Chunyan Xu ^1,2 Ling He ^1,2 Yansong Tu ³ Canhui Guo ^1,2 Hurong Lai ^1,2 Caifeng Liao ^1,2 Chuyang Lin ² Huaijun Tu ^1,2*

• ¹ Graduate School, Jiangxi Medical College, Nanchang University, Nanchang, Jiangxi Province, China
• ² The Department of Geriatrics, The Second Affiliated Hospital of Nanchang University, Nanchang, Jiangxi Province, China
• ³ Bio21 Molecular Science and Biotechnology Institute, Faculty of Science, University of Melbourne, Parkville, Victoria, Australia

Objective: The relationship between surrogate insulin resistance (IR) indices and sarcopenic obesity (SO) remains unclear. This study aimed to assess the association of six IR surrogates—triglyceride‐glucose (TyG), TyG-waist circumference (TyG-WC), TyG-waist-to-height ratio (TyG-WHtR), triglyceride-to-high-density-lipoprotein-cholesterol ratio (TG/HDL), metabolic score for insulin resistance (METS-IR), and Chinese visceral adiposity index (CVAI)—with SO risk in a middle-aged and older population in China. Methods: The study employed longitudinal data from the China Health and Retirement Longitudinal Study (CHARLS) between 2011 and 2015 with 6,395 participants. We applied multivariate logistic regression models to explore the link between six surrogates and SO. Nonlinear relationships were evaluated using restricted cubic spline analysis, and subgroup analyses were conducted for validation. Receiver operating characteristic (ROC) curves were used to assess predictive capabilities. Results: Over the course of a 4-year follow-up period, 319 participants (5.0%) developed SO. All six surrogates were significantly associated with SO in the fully adjusted model. The adjusted odds ratios (ORs) with the 95% confidence interval (95% CI) per standard deviation increase were 1.21 (1.08-1.36) for TyG, 1.56 (1.39-1.75) for TyG-WC, 2.04 (1.81-2.31) for TyG-WHtR, 1.11 (1.01-1.21) for TG/HDL, 1.67 (1.50-1.87) for METS-IR, and 1.74 (1.55-1.97) for CVAI. Notably, TyG-WC, TyG-WHtR, TG/HDL, METS-IR, and CVAI reflected nonlinear relationships with SO. Conversely, TG/HDL did not exhibit a significant association in subgroup analysis. Furthermore, TyG-WHtR had a significantly larger area under the receiver operating characteristic curve than other indices. Conclusions: The results indicated that TyG, TyG-WC, TyG-WHtR, METS-IR, and CVAI were significantly and positively associated with SO incidence. Meanwhile, TyG-WC, TyG-WHtR, METS-IR, and CVAI exhibited nonlinear relationships with SO. Specifically, TyG-WHtR may be the most suitable indicator for predicting SO among middle-aged and older Chinese adults.