Related Experiment Video
Updated: May 5, 2026
A Preclinical Controlled Cortical Impact Model for Traumatic Hemorrhage Contusion and Neuroinflammation
Published on: June 10, 2020
Predicting Sleeping Disorders After mTBI: A Role for Inflammation and Brain Network Biomarkers
Rui Du1, Yue Yu1,2,3,4, Xiwen Zhu2
1Department of Radiology, Zhenjiang First People's Hospital, Zhenjiang, Jiangsu, People's Republic of China.
Purpose:
To predict the occurrence of sleeping disorders (SD) in patients with mild traumatic brain injury (mTBI) 3 months after injury.
Methods:
This study recruited a total of 232 patients with mTBI and underwent a three-month follow-up period. Demographic information, MRI images, and inflammatory factor levels were collected one month after injury and PSQI (Pittsburgh Sleep Quality Index) scores were collected three times respectively on admission, 1 month and 3 months after injury. These mTBI patients were divided into those with SD group (mTBI-SD, n=130) and without SD group (mTBI-ND, n=85) based on PSQI score three months after injury. Differential indicators were used to construct univariate and multivariate logistic regression models, and receiver operating characteristic (ROC) curves were plotted. Pearson correlation analysis was conducted to explore the relationship between the differential indicators and PSQI scores.
Results:
Compared to the mTBI-ND group, patients in the mTBI-SD group exhibited lower levels of OLF.L nodal efficiency, ACG.L nodal efficiency, rich-club connection strength, and feeder connection strength, as well as higher levels of IL-8, IL-10, and TNF-α. In the univariate logistic regression model, OLF.L, ACG.L, rich-club connection strength, IL-8, and TNF-αwere identified as risk factors for the occurrence of SD three months after injury. Their Area Under the Curve (AUC) values were 0.669, 0.589, 0.672, 0.649, and 0.709, respectively. Among them, OLF.L nodal efficiency (78.80%) and rich-club connection strength (76.50%) exhibited higher specificity, while TNF-α (73.82%) demonstrated higher sensitivity. According to the multivariate regression results, the combined model constructed had an ROC-AUC of 0.809, with an accuracy of 75.35%, a sensitivity of 74.62%, and a specificity of 76.47%. The correlation results indicate that OLF.L nodal efficiency, rich-club connection strength and TNF-α are significantly correlated with PSQI scores three months after injury (rOLF.L=-0.461, rrich-club =-0.563, rTNF-α=0.538).
Conclusion:
The logistic regression model and ROC curve based on OLF.L nodal efficiency, rich-club connection strength and TNF-α can effectively predict the occurrence of SD in mTBI patients 3 months after injury.
