Nonlinear relationship between triglyceride-glucose index and cardiovascular mortality with competing risk analysis on populations aged 18–80 years

BACKGROUND: The existing evidence regarding the relationship between the triglyceride-glucose index (TyG index) and cardiovascular mortality risk remains relatively limited and controversial, particularly within the context of competing risk scenarios. This study seeks to investigate this relationship, while further incorporating the impact of non-cardiovascular mortality as a competing risk event to this association. METHODS: Data of eligible participants were extracted from National Health and Nutrition Examination Surveys (NHANES) 1999-2018. Traditional Cox proportional hazards regression and Fine-Gray sub-distribution hazard models were applied to assess the TyG index and cardiovascular mortality relationship. Restricted cubic splines were used to estimate possible non-linearity, while segmented regression and log-likelihood ratio tests were used to identify threshold values and model fit. RESULTS: The final analysis compromised a number of 23,800 participants, with a mean age of 47.75 ± 18.06 years, and female prominent (51.72%). After fully adjusted, it revealed a positive relationship between the TyG index and cardiovascular mortality risk (HR = 1.24, 95%CI 1.08-1.41, P = 0.0017). Furthermore, upon considering non-cardiovascular mortality as competing risk event, the result of Fine-Gray sub-distribution hazard model analysis attenuated but remained significantly positive (sHR = 1.11, 95%CI 1.11-1.11, P < 0.0001). Besides, a non-linear reversed L-shaped relationship was revealed, with a cutoff value determined as 9.4. Below 9.4, the relationship was insignificant (HR = 1.10, 95%CI 0.92-1.31, P = 0.2866), whereas beyond 9.4, the relationship became positive (HR = 1.64, 95%CI 1.21, 2.22, P = 0.0014), and the log-likelihood ratio test confirmed the threshold effect (P = 0.049). Significant interaction was observed in age and body mass index (BMI) subgroups, respectively, with individuals ≤ 65 years and normal BMI category exhibited higher risk in the relationship (P for interaction < 0.05). CONCLUSIONS: The present study reveals a robust positive relationship between the TyG index and cardiovascular mortality among individuals aged 18-80 years despite the influence from non-cardiovascular mortality event. Additionally, the relationship was non-linear with the risk intensifying when TyG index beyond a specific threshold. Besides, individuals younger than 65 years old with normal BMI may be more susceptible in this relationship. CLINICAL TRIAL NUMBER: Not applicable.