Preoperative geriatric nutritional risk index and postoperative acute kidney injury in older patients undergoing noncardiac surgery: evaluation of incremental predictive value beyond serum albumin
Year: 2026
Type: article
Study design: Observational Study
Abstract: Preoperative nutritional assessment may help identify older surgical patients susceptible to acute kidney injury (AKI). Whether the Geriatric Nutritional Risk Index (GNRI) provides information beyond serum albumin remains uncertain. We assessed the association of GNRI with postoperative AKI and whether it improved prediction beyond albumin. This retrospective cohort study used the publicly available INSPIRE database (Seoul National University Hospital, 2011–2020) and included 10,727 adults aged ≥ 65 years undergoing noncardiac surgery. Nutritional risk was defined as GNRI ≤ 98, a conventional but not AKI-validated threshold. The primary outcome was postoperative AKI within 7 days, ascertained by the creatinine component of KDIGO criteria. Because this dataset de-identifies laboratory values by replacing them with coarsened representatives (creatinine was released at only 21 discrete levels), AKI ascertainment is resolution-limited. We used multivariable logistic regression with multiple imputation; prediction models were internally validated by patient-level bootstrap using imputations fitted without the outcome. AKI occurred in 952 patients (8.9%). After adjustment for prespecified preoperative covariates, GNRI ≤ 98 was associated with AKI (adjusted OR 1.33, 95% CI 1.11–1.59; P = 0.002), but the association was attenuated to the null after additional adjustment for continuous serum albumin (adjusted OR 0.97, 95% CI 0.79–1.20; P = 0.78). Adding continuous GNRI to an albumin-containing model yielded no measurable improvement in discrimination (out-of-bootstrap AUC 0.7320 vs. 0.7319; ΔAUC − 0.0001, 95% CI − 0.0016 to 0.0015) or in calibration and net benefit, and a GNRI-based model showed no advantage over an albumin-based model (ΔAUC − 0.0007, 95% CI − 0.0035 to 0.0021). GNRI ≤ 98 was associated with higher adjusted odds of postoperative AKI, but this association was not independent of serum albumin, and we observed no detectable incremental predictive value. Because laboratory values are coarsened, the AKI frequency should be read as a resolution-limited estimate. GNRI may serve as a descriptive nutritional marker; it should not replace serum albumin in AKI prediction models, nor inform triage decisions, before external validation. Not applicable.
Source: BMC Geriatrics
Authors: Yanlin Wu, Min Chen, Lihua Lei, Zhisen Dai
Language: English
Topic: Acute Kidney Injury Research
Subfield: Nephrology
Field: Medicine
Domain: Health Sciences
Open Access status: gold
Estimated APC paid: $3,190
OpenAlex ID: W7236865840
