Objective: Adverse outcomes in hemodialysis patients have been attributed, in part, to the pro-inflammatory state prevalent in this population. This study examines the relationship between blood neutrophil-to-lymphocyte ratio (NLR) with nutrition markers and health outcomes in hemodialysis (HD) patients.
Design: This is a 12-month prospective cohort study that recruited 77 participants from May to Jun 2017.
Settings and subjects: Patients receiving maintenance hemodialysis.
Main outcomes: Hospitalization, transplants and mortality.
Results: Of the 77 participants, 63.8% were hospitalized, 10 (13%) died of cardiovascular diseases and 6 (7.8%) had a kidney transplant. Spearman correlations using baseline values showed an inverse significant correlation between the total number of hospitalizations and BMI kg/m2 (BMI rho = - 0.37, P <0.001); a significant inverse correlation between NLR and albumin (rho = - 0.22, P = 0.028); and a significant direct correlation between baseline NLR and BMI kg/m2 (rho = 0.22, P = 0.028). Participants were grouped by their NLR value into quartiles for outcomes analysis: quartile 1 (NLR ≤ 1.75), quartile 2 (NLR 1.76-2.6), quartile 3 (NLR 2.7-3.9) and quartile 4 (NLR ≥ 4). The percentage of patients with the lowest level of inflammation (NLR ≤ 1.75) was greater for not hospitalized patients than for hospitalized (39.3% vs 16.3%, P = 0.025) and not hospitalized participants had higher BMI kg/m2 (mean ± SD) at baseline compared to those hospitalized (29.11 ± 5.4 vs 26.22 ± 5.34, P = 0.026). In a multivariate cox regression analysis, participants in the lowest quartile (NLR ≤ 1.75) were compared to the rest on hospitalization, mortality and transplant. Years in dialysis, BMI kg/m2 and NLR ≤ 1.75 were significant predictors of hospitalization after adjustment (P = 0.021, P = 0.005, P = 0.039; respectively) and we observed an association of low NLR with a hazard ratio (HR 0.44, 95% CI 0.20-0.96, P = 0.039), BMI (HR 0.90, 95% CI 0.85-0.97, P = 0.005) and years in dialysis (HR 0.90, 95% CI 0.83-0.98, P = 0.021) for hospitalization in overall participants. In a further analysis comparing the effect of low NLR in the subgroup of diabetic vs non-diabetics, it was observed that BMI kg/m2 was a significant predictor for hospitalization in the non-diabetic subgroup (P = 0.040) but not significant in the case of diabetics (P = 0.128) after adjustments. Years in dialysis and NLR ≤ 1.75 were significant predictors of hospitalizations in the subgroup of diabetic before and after adjustment (P = 0.049, P = 0.044; respectively). Having a low NLR decreased 73% the risk for hospitalization (HR 0.27 95% CI 0.07-0.96, P = 0.044) in this subgroup. Survival and hospitalization curves were analyzed by comparing all participants and the diabetic subgroup, in the lowest inflammation quartile vs the rest (NLR ≤ 1.75 vs NLR > 1.75). Participants with NLR ≤ 1.75 had 100% survival rate (log-rank test, P = 0.059) and lower hospitalization rate (log-rank test, P = 0.025); participants with diabetes had lower hospitalization rate (log-rank test, P = 0.039).
Conclusion: NLR at baseline was associated with nutritional markers (albumin, BMI). Low NLR at baseline was a predictor of lower risk for hospitalizations in HD patients with diabetes.
Keywords: Hemodialysis; Hospitalization; Inflammation; Mortality; NLR.