Nutritional Predictors of One-Year Survival in Hemodialysis Patients: Insights from an Age-Matched Cohort Analysis

Abstract Background Malnutrition and protein-energy wasting are major predictors of mortality in hemodialysis patients. While several nutritional markers have been proposed, the independent prognostic value of individual variables remains debated, especially in the context of age-related confounding. Methods We retrospectively analyzed a cohort of 350 hemodialysis patients (237M/128F, mean age ± Standard deviation (SD) : 69.78 ± 13.47 years old). One-year survival was assessed, and a 1:1 age-matched selection (59 survivors and 59 non-survivors) (76M/42F, mean age ± SD :77.70 ± 11.64) was performed. Nutritional variables including GNRI, serum albumin, prealbumin, handgrip strength, body composition parametrs and weight variation were evaluated. Logistic regression models were used to assess their association with survival, with emphasis on simplified multivariate models to avoid overfitting. Results Univariate comparisons showed significantly better nutritional profiles among survivors. After matching for age, GNRI, prealbumin, and handgrip strength remained significantly associated with survival. Multivariate models including GNRI combined with either 6-month weight change or handgrip strength yielded significant and stable associations with survival (OR for GNRI: 1.10, p = 0.001; OR for handgrip: 1.08, p = 0.028). Conclusion In an age-matched hemodialysis cohort, GNRI, dynamic weight change, and muscle strength emerged as independent predictors of one-year survival. Simplified models provide robust prognostic insight and are preferable to overparameterized regressions in small datasets.