Parental determinants of blastocyst euploidy in a predominantly recurrent-pregnancy-loss PGT-A cohort: limited predictive value of female age, BMI, AMH and sperm DNA fragmentation

Abstract

Background Embryonic aneuploidy is a major factor affecting clinical outcomes in assisted reproduction. PGT-A improves embryo selection, but the clinical factors associated with euploidy remain debated. Methods This retrospective study included 588 PGT‑A cycles. We used embryo‑level GLMMs to evaluate the associations of female age, female BMI, AMH, male age, male BMI, and sperm DFI with embryo euploidy. We performed univariable and multivariable analyses, assessed model performance (AUC, R 2 , calibration), and conducted internal validation via patient-clustered bootstrap (200 replicates). Subgroup analyses by female age were performed with FDR correction. Results Among 2,357 blastocysts, 1,299 (55.1%) were euploid. In univariable analysis, female age, male age, and male DFI were significantly associated with euploidy. After multivariable adjustment, only female age (OR = 0.87, p = 0.011) and male DFI (OR = 0.89, p = 0.006) remained independent predictors. Model performance was poor (apparent AUC = 0.571; bootstrap-corrected AUC = 0.574; apparent marginal R 2 = 0.018 and conditional R 2 = 0.036; apparent calibration slope = 0.985; bootstrap-corrected calibration slope = 0.913), indicating weak signal strength and limited generalizability. Subgroup analyses showed some nominal associations, but none survived FDR correction (all q > 0.05). Among 588 cycles with complete covariate data, 90.3% had at least one euploid blastocyst; among all 613 biopsied cycles and 637 started cycles, the rates were 89.89% and 86.50%, respectively. Conclusion Female age and male DFI were independently associated with euploidy, but the model’s discrimination remained poor after optimism correction (bootstrap-corrected AUC = 0.574; apparent calibration slope = 0.985, bootstrap-corrected calibration slope = 0.913), and these variables therefore lack sufficient value for clinical use. Subgroup analyses by female age were exploratory and hypothesis-generating, and no statistically significant age-modified association between male DFI and euploid embryos persisted after FDR correction. Future models incorporating embryo-specific data and larger samples are needed.

The authors' abstract, as published at the source. Frontiers in Endocrinology, 2026 · DOI ↗