Archives of Gerontology and Geriatrics · Feb 3, 2026

Threshold divergence in triglyceride-glucose index and mortality among Chinese older adults across the glucose metabolism spectrum: a cohort study

By Yu L, Yan F, Xu J et al.

BACKGROUND: The triglyceride-glucose (TyG) index is a promising marker of insulin resistance, but its association with mortality in older adults across different glucose metabolism statuses remain unclear. METHODS: This prospective cohort study utilized data from 708,456 adults aged ≥ 65 in the Chinese national Basic Public Health Service program across 4 municipalities (2017-2018). Participants were categorized into three groups by glucose metabolism status: diabetes, prediabetes and non-diabetes. With follow-up until December 31, 2023, we further analysed the association between baseline TyG index and five outcomes in each group: all-cause, cardiovascular disease (CVD), cancer, respiratory disease, and endocrine, nutritional, and metabolic diseases mortality, using Cox proportional hazards regression model and restricted cubic spline. RESULTS: During a median 6.38-year follow-up, 124,945 deaths occurred. A higher TyG index was significantly associated with increased all-cause mortality risk across all glucose metabolism strata. Compared to the lowest quartile, the highest quartile had hazard ratios of 1.42 (95% CI: 1.37 - 1.47) for diabetes, 1.17 (95% CI: 1.11 - 1.23) for prediabetes, and 1.14 (95% CI: 1.12 - 1.17) for non-diabetes populations. Similarly graded increase risk with worsening glucose metabolism were observed for CVD mortality, and respiratory disease mortality. Nonlinear analysis revealed distinct risk patterns: a single high-risk threshold (TyG=8.51) in diabetes, versus dual-threshold "risk windows" in prediabetes and non-diabetes groups. CONCLUSIONS: The TyG index effectively stratifies mortality risk in older adults. Risk interpretation requires consideration of individual glucose metabolism status, enabling tailored preventive strategies for the aging population.

doi.org/10.1016/j.archger.2026.106162PubMed 41719814Biomarkers