Altered lipid profile in uterine leiomyoma: a focus on apolipoprotein A1 reduction and machine learning-based predictive modeling.

Objective: This study aimed to investigate the association between uterine leiomyomas (UL) and specific alterations in serum lipid profiles, and to evaluate the performance of machine learning models incorporating these markers for UL discrimination. Methods: In this age- and body mass index matched case-control study, 200 patients with histologically confirmed UL and 200 controls with normal uteri were enrolled. Fasting serum levels of total cholesterol, triglycerides, low-density lipoprotein, high-density lipoprotein, apolipoprotein A1 (ApoA1), and apolipoprotein B were measured. Logistic regression identified independent risk factors, which were then used to construct predictive models via several machine learning algorithms. Model performance was assessed using receiver operating characteristic curve analysis. Results: Patients with UL exhibited significantly lower serum levels of triglycerides and ApoA1 compared to controls. Multivariate analysis confirmed lower triglyceride and ApoA1 levels, along with higher gravidity and premenopausal status, as independent factors associated with UL. While individual lipid parameters showed limited discriminative power, integrative models combining these with clinical features achieved high performance. The Random Forest model demonstrated superior discriminative ability, with an area under the curve of 0.986. Conclusion: After rigorous confounding control, UL is independently associated with a distinct metabolic phenotype characterized by reduced serum triglyceride and ApoA1 levels. Prediction models integrating these lipid abnormalities with clinical data show promising potential for risk assessment, highlighting a unique interplay between lipid metabolism and UL pathogenesis worthy of further investigation.