AUTHOR=Huang Shulin , Hu Xudong , Hu Lingjuan , Du Tingting , Zhao Yanchun , Xiao Jing 

TITLE=Development of a clinical diagnostic model for atypical endometrial hyperplasia and endometrial carcinoma in women aged 40–60 years based on refined abnormal uterine bleeding patterns

JOURNAL=Frontiers in Oncology

VOLUME=Volume 15 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2025.1684604

DOI=10.3389/fonc.2025.1684604

ISSN=2234-943X

ABSTRACT=ObjectiveTo develop and validate a diagnostic nomogram model for predicting atypical endometrial hyperplasia (AEH) and endometrial carcinoma (EC) in women aged 40–60 years with abnormal uterine bleeding (AUB), incorporating detailed bleeding patterns and clinical parameters.MethodsThis retrospective cohort study included 1,920 patients aged 40–60 years with AUB who underwent hysteroscopic evaluation across four hospital branches from 2021 to 2024. Variables were screened via univariate logistic regression followed by LASSO regression. A multivariate logistic regression model was constructed using seven selected predictors, including age, family history of cancer, endometrial thickness, menstrual blood loss, abnormal menstrual duration, intermenstrual bleeding, and postmenopausal bleeding. Internal validation was performed using bootstrap resampling; external validation was conducted using an independent cohort.ResultsThe final model demonstrated good discriminatory performance (AUC = 0.814 in the training cohort, 0.762 in external validation). Calibration plots and the Hosmer–Lemeshow test indicated good agreement between predicted and observed probabilities. Decision curve analysis confirmed favorable clinical utility. Patients classified as high-risk (score ≥70.362) had a significantly increased likelihood of AEH/EC (OR = 12.46, 95% CI: 7.56–21.01, P < 0.001).ConclusionThis study presents a validated, user-friendly nomogram integrating refined AUB patterns and clinical variables to support early risk stratification for AEH/EC in women aged 40–60 years. The model demonstrates robust predictive performance and may assist in guiding individualized diagnostic strategies, particularly in resource-limited settings.