Pretreatment AMH-AFC stratification of poor-ovarian-response risk in young women with diminished ovarian reserve: a multicentre retrospective cohort study.

Authors

Yanning Lv, Jiahao Fang, Haiyan Yang, Jianying He, Lihua Xia, Honglian Wu

Published in

Frontiers in endocrinology. Volume 17. Pages 1968836. Epub Sep 28, 2026.

Abstract

In women aged 35 years or younger with diminished ovarian reserve (DOR), we evaluated whether two pretreatment markers, anti-Müllerian hormone (AMH) and antral follicle count (AFC), can stratify the risk of poor ovarian response (POR) before stimulation. This retrospective cohort included IVF/ICSI cycles from four centres in Zhejiang, China (2022-2025). Of 1,754 DOR-eligible cycles, 157 were removed in two post-hoc phenotype-cleaning exclusions (PCOS, n = 139; AMH ≥5 ng/mL without PCOS, n = 18), leaving 1,597 cycles. POR was defined as ≤3 oocytes. Cycles were cross-classified by AFC ≤5 and AMH <0.5 ng/mL, and a two-marker logistic model was fitted. POR occurred in 42.0% of cycles. Total AFC (adjusted OR 0.843) and ln (AMH) (OR 0.310) carried the main pretreatment signal; age and basal FSH added little. The two single-marker cells had similar observed POR rates (46.4% vs 49.6%; P = 0.56), but equivalence was not established. The AFC ≤5 boundary was reproduced in the seed-2024 70:30 split and was re-selected in 92.0% of 1,000 bootstrap samples when the cut-off was chosen in-bag and evaluated out-of-bag. The descriptive three-tier rates were 23.9%, 47.1%, and 78.8%; when all 1,754 eligible cycles were retained, the middle-tier rate fell to 39.5%. For the fixed two-marker specification, bootstrap optimism correction gave an AUC of 0.768; centre-held-out AUCs using verified record-level centre labels ranged from 0.725 to 0.800. AMH and AFC therefore provide moderate pretreatment risk stratification within this young DOR cohort, but the absolute tier risks remain sensitive to cohort definition and cut-off choice. External validation and recalibration are required before clinical use.

PMID:
42865567
Bibliographic data and abstract were imported from PubMed on 13 Oct 2026.

Read full publication at:
Please sign in to see all details.

Stats

1-terrible, 9-excellent. How would you rate this publication? Sign in in to submit your rating.

Post a comment

You need to be signed in to post comments. You can sign in here.

Comments

There are no comments yet.