The U-shaped relationship between triglyceride glucose-body mass index and prognosis of sepsis patients: a retrospective study

BACKGROUND: The triglyceride glucose-body mass index (TyG-BMI) has been validated as a reliable indicator of insulin resistance in critically ill patients. Despite its established prognostic value in cardiovascular and metabolic disorders, its role in sepsis outcomes remains controversial. The aim of this study was to clarify the nonlinear relationship between TyG-BMI and all-cause mortality in intensive care unit (ICU)-admitted patients with sepsis. METHODS: Adult patients with sepsis admitted to the ICU for the first time were enrolled from the MIMIC-IV database and stratified into tertiles on the basis of TyG-BMI. The primary outcome was 360-day all-cause mortality, whereas the secondary outcome was 30-day all-cause mortality. Chi-square tests and Kaplan-Meier survival curves were constructed to evaluate survival outcomes across the three groups. Restricted cubic splines, Cox proportional hazards regression models, and exploratory subgroup analyses were further employed to systematically investigate the nonlinear association between TyG-BMI and prognosis in patients with sepsis. RESULTS: A total of 1249 patients were enrolled in this study, with a median age of 64.3 years and a median TyG-BMI of 261.3, among whom 489 (39.2%) were female. Compared with Tertiles 2 and 3, Tertile 1 had the highest mortality rate, with statistically significant associations observed between TyG-BMI and 360-day all-cause mortality (P = 0.028). Restricted cubic spline analysis confirmed a significant U-shaped association between TyG-BMI and 360-day all-cause mortality, with a turning point at TyG-BMI = 289.4. No statistically significant nonlinear association was found between TyG-BMI and the secondary endpoint of 30-day all-cause mortality. In the fully adjusted two‑piecewise Cox regression model (Model 3), each unit increase in TyG‑BMI was associated with a significantly decreased risk of 360‑day all‑cause mortality below the turning point (HR = 0.996, 95% CI 0.993-0.999, P = 0.014), and a significantly increased risk above the turning point (HR = 1.002, 95% CI 1.001-1.004, P = 0.011). Exploratory stratified analyses revealed significant interactions with age, gender, race, and congestive heart failure status, which should be interpreted with caution. CONCLUSIONS: In this cohort, TyG-BMI shows a significant U-shaped association with 360-day all-cause mortality in sepsis patients. The clinical value of TyG‑BMI requires further validation in large‑scale multi‑center studies.