Disease burden, machine learning‑based risk factor identification in uterine cancer: A cross‑sectional and bibliometric analysis

Uterine cancer (UC) is the most common malignancy of the female reproductive system worldwide, with a continuously rising incidence, posing a significant public health challenge. This study systematically evaluated the burden of UC using the Global Burden of Disease (GBD) database. Combined with data from the National Health and Nutrition Examination Survey (NHANES), machine learning (ML) methods – including the Boruta algorithm and 11 predictive models – were employed to identify key risk factors associated with UC. Furthermore, a bibliometric analysis of 2041 relevant publications from 2005 to 2025 in the Web of Science Core Collection was conducted to reveal research hotspots and developmental trends in the field. In 2023, the cumulative incidence and prevalence of UC were highest in the United States, Russia, Italy, and Portugal. The age group 65 to 69 years exhibited the highest incidence and disability-adjusted life years. LightGBM (AUC train set: 1, AUC test set: 0.929), XGBoost (AUC train set: 1, AUC test set: 0.952), and random forest (AUC train set: 1, AUC test set: 0.964) performed well, which identified critical risk factors including molybdenum, triglycerides, cadmium, age, lead, total cholesterol, cobalt, hypertension, uric acid, and mean arterial pressure. Bibliometric analysis indicated that the United States leads in research output, with keywords such as “multicenter,” “endometrial neoplasms,” “disparity,” “race,” and “guideline” representing current research frontiers. This study provides a comprehensive assessment of the disease burden of UC, reveals multi‑factor risk profiles through ML, and outlines evolving research trends, offering evidence‑based insights for targeted prevention strategies, clinical intervention optimization, and resource allocation.