Objective  To explore the predictive value of methylenetetrahydrofolate reductase (MTHFR) gene C677T polymorphism and serum homocysteine (Hcy) level for adverse pregnancy outcomes.

Methods  The clinical data of 698 pregnant women who received antenatal care and delivered at Ji'an Central People's Hospital were retrospectively collected. According to the research purpose, the pregnant women were divided into internal study cohort (n=483) or external validation cohort (n=215). Each cohort was further divided into adverse pregnancy group and normal pregnancy group based on whether adverse pregnancy outcomes occurred. The MTHFR gene C677T polymorphism and serum Hcy level of all pregnant women were detected. Binary Logistic regression analysis was used to investigate the relationships of MTHFR gene C677T polymorphism, Hcy level, and other indicators with adverse pregnancy outcomes. Receiver operating characteristic (ROC) curves were plotted to analyze the predictive efficacy of different indicators when used alone and their combination.

Results  In the internal study cohort, compared with the normal pregnancy group, the detection frequency of the CC genotype at the MTHFR gene C677T locus was lower, while the detection frequencies of the CT and TT genotypes were higher in the adverse pregnancy group (P < 0.001). The proportion of pregnant women with high Hcy levels was higher in the adverse pregnancy group than that in the normal pregnancy group (P < 0.001). Binary Logistic regression analysis showed that age, folic acid level during pregnancy, MTHFR gene C677T polymorphism, and Hcy level were all independent influencing factors for adverse pregnancy outcomes (P < 0.05). The ROC curve indicated that the area under the curve (AUC) of the combined prediction model was 0.938, which was greater than that of each indicator. In the external validation cohort, the AUC of the combined prediction model was 0.917, validating the stability and clinical applicability of the model.

Conclusion  The TT and CT genotypes at the MTHFR gene C677T locus and high Hcy level are all independent risk factors for adverse pregnancy outcomes. The prediction model constructed by combining these factors with age and folic acid level during pregnancy has a high predictive efficacy for adverse pregnancy outcomes.