Triglyceride-glucose-related indices and cardiovascular disease and mortality among individuals with depression: a prospective study from UK Biobank

Background The triglyceride-glucose (TyG) index reflects insulin resistance-related metabolic dysfunction and, when combined with anthropometric measures, may capture adiposity-related cardiovascular risk information. Although associated with depression, cardiovascular disease (CVD), and mortality, their associations with incident CVD and mortality among individuals with depression remain unclear. This study examined these associations and explored potential biological correlates. Methods This prospective cohort study included 53,171 UK Biobank participants with depression and free of CVD at baseline. Seven TyG-related indices were derived: TyG, TyG-body mass index (TyG-BMI), TyG-waist circumference (TyG-WC), TyG-waist-to-height ratio (TyG-WHtR), TyG-a body shape index (TyG-ABSI), TyG-weight-adjusted waist index (TyG-WWI), and TyG-body roundness index (TyG-BRI). Associations were evaluated using Cox regression models. Incremental predictive information beyond a conventional cardiovascular risk model was assessed using several prediction metrics. Exploratory biomarker analyses identified potential biological correlates. Results Over a median follow-up of 13.4 years, 8,516 incident CVD cases, 546 CVD-related deaths, and 4,619 all-cause deaths were documented. For incident CVD, multivariable-adjusted hazard ratios (95% confidence intervals) comparing the highest with the lowest quartile were 1.40 (1.31–1.49) for TyG, 1.80 (1.68–1.92) for TyG-BMI, 1.95 (1.82–2.09) for TyG-WC, 1.89 (1.76–2.03) for TyG-WHtR, 1.77 (1.65–1.90) for TyG-WWI, 1.93 (1.80–2.08) for TyG-BRI, and 1.61 (1.49–1.73) for TyG-ABSI. Each 1-SD increment in these indices was associated with a 24–37% higher risk of CVD mortality and a 7–16% higher risk of all-cause mortality. The associations with incident CVD and CVD mortality showed no significant departure from linearity. Except for TyG and TyG-BMI, adding the other indices to the conventional risk model yielded statistically significant but modest incremental predictive information, with small absolute improvements in prediction metrics. Biomarkers reflecting inflammation (e.g., neutrophils: 6.5–13.1%), metabolism (glycated hemoglobin: 6.5–23.9%), and hepatic (gamma-glutamyltransferase: 5.0–12.3%) and renal (urate: 8.6–25.8%) function statistically accounted for part of their associations with incident CVD. Conclusion Higher TyG-related indices were associated with increased risks of incident CVD and mortality among individuals with depression. Several indices provided modest incremental predictive information, but external validation and implementation-focused studies are needed before clinical application. Inflammatory, metabolic, hepatic, and renal biomarkers may represent candidate biological correlates, although causal mediation cannot be inferred.