Jinyan Fang, Yuanhan Lin, Linjun Zhao, Xue Qiang Zhao, Jinyu Huang
Sepsis is frequently accompanied by severe metabolic disturbances and malnutrition, with nutritional status being an important modifiable factor affecting patient outcomes. The Prognostic Nutritional Index (PNI) and the Geriatric Nutritional Risk Index (GNRI) are multidimensional tools for assessing nutritional status. However, the association between PNI/GNRI and 60-day mortality in patients with sepsis remain inadequately elucidated. This study aimed to evaluate the value of PNI and GNRI for predicting the mortality risk in sepsis based on analysis of data from the MIMIC-IV database. A total of 1649 patients with sepsis were enrolled and stratified into tertiles of PNI and GNRI. We performed Kaplan-Meier survival analysis, multivariate Cox regression analysis, restricted cubic spline (RCS) analysis, complemented by subgroup analyses stratified by age, comorbidities, and treatments. Patients in the PNI- and GNRI-Q3 groups (high-score groups) demonstrated a 53% reduction in 60-day mortality risk relative to those in the PNI- and GNRI-Q1 groups (low-score groups) [hazard ratio (HR) = 0.47, 95% confidence interval (CI): 0.30-0.73]. RCS analysis revealed a nonlinear inverse correlation between PNI and mortality risk, with elevated mortality risk (HR > 1) observed at low scores of PNI (< 20). Subgroup analyses indicated that PNI exhibited more pronounced protective effects in patients aged ≤ 65 years (HR = 0.33) and those treated with β-lactam antibiotics (HR = 0.38). In contrast, GNRI showed better predictive value in elderly patients (> 65 years, HR = 0.61) and those with diabetes (HR = 0.37). Both PNI and GNRI are independent predictors of 60-day mortality in patients with sepsis. By integrating nutritional and immunological parameters, they provide valuable prognostic value beyond conventional measures. Incorporating PNI and GNRI into risk evaluation frameworks may help identify patients who may benefit from early nutritional interventions.