Abstract
The dehydroepiandrosterone (DHEA)-induced mouse model is widely used to study polycystic ovary syndrome (PCOS), yet findings on antral follicle (AF) dynamics remain inconsistent. Among 12 representative studies-classified as mechanistic (n = 1), intervention (n = 7), or mixed-type (n = 4)-half reported increased AF counts, while the others reported reductions. This variability likely stems from methodological differences, including DHEA dosage, solvent, animal age, treatment duration, follicle classification, and estrous cycle control. AMH, a key regulator of folliculogenesis, was reported in only one study, limiting mechanistic interpretation. Our subgroup and forest plot analyses suggest that ethanol solvents, younger animals, and longer treatment durations may increase the likelihood of AF elevation, though none reached statistical significance and these results should be interpreted as exploratory due to small sample sizes and study heterogeneity. We propose standardizing key variables such as AMH measurement, follicle classification, histological sectioning, estrous cycle staging, and DHEA administration protocols. Although the DHEA model recapitulates reproductive hallmarks of PCOS-such as hyperandrogenism and follicular arrest-it incompletely reflects the metabolic and neuroendocrine features of human PCOS, limiting its translational fidelity. These findings highlight the need for greater methodological transparency and standardization to improve the translational value of preclinical PCOS models.