Psoas muscle to L3 vertebral body ratio, a CT-based psoas assessment metric, associates with nutritional status and predicts prolonged hospitalization in geriatric trauma patients: a 5-year retrospective study

The incidence of geriatric trauma has been increasing annually. Geriatric individuals tend to suffer a greater risk of malnutrition after they experience trauma. However, there is no well-verified and suitable tool for assessing the nutritional state of this population. Psoas volume assessment based on computed tomography (CT) can reflect the whole-body muscle level and nutritional state, from which the psoas muscle to L3 vertebral body ratio (PML3) is objective and highly accessible. Our study aimed to investigate the association between the PML3 and nutritional status in geriatric trauma patients and clarify the ability of the PML3 to predict prognosis in these patients. A 5-year retrospective study of geriatric trauma patients in our institution older than 65 years was conducted. CT imaging of these patients at admission was utilized to calculate the PML3. Simple linear regression analysis was conducted to evaluate the associations between the PML3 and other nutritional indicators. Furthermore, we conducted univariable correlation and multivariable logistic regression analyses to investigate the association between the PML3 and prognostic outcomes (in-hospital mortality and prolonged hospitalization) of geriatric trauma patients. Receiver operating characteristic (ROC) curve analysis was conducted to assess the performance in predicting prognostic outcomes. A total of 439 geriatric trauma patients from January 2018 to December 2022 were included in our study; 286 patients (65.1%) were males, and the median age was 69 years. Thirty-one patients (7.1%) died during hospitalization. The PML3 distribution in this population was significantly associated with nutritional indicators, including the albumin, cholesterol concentrations and the Prognostic Nutritional Index. The prognostic outcome results revealed that the PML3 was independently associated with prolonged hospitalization in geriatric trauma patients (OR = 0.013, 95% CI: 0.005–0.034). However, no significant independent association existed between the PML3 and in-hospital death (OR = 0.160, 95% CI: 0.019–1.375). ROC analysis revealed that compared with WBC (AUROC = 0.615) and neutrophils (AUROC = 0.634), the PML3 (AUROC = 0.903) had the best performance in predicting prolonged hospitalization. Further subgroup analysis also suggested good utility of the PML3 in predicting prolonged hospitalization among geriatric trauma patients admitted or not admitted to the ICU. CT-based psoas muscle assessment using the PML3 as a metric was closely associated with the nutritional status of geriatric trauma patients. Additionally, the PML3 was independently associated with prolonged hospitalization and demonstrated promising predictive value for extended hospitalization in geriatric trauma patients. Not applicable.