Association and predictive value of geriatric nutritional risk index, body composition, or bone mineral density in hemodialysis patients.

2020 
AIM Assess the association and predictive value of geriatric nutritional risk index (GNRI), body composition, and bone mineral density (BMD) in hemodialysis (HD) patients. METHODS Laboratory data, body composition parameters measured via body composition monitor, and radius, lumbar spine, femoral neck BMD measured using dual energy X-ray absorptiometry were assessed in all subjects on HD or online hemodiafiltration (HDF) at baseline. Regression analysis for GNRI, Cox proportional hazard analyses and comparison of multiple receiver operating characteristic (ROC) curves were performed. RESULTS Among all 264 patients, age was 65±12 years and dialysis vintage was 79 (39-144) months. GNRI tertile (T)1, T2, and T3 were 88 (85-91), 94 (93-95), and 98 (97-101), respectively. Patients in GNRI T1 had lower fat tissue index (FTI), lean tissue index, and femoral neck, lumbar spine, and distal mid-third radius BMD, but higher overhydration/extracellular fluid than patients in GNRI T2 or T3 (P<0.05). GNRI was significantly associated with FTI, lean tissue index, and femoral neck, lumbar spine, and distal mid-third radius BMD (P<0.01). GNRI was a significant predictor of 2-year all-cause mortality (HR 0.92, P<0.05). Area under the ROC curve for all-cause mortality using traditional risk factors (age, sex, diabetes mellitus, cardiovascular disease, use of vasopressors for dialysis-related hypotension, and C-reactive protein) was 0.67 and changed by adding GNRI (0.78, P<0.05), FTI (0.75), or femoral neck BMD (0.66), respectively. CONCLUSION Associations between GNRI, body composition, and BMD were confirmed in HD patients. Combining GNRI with traditional risk factors improved mortality prediction in HD patients. This article is protected by copyright. All rights reserved.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    29
    References
    2
    Citations
    NaN
    KQI
    []