METS-IR predicts new-onset stroke in older adults with stages 0-3 cardiovascular-kidney-metabolic syndrome: a prospective, multicohort, clinical study with statistical and machine learning.

BACKGROUND: Cardiovascular-Kidney-Metabolic (CKM) syndrome is a systemic disorder characterized by the complex interplay of chronic kidney disease, cardiovascular disease, and metabolic abnormalities. The metabolic score for insulin resistance (METS-IR) has emerged as a validated surrogate for IR assessment. This study aims to undertake a cross-national analysis to explore the association between METS-IR and stroke risk in older adults with CKM syndrome stages 0-3. METHODS: This study employed data from the China Health and Retirement Longitudinal Study (CHARLS) and the National Health and Nutrition Examination Survey (NHANES), which incorporated multiple general health examination datasets. The external validation cohort was recruited from Shengjing Hospital of China Medical University between 2011 and 2020. RESULTS: Cox regression analysis revealed a significant association between elevated METS-IR scores and stroke risk in older adults with stages 0-3 CKM syndrome, with consistent results validated in the clinical cohort. Restricted cubic spline analysis showed a nonlinear association between METS-IR and stroke incidence in specific models. Sensitivity and subgroup analyses demonstrated consistent findings, thus confirming the robustness of the results. Moreover, K-means clustering coupled with logistic regression identified persistently high METS-IR levels as an independent stroke risk factor. Machine learning algorithms (Logistic Regression, XGBoost, and Random Forest) validated these findings, confirming METS-IR's predictive utility for stroke in adults with stages 0-3 CKM syndrome. CONCLUSION: A significant association exists between elevated METS-IR and new-onset stroke in older adults with stages 0-3 CKM syndrome.