Introduction
Bloodstream infections (BSIs) pose a severe threat to elderly patients, who are disproportionately affected by malnutrition and frailty. The Geriatric Nutritional Risk Index (GNRI), a simple calculator using albumin, height, and weight, has emerged as a practical tool for predicting morbidity and mortality in hospitalized seniors. This retrospective study from a large Beijing tertiary hospital assesses whether GNRI can forecast adverse outcomes in elderly patients with BSIs and, if so, how clinicians might leverage this information to guide nutritional interventions.
Methods
From January 2020 to December 2021, a two-year retrospective analysis identified 464 elderly inpatients (age ≥60) with healthcare-associated BSIs. GNRI was calculated at admission and categorized into four risk levels: no risk (GNRI > 98), low risk (92–98), moderate risk (82–92), and major risk (<82). The primary outcome was all-cause in-hospital mortality. Statistical approaches included univariate and multivariate logistic regression to adjust for potential confounders, as well as restricted cubic spline (RCS) analysis to explore potential nonlinear associations between GNRI and mortality. A robust validation strategy incorporated 100 replications of case–control subsamples to examine model stability, and model discrimination was assessed by ROC curves (AUC). Data were drawn from a 3,500-bed tertiary center in Beijing, with ethical approval noted in the study protocol.
Key Findings
Among the cohort, 203 patients (43.8%) had GNRI > 98 (no nutritional risk), 70 (15.1%) were low risk, 118 (25.4%) moderate risk, and 73 (15.7%) major risk. Higher GNRI-related malnutrition levels correlated with increased in-hospital mortality. In multivariate analysis, GNRI remained an independent predictor of mortality: GNRI 98 yielded an odds ratio (OR) of 3.16 (95% CI: 1.52–6.58; P = 0.002), and GNRI 82 to <92 (moderate risk) yielded an OR of 1.91 (95% CI: 1.00–3.62; P = 0.049). Conversely, each unit increase in GNRI slightly reduced mortality risk (OR: 0.96; 95% CI: 0.94–0.98; P = 0.001). The ROC analysis demonstrated good discriminative ability of the model (AUC around 0.824). Cross-validated results showed stable performance with average AUCs near 0.828 for training and 0.825 for testing datasets.
Nonlinear Relationship Between GNRI and Mortality
The restricted cubic spline analysis revealed a nonlinear, inverse association: mortality risk declined as GNRI increased, with the curve leveling off once GNRI reached roughly 96–98. This plateau suggests that beyond a certain nutritional risk threshold, further GNRI improvements may confer diminishing additional survival benefit in the context of BSI in elderly patients. These findings imply that GNRI is not only a binary risk marker but also a nuanced continuum that can help stratify patients who may benefit most from intensified nutritional support.
Clinical Implications
The study supports GNRI as a practical, bedside tool for early risk stratification in elderly BSI patients. Clinicians can identify high-risk individuals at admission, prompting timely nutritional assessment and targeted interventions. For those with moderate to major GNRI risk, proactive strategies—such as tailored enteral or parenteral nutrition, optimized macronutrient and micronutrient delivery, and close monitoring of glycemic control and inflammatory markers—may improve outcomes. The findings also highlight the role of invasive devices, such as indwelling urinary catheters and central venous lines, as contributors to poor prognosis, underscoring the need for judicious device use and rigorous infection control alongside nutritional care.
Limitations and Future Directions
Despite its strengths, the study has limitations: it is a single-center retrospective analysis, GNRI was measured only at admission, and potential Neyman bias may affect results. Although robust internal validation was used, multi-center prospective studies are needed to confirm generalizability across diverse populations and settings. Future work should explore GNRI trajectories over hospitalization and their relationship with long-term outcomes, as well as the impact of specific nutritional regimens on GNRI-improved risk profiles in elderly BSI patients.
Conclusion
In this two-year retrospective cohort of elderly patients with BSIs, GNRI emerged as a significant, independent predictor of in-hospital mortality, with a nonlinear relationship showing pronounced risk reduction as GNRI rises toward the 96–98 range. Early GNRI-based nutritional assessment can help clinicians identify high-risk patients who may benefit most from timely, individualized nutritional support, potentially reducing BSI-related mortality in the aging population.
