This study is the first retrospective analysis of the association between SHR levels and new-onset AF post-CABG, confirming SHR as an independent predictor of POAF with a dose-dependent increase in AF risk. These results suggest SHR may serve as a novel predictor of AF post-CABG.

Existing studies indicate SHR, reflecting acute stress-induced glucose dysregulation, is strongly associated with AF prognosis. Cheng et al. observed a U-shaped relationship between SHR and all-cause mortality in critically ill AF patients, indicating synergistic damage from excessive stress and metabolic disorders [12]. Zhou et al. linked postoperative glycemic variability (GV) to POAF risk, though GV’s clinical utility is limited by data continuity requirements and inability to distinguish chronic hyperglycemia’s impact [13]. SHR, derived from single-admission blood glucose and HbA1c, combines operational simplicity with pathophysiological relevance.

Research on SHR’s role in cardiovascular disease has expanded to various pathological scenarios. Chen et al. found higher postoperative SHR levels (> 1.40) correlated with increased in-hospital and post-discharge mortality [14]. In acute myocardial infarction, SHR ≥ 1.4 was associated with a 2.8-fold increase in 30-day mortality (OR = 2.80, 95% CI: 1.92–4.10) [14]. SHR also predicts mortality in diabetic nephropathy and chronic kidney disease and rehospitalization in heart failure [15, 16]. In cerebrovascular disease, SHR predicts mortality risk [17].

Huang et al. [18] reported a U-shaped association between the stress hyperglycemia ratio (SHR) and all-cause mortality in patients with acute myocardial infarction (AMI), with an inflection point at SHR = 1.09. In those with concomitant atrial fibrillation (AF), a low SHR (≤ 1.09) was linked to a markedly higher risk of death, suggesting that relative hypoglycemia may exacerbate arrhythmic risk via neuroendocrine activation and supporting an SHR of 1.09 as a stratification threshold. Koracevic et al. [19], also in an AMI cohort, observed a positive linear relationship between admission hyperglycemia (≥ 8.0 mmol/L) and both AF incidence and in-hospital mortality, reporting that AMI patients with hyperglycemia and AF had a 14.5-fold greater mortality risk compared to those without either condition.

Our study extends this work to the post–coronary artery bypass grafting (CABG) population and, for the first time, demonstrates a linear dose-response relationship between SHR and new-onset postoperative AF (POAF) risk (OR = 1.63), thus identifying SHR as a novel metabolic biomarker for early POAF warning. Together, the studies by Huang, Koracevic, and our own systematically reveal the prognostic value of stress hyperglycemia metrics in cardiovascular events. However, the pattern of association varies by clinical context and pathophysiology: AMI research highlights a nonlinear SHR–mortality relationship with AF as an effect modifier, whereas the CABG setting emphasizes surgery-related metabolic perturbations driving POAF. These findings underscore the need for stage-specific glycemic management strategies–AMI patients should guard against the synergistic lethality of low SHR and AF. In contrast, CABG patients may benefit from dynamic SHR monitoring to optimize perioperative anti-inflammatory care. Future investigations should integrate multi-omic biomarkers and real-time glucose monitoring to develop precise metabolic–electrophysiological risk-prediction models.

In this study, we constructed a multivariate logistic regression model to confirm that SHR, as a comprehensive indicator reflecting the balance between acute metabolic stress and chronic blood sugar regulation, was independently and robustly positively associated with AF after CABG. It is particularly noteworthy that after fully adjusting for 21 potential confounders including demographic characteristics, underlying diseases, critical scores, laboratory indicators and treatment measures, the highest SHR group still showed an additional risk increase of 31% compared with the lowest group (OR = 1.31, 95% CI: 1.03–1.67, P = 0.0275), and each unit SHR increase corresponded to a 1.63-fold risk multiplier effect (OR = 1.63, 95% CI: 1.19–2.23, P = 0.00 23). There was a significant dose-response relationship between SHR and POAF after CABG, and the association was in a strict linear pattern (RCS analysis, overall association P < 0.05, nonlinear test P > 0.05). This finding suggests that SHR may surpass traditional blood glucose indicators and have unique predictive value in metabolic risk assessment in cardiac surgery patients.

Subgroup analysis confirmed that SHR, as an independent predictor of POAF, is robust across populations, and its effect is enhanced in women, hypertension, and CKD patients, suggesting synergistic damage from metabolic stress and underlying diseases. Although SHR consistently predicts POAF risk in the elderly (≥ 65 years old), women, and hypertensive patients, its effect is weakened in the subgroup of acute kidney injury (AKI). This is consistent with previous studies that AKI may mask the effects of metabolic stress markers through competing mechanisms such as uremic cardiomyopathy or inflammation. Even in diabetic patients, the effect of SHR on POAF is still significant.

