Inflammation at the crossroads of reproduction: SIRI as a prognostic signature of female infertility in hybrid regression–machine learning models

BACKGROUND: Female infertility is a critical global health concern, with a rising prevalence and significant psychosocial consequences. This study aimed to investigate the association between Systemic Inflammatory Response Index (SIRI) and female infertility. METHODS: This cross-sectional study enrolled 3059 reproductive-aged women (18-45 years) to examine the association between SIRI and female infertility using data from the National Health and Nutrition Examination Survey (NHANES) spanning from 2015 to 2020. Multivariable logistic regression generalized additive models (GAM), restricted cubic splines (RCS), and threshold effect analyses were leveraged. Machine learning approaches were also utilized to validate predictive performance and identify key features. RESULTS: Elevated SIRI was independently associated with increased odds of infertility. In the fully adjusted logistic model, each unit increase in SIRI corresponded to a 34% increase in infertility risk (OR 1.34, p = 0.001). Women in the highest SIRI quartile had more than double the odds of infertility compared to those in the lowest quartile (OR 2.08, p < 0.001), with a significant dose-response trend (p trend < 0.001). GAM and RCS models confirmed a monotonic and curvilinear association, respectively. Threshold analysis revealed a critical inflection point at SIRI = 1.66. Machine learning validation identified SIRI as one of the most influential predictors, with XGBoost achieving the highest (AUC = 0.866). CONCLUSION: These findings support the role of chronic systemic inflammation in female infertility and highlight SIRI as a valuable biomarker for risk prediction and clinical assessment.