Abstract
BACKGROUND: Glycemic variability (GV), a prevalent phenomenon of abnormal blood glucose levels in intensive care unit (ICU) is correlated with cardiac arrhythmias and in-hospital mortality. The specific link between GV and new-onset atrial fibrillation (AF), as well as in-hospital mortality, is not yet fully understood, particularly in critically ill patients without diabetes. Additionally, the potential mediating role of AF in this relationship has not yet been definitively established. This study investigated the association between GV, new-onset AF, and in-hospital mortality among these patients. METHODS: Patients without diabetes from the Medical Information Mart for Intensive Care-IV database were included in the analysis. GV was quantified as the coefficient of variation (CV), derived from the ratio of blood glucose’s standard deviation (SD) to its mean. To evaluate the relationship between GV levels and AF incidence, plus in-hospital death, multivariate logistic regression, restricted cubic spline (RCS) analyses, and a mediation effect analysis were used. RESULTS: The research involved 17,643 patients without diabetes, with a median age of 64, and 54.5% of them were male. The 30-day in-hospital ICU mortality rate was 14%, and the AF incidence rate was 9.7%. The adjusted logistic regression analysis revealed a significant link between log-transformed CV values and increased likelihood of atrial fibrillation (odds ratio [OR] 1.44; 95% confidence interval [95% CI] 1.32–1.56). Furthermore, this metric showed an even stronger relationship with in-hospital mortality risk (OR 2.09; 95% CI 1.94–2.24), with both findings reaching statistical significance. The RCS models indicated a potential nonlinear association between GV and these outcomes, with higher GV levels being associated with a greater probability of AF incidence and in-hospital death. The mediation analysis showed that the indirect effect of GV on in-hospital death was 36.26% after adjustment for covariates, suggesting AF may act as a mediator in the relationship between GV and in-hospital mortality. CONCLUSIONS: Elevated GV was associated with a higher incidence of new-onset AF and in-hospital mortality in non-diabetic critically ill patients, and new-onset AF may partly be a mediator of the association between GV and in-hospital mortality. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s13098-025-02078-9.