A new study analyzing nearly 19,000 intensive care unit patients found that patients with elevated triglyceride-glucose index levels had significantly higher risk of dying within 28 days of admission.
The triglyceride-glucose (TyG) index, a widely accepted surrogate for insulin resistance, is calculated using fasting triglyceride and glucose values. The study researchers reported that each 1-unit increase in TyG was linked to a 58% higher risk of death (hazard ratio [HR] = 1.58, 95% confidence interval [CI] = 1.25–2.00). Above the 8.84 threshold, this risk nearly doubled (HR = 1.99, 95% CI = 1.48–2.66), wrote the study authors, led by Bo Li of the Department of Cardiology at the Sixth Affiliated Hospital, School of Medicine at the South China University of Technology.
In contrast, TyG increases below this level were not significantly associated with mortality.
The investigators used a large critical care database that covered more than 200 U.S. hospitals. The mean baseline TyG index was 8.92. Of the patients studied, 973 (5.15%) died within 28 days. Participants were grouped by TyG tertiles, and the researchers found that mortality in the highest tertile (at least 9.16) was 6.77%—nearly double the rate of the lowest tertile (up to 8.57), at 3.66%.
Kaplan-Meier survival curves confirmed shorter survival times with higher TyG levels. Statistical modeling identified a nonlinear association with a clear inflection point at 8.82. Subgroup analyses showed the association held among patients under 60 years (HR = 1.89, 95% CI = 1.13–3.14, P = .0147) and those with body mass index (BMI) of less than 25 kg/m² (HR = 1.64, 95% CI = 1.05–2.54, P = .0282).
Higher TyG levels were also linked to increased rates of comorbidities such as diabetes, acute myocardial infarction, and sepsis. Patients with elevated TyG tended to be younger, with higher BMI and poorer metabolic profiles, including elevated blood urea nitrogen and reduced HDL cholesterol.
The researchers noted that the TyG index is simpler to calculate than other insulin resistance measures. They suggested its use as a practical, early risk indicator in critical care settings.
“This finding has significant clinical implications,” the authors wrote. “By recognizing this nonlinear effect, health care providers can modify treatment strategies based on changes in the TyG index.”
To assess the potential impact of unmeasured confounding, the researchers applied E-value calculations. Results suggested that observed associations were robust even when accounting for unknown variables.
The authors disclosed no conflicts of interest.
Source: Frontiers in Endocrinology