Applicability and predictive validity of the global leadership initiative on malnutrition criteria for older patients with sepsis according to different muscle mass assessment methods

To evaluate the applicability of the Global Leadership Initiative on Malnutrition (GLIM) criteria in older patients with sepsis and to compare the predictive validity for 28-day mortality of different muscle mass assessment methods in the emergency department. Prospective cohort study. Emergency department. Older patients (≥65 years) with sepsis. Muscle mass was assessed using three methods: (1) the skeletal muscle index at the third lumbar vertebra (L3) on computed tomography (CT) scans; (2) calf circumference (CC), and (3) mid-upper-arm circumference (MAC). Cox regression analysis was performed to assess the association between the GLIM criteria and 28-day all-cause mortality. Additionally, the C-statistic, net reclassification improvement (NRI), and integrated discrimination improvement (IDI) were used to evaluate the predictive validity of the three instruments. Survival curves were assessed using the Kaplan-Meier method and compared using the log-rank test. A total of 598 patients with sepsis were included. The prevalence of malnutrition according to GLIM-CT, GLIM-CC, and GLIM-MAC was 53.3%, 63.0%, and 40.8%, respectively. Cox regression analysis revealed that the GLIM criteria were independent risk factors for all-cause 28-day mortality. Incorporation of GLIM-CT, GLIM-CC, or GLIM-MAC into a base model significantly improved the C-statistic. The model including GLIM-CT had the highest C-statistic, improving the C-statistic of the base model from 0.780 (95% confidence interval CI: 0.741-0.819) to 0.823 (95% CI: 0.789-0.857). This improvement in risk prediction was also confirmed via category-free NRI and IDI, suggesting that GLIM-CT had the best performance. Kaplan-Meier survival analysis showed that patients with malnutrition defined according to the GLIM criteria had a greater probability of 28-day mortality (log-rank, P < 0.001). Malnutrition, defined via any of the three methods, was predictive of 28-day mortality among older patients with sepsis in the emergency department. GLIM-CT had the best predictive validity.