Relationship between stress hyperglycemia ratio and the incidence of atrial fibrillation in patients after coronary artery bypass grafting: a retrospective study based on the MIMIC-IV database

Abstract

Background: The stress hyperglycemia ratio (SHR) is a clinical index that quantifies acute stress-induced hyperglycemia by comparing immediate blood glucose levels with chronic glucose control (reflected by HbA1c). It is especially valuable in cardiovascular disease and surgical prognosis. However, there is still a lack of research on the relationship between SHR and new-onset atrial fibrillation (AF) in patients after coronary artery bypass grafting (CABG). This study investigates the impact of postoperative SHR on AF risk following CABG.

Methods: This study is a retrospective cohort analysis conducted through the MIMIC-IV database, which included adult patients who underwent CABG and were admitted to the ICU. These patients were categorized into three distinct groups according to the tertiles of the baseline SHR level, and the primary outcome was the incidence of postoperative atrial fibrillation (POAF). We employed logistic regression models, restricted cubic splines (RCS), threshold effect analysis, ubgroup analysis, Boruta algorithm, lasso algorithm, and receiver operating characteristics (ROC) to analyze the relationship between SHR and POAF incidence comprehensively.

Results: 2112 patients undergoing CABG were included in this study, with a median age of 69 years (IQR: 62-76), of whom 1643 (77.79%) were male. Logistic regression results showed that the incidence of AF was significantly increased in patients in the highest third of the SHR group compared with the lowest third group (OR = 1.31, 95%CI = 1.03-1.67; P = 0.0275). SHR was an independent risk factor for the incidence of POAF (OR = 1.63, 95%CI = 1.19-2.23; P = 0.0023). At the same time, RCS analysis showed that SHR was positively and linearly correlated with the incidence of POAF in patients after cardiac surgery (P = 0.009, P for Nonliner = 0.848). Threshold effect analysis identified no significant threshold and further supported a linear relationship between SHR and POAF. In addition, SHR was double-screened by Boruta and Lasso algorithms, indicating that it was statistically and biologically significantly associated with AF after CABG.

Conclusion: SHR is significantly related to AF after CABG. As SHR increases, the risk of POAF increases. Incorporating SHR into post-CABG risk assessment enhances AF prediction, offering a valuable reference for clinical decision-making. It may also be a potential biomarker for studying pathological mechanisms in patients after cardiac surgery. In the future, combining multi-omics data with clinical intervention trials is necessary to verify its clinical application value further.

Keywords: Coronary artery bypass grafting; MIMIC-IV database; Postoperative atrial fibrillation; Stress hyperglycemia ratio.

Fig. 1

Flow chart. AF, atrial fibrillation; POAF, postoperative atrial fibrillation. CABG, Coronary Artery Bypass Grafting

Fig. 2

Restricted cubic spline of SHR and POAF

Fig. 3

Receiver operating characteristic curve of model 4

Fig. 4

Forest plot showing the association between SHR and POAF in different subgroups

Fig. 5

Selected features of new-onset AF in SHR and patients after CABG. (A) The graphical representation of LASSO regression analysis results. (B) The cross-validation visualization of LASSO regression analysis results. The Boruta algorithm is utilized for feature selection. The x-axis denotes the names of the variables, while the y-axis represents the Z scores of each variable. The box plot illustrates the distribution of Z scores for each variable during the model computation process. The feature selection network visualization. The blue part is the result of LASSO regression analysis, while the red part is the result of Boruta algorithm. The purple part is the shared variables that are consistently identified by the two algorithm results