Cardiovasc Diabetol. 2026 Jan 20. doi: 10.1186/s12933-025-03066-z. Online ahead of print.
ABSTRACT
BACKGROUND: Triglyceride-glucose (TyG)-related indices, which have emerged as promising prognostic markers in multiple vulnerable populations, are closely linked to the incidence and progression of cardiovascular disease (CVD). Nevertheless, their associations with CVD morbidity and mortality, as well as the underlying biological mechanisms, remain unclear in pre-frail or frail populations. We aimed to evaluate the associations of multiple TyG-related indices with incident CVD and cardiovascular mortality among pre-frail or frail individuals, and to explore potential mediating biological pathways.
METHODS: In this prospective cohort analysis using UK Biobank data, 136,004 pre-frail or frail individuals [median age (interquartile range): 57.0 (49.0, 63.0) years; 54.4% female] without baseline CVD were identified. Six composite TyG-related indices integrating anthropometric measures [including TyG-body mass index (BMI), waist circumference (WC), waist-to-height ratio (WHtR), body roundness index (BRI), a body shape index (ABSI), and weight-adjusted waist index (WWI)] were evaluated to assess their associations with CVD morbidity and mortality via Cox proportional hazards models, with findings reported as hazard ratios (HRs) and 95% confidence intervals (CIs). The net reclassification index (NRI), integrated discrimination improvement index (IDI), and C-index were utilized to assess the incremental predictive value of these indices. Mediation analyses were further conducted to quantify the contribution of biomarkers to the observed associations.
RESULTS: During a median follow-up of 13.3 years, 23,494 participants developed CVD, and 2453 deaths were attributed to CVD. All TyG-related indices were positively associated with incident CVD among individuals with pre-frailty or frailty. In fully adjusted models, HRs with 95% CIs for participants in the highest versus lowest quartiles were 1.36 (1.30-1.42) for TyG-BMI, 1.48 (1.42-1.56) for TyG-WC, 1.36 (1.30-1.42) for TyG-WHtR, 1.40 (1.34-1.46) for TyG-BRI, 1.24 (1.18-1.30) for TyG-ABSI, and 1.28 (1.22-1.34) for TyG-WWI. Comparable associations were observed for cardiovascular mortality. All indices demonstrated linear associations with CVD incidence, except TyG-BMI and TyG-ABSI, which exhibited nonlinear dose-response relationships (P for nonlinearity < 0.01). In predictive analyses, the NRI, IDI, and C-index were significantly improved for all indices, with TyG-WC and TyG-BRI exhibiting the strongest predictive performance. Biomarkers reflecting inflammation, liver function, and kidney function partially mediated these associations. Key mediators, including C-reactive protein, albumin, cystatin C, and urate, accounted for more than 10% of the observed effects.
CONCLUSION: TyG-related indices, particularly TyG-WC and TyG-BRI, were independently associated with cardiovascular morbidity and mortality in pre-frail or frail individuals. These indices may serve as simple, scalable tools for identifying cardiovascular risk in pre-frail or frail individuals. Mediation analyses highlight potential avenues for targeted interventions aimed at reducing cardiovascular risk.
PMID:41559713 | DOI:10.1186/s12933-025-03066-z