Comment on “AMH predicts miscarriage in non-PCOS but not in PCOS related infertility ART cycles”

Arkfeld et al. reported that AMH predicts miscarriage in non-PCOS but not in PCOS ART cycles. We raise concerns regarding the biological plausibility and data interpretation of their findings. Extremely low AMH (<1 ng/mL) is physiologically uncommon in PCOS, suggesting potential misclassification or temporal mismatch in measurements. While the original study attributed the null finding to generally elevated AMH in PCOS (a dilution effect), our concern is distinct: biologically discordant AMH <1 ng/mL values in PCOS-labeled patients may distort subgroup estimates and artificially produce a null association. Additionally, the SART CORS database uses heterogeneous, clinic-reported PCOS diagnoses, and sample sizes for low-AMH PCOS patients were not reported. These factors may explain the apparent lack of association, highlighting the need for cautious interpretation and sensitivity analyses using standardized criteria.