Geriatric nutritional risk index, muscle function, quality of life and clinical outcome in hemodialysis patients

Background & aims The geriatric nutritional risk index (GNRI) has been reported as a useful predictor of prognosis in maintenance hemodialysis (MHD) patients, demonstrating GNRI less than 90 as a marker of a poorer nutritional status and significantly increased mortality. We tested whether GNRI as a whole associated stronger with clinical and laboratory surrogates of nutrition and inflammation, muscle function, health-related quality of life (QoL), and predicts all-cause and cardiovascular (CV) morbidity and mortality in this population better than its individual components (albumin and body weight to ideal body weight ratio). Methods A prospective observational study with a median follow-up of 30 months (interquartile range - 19–41 months) was performed on 352 MHD outpatients (38.0% women) with a mean age of 67.4 ± 13.2 years. All-cause and cardiovascular hospitalization and mortality, GNRI, handgrip strength (HGS), body composition parameters (anthropometry and bioimpedance) and short form 36 (SF-36) quality-of-life scores were measured. Multivariate linear regression analyses were performed to obtain adjusted correlations. Receiver operating characteristic (ROC) curves were generated and multivariate Cox proportional hazards models were applied to identify the predictive value of GNRI and its components separately. Results GNRI positively correlated with total score (r = 0.15, P < 0.05), the physical health dimension (r = 0.14, P < 0.05), the general health (r = 0.18, P < 0.01) and some other scales of the SF-36. A significant correlation of GNRI with HGS in male patients didn't stand up to multivariable adjustments. For each one unit increase in baseline GNRI levels, the first hospitalization hazard ratio (HR) after adjustments for confounders was 0.98 (95% confidence interval (CI), 0.97 to 0.99) and the first CV event HR was 0.98 (95% CI, 0.97 to 0.99); all-cause death HR was 0.97 (95% CI, 0.96 to 0.99) and CV death HR was 0.97 (95% CI, 0.95–0.99). Albumin was related to QoL and clinical outcomes with higher strength and magnitude than GNRI. Conclusions Despite the significant relationship with clinical outcomes and QOL, GNRI is not better and is even slightly worse than albumin's performance. This raises doubts as to the clinical utility of GNRI as a prognostic tool in the MHD population.