Association between geriatric nutritional risk index and stroke risk in hemodialysis patients: 10-Years outcome of the Q-Cohort study.

BACKGROUNDS AND AIMS: The geriatric nutritional risk index (GNRI), which is calculated using the serum albumin level and body mass index, is a nutritional marker associated with an increased risk of cardiovascular events in patients who are receiving hemodialysis. However, no studies have examined the association between the GNRI level and the incidence of stroke in this population. METHODS: Three thousand forty-five patients were registered in the Q-Cohort Study, which is a multicenter, observational cohort of hemodialysis patients. The main outcomes were brain infarction and brain hemorrhage. The main exposure was GNRI levels at baseline. Patients were divided into quartiles on the basis of baseline GNRI levels: Q1, <90.7; Q2, 90.7-95.5; Q3, 95.6-99.8; Q4, >99.8. The risk of brain infarction or hemorrhage was estimated using the multivariable-adjusted Cox proportional hazard risk models and restricted cubic spline analyses. RESULTS: During the 10-year follow-up period, 326 patients developed brain infarction and 149 patients developed brain hemorrhage. Cox proportional hazard risk models showed that the risk of brain infarction and hemorrhage in Q1 was significantly higher than that in Q4 group. The hazard ratios [95% confidence intervals] were 1.49 [1.05-2.12] and 1.89 [1.11-3.20], respectively. Restricted cubic spline curves showed that a lower GNRI was incrementally associated with an increased risk for both brain infarction and brain hemorrhage. CONCLUSIONS: Our results suggest that a lower GNRI is an independent risk factor for both brain infarction and hemorrhage in patients who are receiving maintenance hemodialysis.