Evaluative performance of triglyceride-glucose-frailty index versus surrogate insulin resistance indices for advanced cardiovascular-kidney-metabolic syndrome, cardiometabolic multimorbidity, and mortality: the role of estimated glomerular filtration rate

Abstract

Cardiovascular-kidney-metabolic (CKM) syndrome and cardiometabolic multimorbidity (CMM) carry substantial mortality, yet whether a composite index integrating insulin resistance and frailty shows stronger associations with advanced CKM syndrome, CMM, and mortality than conventional metabolic markers remains unclear. This study evaluated the associations of the triglyceride-glucose-frailty index (TyG-FI) with CKM, CMM, and all-cause and cardiovascular mortality, and explored the role of baseline estimated glomerular filtration rate (eGFR). The analysis included 11,228 adults aged 20–79 years from the National Health and Nutrition Examination Survey 2001–2018. TyG-FI was calculated as the TyG index multiplied by the frailty index. Survey-weighted logistic and Cox proportional-hazards models were used to estimate odds and hazard ratios, with sequential adjustment for demographic, socioeconomic, and behavioral confounders. Restricted cubic splines examined non-linear relationships. Overall model performance was assessed using receiver operating characteristic curves with DeLong’s test, time-dependent AUC, Harrell’s C-index, calibration, and decision curve analysis. Exploratory mediation analyses quantified the proportion of mortality associations accounted for by baseline eGFR. Robustness was verified through multiple sensitivity and subgroup analyses. Over a median follow-up of 90 months (1,220 all-cause and 379 cardiovascular deaths), higher TyG-FI quartiles were associated with graded decreases in survival. Compared with the lowest TyG-FI quartile, the highest quartile yielded markedly elevated odds for advanced CKM syndrome (OR = 4.57, 95% CI: 3.50–5.99) and CMM (OR = 5.09, 95% CI: 2.83–9.16), and higher hazard for all-cause (HR = 3.90, 95% CI: 2.96–5.13) and cardiovascular death (HR = 5.82, 95% CI: 3.38–10.03). Among participants with CMM, the relationship remained for all-cause death (HR = 2.69, 95% CI: 1.69–4.28). The dose–response patterns were non-linear for outcomes (P for non-linearity < 0.001). TyG-FI demonstrated higher discrimination than most surrogate insulin resistance indices, and lower baseline eGFR partly accounted for 16.45% to 26.74% of the associations with mortality. Sensitivity analyses corroborated the consistency of all findings. TyG-FI, which captures both metabolic dysfunction and physiological frailty, showed stronger associations with advanced CKM syndrome, prevalent CMM, and mortality than most surrogate insulin resistance indices. In exploratory analyses, lower baseline eGFR partly accounted for a proportion of the association with mortality. TyG-FI may offer a pragmatic tool for early mortality risk stratification in individuals with high cardiometabolic risk.