This invention, through lipidomics and clinical sample validation, found that prospective follow-up of pre-
pregnancy serum samples from women with a history of recurrent
pregnancy loss (RPL) revealed that the pre-
pregnancy serum levels of
phosphatidylinositol PI (18:1 / 18:1) (i.e., PI 36:2),
phosphatidylcholine (29:0-6:0) (i.e., PC 35:0),
phosphatidylethanolamine (18:1p-18:2) (i.e., PE 36:3p), triglycerides (28:1-10:4-11:2) (i.e., TG 49:7), triglycerides (14:0-18:2-20:5) (i.e., TG 52:7), and
wax esters (22:1) (i.e., WE 22:1) were significantly lower in these women than in those who subsequently achieved successful pregnancies. This confirms that the pre-pregnancy serum levels of these six lipids (PI 36:2, PC 35:0, PE 36:3p, TG 49:7, TG 52:7, WE 22:1) were significantly lower in these women than in those who subsequently achieved successful pregnancies. Decreased levels of 22:1 lipids were significantly associated with the risk of
miscarriage in subsequent pregnancies. The
area under the curve (AUC) of the predictive model constructed by combining six lipids reached 0.85, demonstrating good predictive
efficacy. Based on these findings, this lipid combination can serve as a reliable predictor of
miscarriage risk in subsequent pregnancies in RPL patients and can be used to develop diagnostic reagents or kits for preconception
risk screening.