Association of Blood Glucose Variability During the First 3 Days After ICU Admission With 28-Day All-Cause Mortality in Critically Ill Patients With Acute Myocardial Infarction

Abstract

Background

This investigation focused on assessing the link between blood glucose variability (BGV) and 28-day all-cause mortality in individuals with critically ill AMI, as well as determining whether this relationship differed based on diabetes status.

Methods

The MIMIC-III database yielded 1720 critically ill patients with AMI for inclusion. The coefficient of variation (CV) of glucose was utilized to quantify BGV, derived from all blood glucose determinations recorded throughout the initial 72-h post-ICU admission. The 28-day all-cause mortality constituted the primary outcome.

Results

Among the 1720 critically ill patients with AMI, 28-day mortality following ICU admission occurred in 200 cases (11.63%). After full adjustment, a one-unit increment in log-transformed glucose CV was linked to greater odds of 28-day all-cause mortality (OR: 1.50, 95% CI: 1.13–1.99, p-value = 0.01). A positive link was verified through restricted cubic spline (RCS) analysis, and no nonlinear trend was apparent (p for nonlinearity = 0.205). Moreover, in individuals without diabetes, glucose CV was positively linked to 28-day all-cause mortality (OR: 2.49, 95% CI: 1.58–3.93, p-value < 0.001), whereas no significant connection was identified in individuals with diabetes (OR: 0.93, 95% CI: 0.62–1.39, p-value = 0.71). These observations showed consistency when evaluated through subgroup and sensitivity analyses.

Conclusion

Among critically ill AMI patients, glucose CV within 72-h post-ICU admission independently forecasted 28-day all-cause mortality. This association reached statistical significance in patients without diabetes, yet failed to do so in those with diabetes.