Frontiers in Endocrinology · Oct 30, 2026
Revisiting common female infertility factors traditionally associated with impaired oocyte quality: insights from IVF outcomes and AI-derived oocyte score performance
By Maite del Collado, Daniela Paes de Almeida Ferreira Braga, Amanda Setti et al.
This retrospective cohort study evaluated whether common female infertility factors are associated with oocyte and embryo competence and cumulative reproductive outcomes in intracytoplasmic sperm injection (ICSI) cycles, and assessed the performance of an AI-derived composite oocyte score across these populations. The study included 14,420 injected oocytes from 1,856 autologous ICSI cycles performed between 2020 and 2025. Outcomes were compared between controls and women with polyendocrine metabolic ovarian syndrome (PMOS), endometriosis, overweight, or obesity; controls were additionally stratified by maternal age (<38 versus ≥38 years). Age-based propensity-score weighting was applied for pairwise female-factor comparisons, followed by adjusted mixed-effects regression models accounting for patient-and retrieval-cycle-level clustering. Endometriosis and obesity were not associated with consistent reductions in laboratory or clinical outcomes, whereas overweight was associated with lower top-quality blastocyst and cumulative reproductive outcomes. Advanced maternal age showed the most consistent detrimental pattern, including lower MAGENTA and KIDScore values, reduced blastocyst development and euploidy rates, and lower cumulative pregnancy rates. PMOS cycles showed largely preserved laboratory outcomes, higher MII yield, and higher observed cumulative pregnancy rates. The composite MAGENTA score was associated with higher odds of fertilization (OR 1.53), blastocyst formation (OR 1.67), and top-quality blastocyst formation (OR 1.26), and with higher KIDScore (β = 0.21). These associations were generally observed across the evaluated female-factor groups, and MAGENTA provided modest incremental discrimination beyond clinical variables. Overall, female infertility factors appear to affect reproductive competence through distinct patterns rather than a uniform impairment in oocyte quality, with maternal age showing the most consistent adverse association and AI-derived oocyte assessment providing complementary predictive information across heterogeneous populations.