Abstract:Objective To investigate the differences in the predictive efficacies of 4 surrogate indicators of insulin resistance (IR) for incident cardiometabolic multimorbidity (CMM). Methods A longitudinal study was conducted using data from individuals aged≥45 years in the China Health and Retirement Longitudinal Study (CHARLS) from 2011 to 2020. Multivariable logistic regression models were used to explore the associations between 4 IR surrogate indicators-triglyceride-glucose index (TyG), lipid accumulation product index (LAP), metabolic score for insulin resistance (METS-IR), and metabolic score for visceral fat (METS-VF)-and CMM. Restricted cubic spline (RCS) models were employed to examine potential non-linear associations. The predictive performance of each indicator was assessed using receiver operating characteristic (ROC) curves and the area under curve (AUC). Results A total of 6 100 participants were enrolled in this study, and 649 incident cases of CMM were recorded during the 9-year follow-up. The multivariable logistic regression results showed that for each increase of 1 interquartile range (IQR) in TyG, LAP, METS-IR, and METS-VF, the risk of CMM increased by 50.8% (odds ratio [OR]=1.508, 95% confidence interval [95%CI] 1.364-1.667), 11.3% (OR=1.113, 95%CI 1.054-1.176), 37.6% (OR=1.376, 95%CI 1.247-1.518), and 190.7% (OR=2.907, 95%CI 1.826-4.629), respectively. TyG and METS-VF had linear relationships with CMM risk (Poverall<0.001, Pnon-linear>0.05), while LAP and METS-IR had non-linear relationships with the risk of CMM (Poverall<0.001, Pnon-linear<0.001). LAP had the highest predictive value for CMM, with an AUC of 0.669 (95%CI 0.648-0.690), followed by METS-IR, TyG, and METS-VF. Conclusion TyG, LAP, METS-IR, and METS-VF are all closely related to the risk of incident CMM and have certain predictive value. Among them, LAP has the best predictive efficacy, and it is also cost-effective and easily accessible, making it suitable for preliminary screening of high-risk populations for CMM in clinical settings.