Abstract
BACKGROUND: Objective biochemical markers for stratifying brachial plexus injury (BPI) severity remain lacking. Platelet-activating factor (PAF), an inflammatory mediator implicated in neurological damage, has not been evaluated in traumatic BPI. We investigated whether serum PAF levels predict complete injury and correlate with nerve conduction impairment. METHODS: This retrospective case-control study analyzed 102 traumatic BPI patients (43 complete, 59 incomplete) and 51 matched controls enrolled between January 2020 and December 2023. Serum PAF and nerve conduction velocities (NCV) of median, radial, and ulnar nerves were measured. Multivariable logistic regression, restricted cubic spline (RCS) modeling, and ROC analysis were performed. RESULTS: BPI patients exhibited elevated PAF versus controls (171.8±26.1 vs. 147.7±12.1 pg/mL, P<0.001) and reduced mean NCV (19.8±2.6 vs. 64.0±6.1 m/s, P<0.001).PAF correlated inversely with NCV (ρ=-0.50, P<0.001). Complete injuries showed higher PAF (189.3±23.5 vs. 158.9±19.8 pg/mL, P<0.001) and lower NCV (all P<0.001). PAF (OR=1.151 per pg/mL, 95% CI: 1.033–1.284, P=0.011) and mean NCV (OR=0.012, P=0.014) independently predicted complete injury. RCS analysis revealed a nonlinear threshold at 165 pg/mL (OR=5.247, 95% CI: 2.808-9.803). PAF achieved good discrimination (AUC=0.840, sensitivity 84%, specificity 68%), though inferior to NCV (AUC=0.982). A multivariable model combining PAF, NCV, and injury-to-detection interval reached near-perfect discrimination (AUC=0.997). CONCLUSIONS: Serum PAF independently predicts complete BPI with a nonlinear threshold at 165 pg/mL. While electrophysiological assessment remains the gold standard, PAF demonstrates potential as a rapid, minimally invasive screening tool in acute settings where immediate nerve conduction studies are unavailable, particularly for triaging patients requiring urgent intervention. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12883-026-04718-7.