Hua Liu,1,2,* Jinrong Wang,1,2,* Wenming Wang,1 Min Ruan,3 Jiangang Liu4 

1Department of Neurosurgery, Affiliated Kunshan Hospital of Jiangsu University, Kunshan, Jiangsu, 215300, People’s Republic of China; 2Kunshan Biomedical Big Data Innovation Application Laboratory, Kunshan, Jiangsu, 215300, People’s Republic of China; 3Department of Neurosurgery, Kunshan Hospital of Traditional Chinese Medicine, Kunshan Affiliated Hospital of Nanjing University of Chinese Medicine, Kunshan, Jiangsu, 215300, People’s Republic of China; 4Department of Neurosurgery, The First Affiliated Hospital of Soochow University, Suzhou, Jiangsu, 215000, People’s Republic of China

*These authors contributed equally to this work

Correspondence: Min Ruan, Department of Neurosurgery, Kunshan Hospital of Traditional Chinese Medicine, Kunshan Affiliated Hospital of Nanjing University of Chinese Medicine, 91 East Qianjin Road, Yushan Town, Kunshan, Jiangsu, 215300, People’s Republic of China, Email rei128@163.com Jiangang Liu, Department of Neurosurgery, The First Affiliated Hospital of Soochow University, 188 Shizi Street, Gusu District, Suzhou, Jiangsu, 215000, People’s Republic of China, Email liujiangang@suda.edu.cn

Background: Traumatic brain injury (TBI) is a leading cause of disability and death worldwide, involving complex pathophysiological responses such as metabolic disturbance and systemic inflammation. This study aimed to evaluate the prognostic value of selected metabolic and inflammatory biomarkers in predicting short- and medium-term mortality in patients with moderate-to-severe TBI.
Methods: We conducted a retrospective cohort study of patients with TBI admitted between March 29, 2018, and July 31, 2023. Clinical data, including a panel of metabolic (eg, triglyceride-glucose index [TYG], APOB/A1 ratio) and inflammatory biomarkers (eg, neutrophil-to-platelet ratio [NPR]), were collected within 24 hours of admission. Mortality was assessed at 14 days, 30 days, and hospital discharge. Multivariate Cox regression models and ROC curve analysis were used to assess prognostic associations and model performance.
Results: A total of 2555 patients were enrolled, of whom 579 (22.67%) underwent surgical treatment. Multivariate Cox proportional hazards regression analysis revealed that the triglyceride-glucose index (TYG) was an independent predictor of short-term mortality in TBI patients, while the neutrophil-to-platelet ratio (NPR) and apolipoprotein B/A1 (APOB/A1) ratio were independent predictors of both short- and mid-term mortality. In addition, surgical treatment was associated with an increased risk of mid-term mortality, while tracheostomy significantly reduced mortality risk across all time points. Receiver operating characteristic (ROC) curve analysis showed that the regression model incorporating inflammatory markers had the highest areas under the curve (AUCs) of 0.904, 0.897, and 0.897, demonstrating superior performance in predicting short- and mid-term mortality. Additionally, in the subgroup analysis of non-operation patients, TYG and NPR had a more significant impact on mortality risk.
Conclusion: Metabolic and inflammatory biomarkers, including TYG, NPR, and APOB/A1 ratio, provide valuable prognostic information in patients with TBI. These markers may assist clinicians in early risk stratification and personalized treatment planning.

Keywords: traumatic brain injury, triglyceride-glucose index, neutrophil-to-platelet ratio, prognosis, cox proportional hazards model