摘要

Quantitative method and multiple statistical methods were used to compare the effect of folic acid and its related metabolites on risk assessment and prediction of neural tube defects. First, a previously established HPLC-MS/MS method was employed to determine the levels of these metabolites in maternal serum samples of 121 cases and 118 controls. Then two-sample t test (p<0.05) was used to obtain differential metabolites between these two groups. At last, Logistic regression and receiver operating characteristic (ROC) analysis were employed to calculate the odds ratio, sensitivity and specificity, positive and negative predictive rate of these metabolites. Results showed that among the 4 differential metabolites, 5-methyltetrahydrofolate and homocysteine both got higher odds ratio, positive and negative predictive rate than folic acid. The method was suitable for screening and assessment of disease biomarkers.