Correlation Between Various Insulin Resistance Indices and Sarcopenia: Evidence from CHARLS

Background Evidence on the association between surrogate insulin resistance (IR) indices and sarcopenia (SA) remains limited. Seven alternative indicators for assessing insulin resistance were included in this study, primarily categorized into three groups: lipid ratios (TG/HDL); the TyG index and its derivatives combined with body composition (TyG-WC, TyG-BMI, TyG-WHtR); indicators reflecting fat distribution (LAP), and Chinese visceral adiposity index (CVAI)—in relation to SA risk among Chinese adults. Methods Longitudinal data from the China Health and Retirement Longitudinal Study (CHARLS) was used, with information collected between 2011 and 2015., which included 5,329 participants. Multivariable logistic regression assessed associations between each index and incident SA. Results Over a 4-year period, 330 participants (6.2%) developed SA. In fully adjusted models (Model 3), all seven IR surrogates were significantly associated with SA. Per standard-deviation increase, adjusted odds ratios (OR, 95% CI) were: TG/HDL 0.84 (0.74–0.95), TyG 0.65 (0.53–0.80), TyG-BMI 0.14 (0.10–0.19), TyG-WHtR 0.53 (0.45–0.61), TyG-WC 0.50 (0.43–0.57), CVAI 0.41 (0.35–0.48), and LAP 0.60 (0.51–0.70). Notably, all indices except LAP showed evidence of nonlinearity in relation to SA risk. Additionally, TyG-BMI yielded the largest AUC among the seven indices. Conclusions TG/HDL, TyG, TyG-WC, TyG-BMI, TyG-WHtR, LAP, and CVAI were inversely associated with incident SA. TG/HDL, TyG, TyG-WC, TyG-BMI, TyG-WHtR, and CVAI exhibited nonlinear relationships with SA. TyG-BMI may be the most informative predictor of SA.