Independent effects of the glucose-to-glycated hemoglobin ratio on mortality in critically ill patients with atrial fibrillation

Abstract

The glucose-to-glycated hemoglobin ratio (GAR) represents stress hyperglycemia, which has been closely associated with adverse outcomes in cardio-cerebrovascular diseases. No studies have examined the association between stress hyperglycemia and atrial fibrillation (AF) in critically ill patients. This study aims to explore the relationship between GAR and the prognosis of critically ill patients with AF. A retrospective cohort of patients was selected from the Medical Information Mart for Intensive Care IV (MIMIC-IV) database. The GAR was calculated based on fasting blood glucose and glycated hemoglobin levels measured after admission. The primary outcome was the 30-day mortality rate, with secondary outcomes being the 90-day and 365-day mortality rates. The GAR was divided into tertiles, and Kaplan–Meier analysis was employed to compare differences in mortality rates between groups. The Cox proportional hazards model and restricted cubic splines (RCS) were utilized to evaluate the relationship between the GAR and mortality. Subsequently, a segmented regression model was constructed to analyze threshold effects in cases where nonlinear relationships were determined. In this cohort, the second tertile of the GAR exhibited lower mortality rates at 30 days (10.56% vs 6.33% vs 14.51%), 90 days (17.11% vs 10.09% vs 17.88%), and 365 days (25.30% vs 16.15% vs 22.72%). In the third tertile, the risk of mortality at 30 days increased by 165% (HR = 2.65, 95% CI 1.99–3.54, p < 0.001), at 90 days increased by 113% (HR = 2.13, 95% CI 1.68–2.70, p < 0.001), and at 365 days increased by 70% (HR = 1.70, 95% CI 1.68–2.70, p < 0.001). The association between the GAR and patient mortality demonstrated a “J-shaped” non-linear correlation. Once the GAR exceeded 15.915, each incremental unit increase in the ratio was associated with a 27.2% increase in the risk of 30-day mortality in critically ill atrial fibrillation patients (HR = 1.262, 95% CI 1.214–1.333, p < 0.001). The GAR is associated with both short-term and long-term mortality in critically ill patients with AF in a J-shaped relationship. Both low and excessively high GAR values indicate poor prognosis.