NUTRITIONAL RISK INDEX FOR JAPANESE HEMODIALYSIS PATIENTS (NRI-JH) CATEGORIES PREDICT MORTALITY AND DISTINCT METABOLOMIC PROFILES IN CONVENTIONAL AND EXTENDED-HOURS HEMODIALYSIS

The Nutritional Risk Index for Japanese Hemodialysis patients (NRI-JH), developed by the Japanese Society for Dialysis Therapy (JSDT), is a composite nutritional prognostic score that identifies protein-energy wasting (PEW) in Japanese patients, with Medium- and High-risk groups corresponding to PEW. Its prognostic utility under intensive dialysis regimens and its relationship with metabolomic profiles remain unclear.

The NRI-JH was assessed using pre-dialysis blood samples collected between January–March 2020 at two facilities providing in-center daytime extended-hours HD (6–8 h/session) and one facility with conventional HD (3–5 h/session) in Japan, including 286 patients on conventional HD and 186 on extended-hours HD. Samples were obtained at the beginning of the dialysis week (first session after the 2-day interval). NRI-JH was calculated based on four components: low body mass index (<20 kg/m²), low serum albumin (modified bromocresol purple; <3.4 g/dL if <65 years, <3.2 g/dL if ≥65 years), abnormal total cholesterol (<130 or ≥220 mg/dL), and low serum creatinine (age- and sex-specific cutoffs). Patients were stratified into Low (0–7) vs Combined Medium/High (≥8) risk groups. Survival was analyzed separately by modality using Kaplan–Meier curves, followed by Cox proportional hazard models adjusted for age, sex, diabetes, cardiovascular disease, malignancy, and vascular access. Metabolomic profiling was performed on the same samples using capillary electrophoresis time-of-flight mass spectrometry (CE-TOFMS), and metabolite levels were compared between the two risk groups within each modality.

The mean age was 68.6 ± 11 years in conventional HD and 66.0 ± 11 years in extended-hours HD. The proportion of patients classified as Combined Medium/High risk was 26.2% in conventional HD and 11.8% in extended-hours HD. Across both modalities, the Combined Medium/High risk group was associated with higher mortality than the Low risk reference. This association was evident in conventional HD (adjusted HR 2.05, 95% CI: 1.24–3.40) and persisted in extended-hours HD with an even greater effect size (adjusted HR 3.78, 95% CI: 1.51–9.40), although the interaction between modality and NRI-JH risk was not significant (P for interaction = 0.41). Metabolomic analysis showed that branched-chain amino acids (BCAAs) were significantly lower in Combined Medium/High risk group compared with Low risk group in conventional HD, whereas no significant differences were observed in extended-hours HD. Among uremic compounds, hippurate and guanidinosuccinic acid were consistently lower in Combined Medium/High risk compared with Low risk across both modalities. Known protein-bound solutes (e.g., indoxyl sulfate) and other small water-soluble compounds (e.g., ADMA) showed no consistent differences between the two risk groups.

NRI-JH identified Medium-to-High-risk patients—corresponding to PEW—who exhibited higher mortality than Low risk patients across dialysis modalities. In conventional HD, this group also demonstrated lower BCAA levels, suggesting potential benefits from enhanced nutritional support. In extended-hours HD, patients in the Medium-High category remained at elevated risk, indicating that strategies to further mitigate uremic retention may be needed in conjunction with comprehensive care approaches.