Background: Prior research indicates a potential link between dyslipidemia and endometriosis (EMs). However, the relationship between remnant cholesterol (RC) and EMs has not been thoroughly investigated. Consequently, looking into and clarifying the connection between RC and EMs was the primary goal of this study.
Methods: Following the screening of participants from the NHANES dataset spanning 2001 to 2006, a total of 1,840 individuals were incorporated into this research. A weighted multivariable logistic regression analysis was first performed to investigate the relation between RC and the likelihood of encountering EMs. To assess the degree of consistency in the link between RC and EMs across different populations, additional subgroup analyses were performed. In addition, the researchers used the extreme gradient boosting (XGBoost) technique and the area under the receiver operating characteristic curve (ROC) to evaluate how well RC recognized EMs. Lastly, both linear and nonlinear relationships were validated using generalized additive models (GAM), while dose-response connections were investigated through restricted cubic spline models.
Results: After accounting for all potential confounders, a strong correlation between RC and EMs was identified. In particular, an increase of one unit in RC was linked to a 135% rise in the likelihood of developing EMs. Analyses of subgroups revealed that these relationships remained stable across the majority of subgroups (interaction P-value > 0.05). Multivariable logistic regression demonstrated RC's independent predictive value, maintaining statistical significance after adjusting for confounders. The AUC of 0.614 suggests RC's moderate ability to discriminate EMs, outperforming traditional markers like LDL-C in sensitivity and specificity. Furthermore, XGBoost analysis identified RC as the most critical predictor among lipid-related and demographic variables. The relationship was further validated through GAM, which visually confirmed a linear trend, and RCS, which provided statistical evidence of linearity.
Conclusion: This study reveals a clear connection between RC and the likelihood of having EMs within the US population, suggesting RC as a potential marker for further investigation in understanding endometriosis risk.
Keywords: Cross-sectional study; Endometriosis; Extreme gradient boosting; Generalized additive model; NHANES; Receiver operating characteristic curve; Remnant cholesterol.
© 2025. The Author(s).