Divergent mortality associations of TyG-BMI in middle-aged and older adults: Parallel analyses of Chinese and US populations.
TL;DR
The association between TyG-BMI and all-cause mortality differed between Chinese and US middle-aged and older adults, and TyG-BMI provided little or no incremental prognostic discrimination beyond BMI alone, which supports cautious, population-specific interpretation of TyG-BMI.
Abstract
Background
To compare the association between TyG-BMI and all-cause mortality in Chinese and US middle-aged and older adults using parallel, dataset-specific analytical approaches.
Methods
A comparative analysis was conducted using data from the China Health and Retirement Longitudinal Study (CHARLS) and the US National Health and Nutrition Examination Survey (NHANES) linked to mortality records. A total of 7478 participants from CHARLS and 8363 from NHANES were included after excluding those aged <45 years and with missing data. Multivariable Cox proportional hazards models, restricted cubic splines (RCS), and propensity score matching (PSM) were utilized to evaluate the associations between TyG-BMI and all-cause mortality.
Results
During follow-up, distinct mortality association patterns were observed. In CHARLS, TyG-BMI showed a significant L-shaped inverse association (P for non-linearity <0.001), and participants in the highest quartile had a lower mortality risk than those in the lowest quartile (HR = 0.59, 95% CI: 0.46-0.75). The inverse Q4-versus-Q1 association remained statistically significant after PSM (HR = 0.52, P < 0.001). In NHANES, TyG-BMI showed a U-shaped association (P for non-linearity <0.001), with a nadir at 240.87, while the Q4-versus-Q1 association remained null after PSM (HR = 0.94, P = 0.44). Age significantly modified the association in NHANES after correction for multiple testing (P for interaction <0.001), with high TyG-BMI associated with lower mortality among adults aged ≥65 years (HR = 0.59, 95% CI: 0.48-0.73), whereas no significant association was observed among those aged <65 years (HR = 1.15, 95% CI: 0.81-1.65). Additional analyses showed that BMI alone also exhibited significant non-linear mortality associations, and TyG-BMI did not materially improve prognostic discrimination compared with BMI alone.
Conclusions
The association between TyG-BMI and all-cause mortality differed between Chinese and US middle-aged and older adults. However, TyG-BMI provided little or no incremental prognostic discrimination beyond BMI alone. These findings support cautious, population-specific interpretation of TyG-BMI rather than its use as a prognostically superior alternative to BMI.