IntroductionIdentifying controllable risk factors for large-artery atherosclerosis (LAA) stroke is crucial due to its significant role as a leading cause of ischemic stroke. We aimed to validate the correlation of serum vitamin B12 with LAA stroke. MethodsInpatients with LAA stroke and healthy controls were retrospectively collected for a case-control study from January 2020 to May 2022. Serum vitamin B12 concentration and other blood indicators, demographic, lifestyle factors and comorbidities were investigated. Logistic regression analysis was used to identify the correlation of serum vitamin B12 concentrations with LAA stroke, meanwhile adjusted for confounding factors. ResultsPatients with LAA stroke had significantly lower serum vitamin B12 concentrations in comparison to those of controls. In the fully adjusted model, vitamin B12 (per 1 interquartile range increase, odds ratio [OR] = 0.84, 95 % confidence interval [CI]: 0.77–0.91), vitamin B12 < 200 pg/mL (OR=7.70, 95 %CI: 2.19–27.03) and vitamin B12 < 300 pg/mL (OR=4.19, 95 %CI: 1.82–9.66) were independently factors for LAA stroke. Furthermore, the optimal cut-off values for vitamin B12 to predict LAA stroke were 305.25 pg/mL (area under the curve [AUC] = 0.71) when unadjusted and 308.25 pg/mL when adjusted for age and sex (AUC=0.68). Lower vitamin B12 concentrations were significantly associated with male sex, smoking, older age, higher neutrophil count, higher creatinine, lower folate and higher total homocysteine. ConclusionResults indicate that low concentration of serum vitamin B12 may be a strong predictor for the risk of LAA stroke.