Main findings
To the best of our knowledge, this is the first study to comprehensively investigated the prognostic impact of different types of LM on the subsequent pregnancy outcomes. The results of this retrospective cohort study indicate a relationship between a history of LMs and the adverse outcomes during the subsequent pregnancy. Notably, we observed that women with a previous LM caused by cervical incompetence, trauma and unexplained factors had a significantly higher risk of recurrent LM, which was in accordance with prior studies (3, 12, 13). A history of an unLM was also significantly associated with an elevated EM rate and a ceLM was associated with a higher preterm birth rate in subsequent embryo transfer. More importantly, the live birth rate was resultantly decreased in both unLM and ceLM group. In addition, the pregnancy outcomes were similar between feLM group and general IVF group, which suggesting that one LM because of fetal abnormality was more likely to be accidental and would not conspicuously influence the subsequent pregnancy outcome.