Regression analysis of serum microRNAs in predicting early neurological deterioration in patients with branch atheromatous disease
Nie Li
Qiao Xiangliang
Cheng Wei
Hu Suhua
Hu Huanhuan
Abstract：Objective To perform a regression analysis on the predictive value of serum micro-RNAs(miR)for early neurological deterioration(END)in patients with branch atheromatous dis-ease(BAD).Methods A total of 134 BAD patients admitted in our department from February 2020 to February 2023 were enrolled,and according to the END status,they were divided into de-terioration group(28 cases)and non-deterioration group(106 cases).Serum levels of miR-130a,miR-210,miR-141-3p and miR-29a-3p were measured at admission.NIHSS score was used to eval-uate the END status at admission and at 7 d after admission.Binary logistic regression analysis was adopted to construct a model of above four miRs in predicting END in BAD patients.ROC curve was plotted to assess the predictive value of the four miRs alone or combined together for END.Results The serum levels of miR-130a and miR-210 were significantly higher,while those of miR-141-3p and miR-29a-3p were obviously lower in the deterioration group than the non-deterioration group(P＜0.01).Logistic regression analysis indicated that serum miR-130a,miR-210,miR-141-3p and miR-29a-3p were independent predictors of END in BAD patients(P＜0.05,P＜0.01).ROC curve analysis showed that the AUC value of the four miRs combined together in predicting END in BAD patients was 0.977(95％CI:0.936-0.995),with a sensitivity of 96.43％and a specificity of 90.57％,and the combined detection exhibited better predictive effi-ciency than each indicator alone(P＜0.01).Conclusion Serum miR-130a,miR-210,miR-141-3p and miR-29a-3p have certain value in predicting END in BAD patients,and their combined detec-tion can enhance its predictive efficiency.
Keywords：atherosclerosisneurologic manifestationsmicroRNAsregression analysisforecasting
Publication Date：2025-08-15
Online Publishing Date：2025-08-21(First online date of this platform, not the publication date of the document)
Pages：4( 1047-1050 )