Metabolic Reprogramming in Recurrent Spontaneous Abortion: Key Biomarkers Identification and Diagnostic Model Development.

Recurrent spontaneous abortion (RSA) involves complex and often unexplained pathogenesis. This study investigated metabolic reprogramming in RSA to identify diagnostic biomarkers. We integrated GEO datasets (GSE26787, GSE165004), identified metabolic reprogramming-related differentially expressed genes, and intersected them with WGCNA modules. Module genes were refined using logistic regression, SVM-RFE, and LASSO to construct a diagnostic model validated by ROC, nomogram, calibration, and decision curve analysis. The role of SREBF2 was explored by PCR, western blot, CCK-8, and transwell assays in vitro. Immune infiltration (ssGSEA, CIBERSORT) and single-cell RNA sequencing (GSE214607) explored immune-metabolic crosstalk. Results demonstrate critical involvement of metabolic reprogramming in RSA. Four model genes (SREBF2, PPARG, SQLE, UCP2) were upregulated in RSA and enriched in lipid/cholesterol pathways. Experimental validation confirmed SREBF2 upregulation in RSA decidua. Its overexpression in endometrial stromal cells impaired cell proliferation, decidualization, and trophoblast migration. Immune infiltration revealed correlations between model genes and specific immune subsets. Single-cell analysis identified altered immune cell proportions and cell-type-specific expression patterns, predominantly in macrophage subsets. This study indicates that lipid-centric metabolic remodelling may concurrently affect decidual function, trophoblast behaviour, and immune tolerance in RSA, offering diagnostic and therapeutic insights.