( p = 0.003), transferrin ( p < 0.0001), and cholesterol ( p = 0.009) were negatively correlated with MIS. Correlation coefficients of hospitalization indices, peritonitis rate, anemia indices, erythropoietin requirements, albumin, prealbumin, CRP, and anthropometric parameters were higher with MIS than with SGA. In logistic regression analysis, a higher MIS was independently associated with a higher risk of future hospitalization ( p = 0.029, odds ratio 2.14, confidence interval 1.082 – 4.146).Conclusions This study demonstrated that MIS significantly correlated with clinical, nutritional, inflammatory, and anthropometric parameters and anemia indices in PD patients, and that those correlations were stronger than those with SGA.