At the same time, this study systematically verified the independent predictive value of SHR for POAF after CABG through dual feature screening of LASSO regression and Boruta algorithm. After strict variable screening (LASSO’s L1 regularization constraint and Boruta’s random forest importance assessment), the two algorithms jointly locked SHR, Charlson score, sofa score, age, heart failure and mechanical ventilation time as core predictors. Among them, SHR was included in both methods as the only metabolic-related indicator, and its predictive efficiency was highly consistent with the results of multivariate logistic regression (OR = 1.63), suggesting its core position in POAF risk stratification.

From the perspective of pathophysiological mechanisms, the association between SHR and POAF may involve multiple pathway interactions. They affect the following aspects: (1) During stress-induced hyperglycemia, NOX2 is activated by phosphorylation of PKCβII. The ROS produced by it not only oxidizes calmodulin but also activates CaMKII (calcium/calmodulin-dependent protein kinase II) to cause hyperphosphorylation of ryanodine receptor 2 (RyR2), triggering sarcoplasmic reticulum calcium leakage [20]. The oxidative modification of intracellular calcium regulatory proteins leads to a decrease in atrial mitochondrial membrane potential and energy metabolism disorders. Reduced mitochondrial membrane potential leads to decreased ATP synthesis, inhibiting SERCA2a activity, further aggravating cytoplasmic calcium overload, and delaying afterdepolarization, leading to atrial electrical remodeling [21,22,23,24]. The chronic hyperglycemia reflected by SHR is promoted by upregulating the TGF-β1/Smad signaling pathway, promoting the transdifferentiation of atrial fibroblasts into myofibroblasts, and accelerating collagen deposition and interstitial fibrosis [25,26,27]. The fibrotic matrix not only forms an anatomical barrier but also destroys the continuity of electrical conduction and forms micro-reentry loops, thus maintaining the persistence of AF [28]. (2)Stress-induced hyperglycemia is mainly caused by the paraventricular nucleus (PVN) of the hypothalamus, which simultaneously activates the sympathetic-SAM axis and the HPA axis [29,30,31]. Sympathetic nerve excitement directly increases blood glucose, inhibits insulin secretion through α2-adrenergic receptors, and exacerbates insulin resistance by releasing free fatty acids [32]. Insulin resistance is a significant risk factor for AF and may promote the development of [33]. At the same time, chronic insulin resistance is closely related to forming an atrial fibrotic matrix [34].

It is worth noting that the dose-response relationship in this study showed a linear trend without an inflection point, suggesting that even if SHR is within the clinical normal range, its subtle fluctuations may still affect postoperative electrical stability, which echoes the conclusion of Sim et al. on the correlation between dynamic changes in perioperative blood glucose and arrhythmia risk [35]. Compared with previous studies that focused on the impact of isolated blood glucose indicators (such as fasting blood glucose and HbA1c) [36], this study innovatively introduced SHR as an integrated biomarker. The advantage of this indicator is that it captures both the intensity of acute stress response and the basal metabolic reserve capacity, which may more accurately reflect the “double-hit” effect - the interaction between acute metabolic disorders caused by surgical trauma and the patient’s inherent glucose metabolism defects. In addition, the non-threshold effect shown by restricted cubic spline analysis suggests that the traditional dichotomous cutoff value may underestimate the continuous risk attribute of SHR, which has important revision significance for the existing risk stratification system.

In clinical practice, these results support targeted blood glucose management for patients with high SHR. For example, dynamic blood glucose monitoring and insulin dose adjustment in the early postoperative period (e.g., within 120 h) may reduce the risk of arrhythmias. In addition, four metabolites (acetylglutamine, ornithine, methionine, and arginine) found in pericardial fluid and serum provide evidence for the mechanistic association between SHR and POAF [37]. These metabolites are involved in the urea cycle and redox homeostasis, and hyperglycemia can disrupt the above processes.

In summary, SHR is an essential factor affecting the occurrence of AF in patients after CABG. It is a new, cost-effective risk stratification tool for POAF after CABG. Integrating it into the clinical process is expected to improve the prognosis of high-risk patients through early identification and individualized prevention. It may be a potential new indicator for predicting the incidence of AF after CABG. Early detection and intervention of SHR can be used as a clinical prognostic assessment tool and play an essential role in future treatment strategies.

However, our research is not without limitations. First, due to limited measurement data, this study did not include cardiac-specific biomarkers (such as troponin, BNP) and imaging indicators (such as left ventricular ejection capacity and left atrial size.), inflammatory markers (such as IL-6, CRP,) and oxidative stress mediators (such as MDA, SOD), etc. The unmeasured indicators of cardiac function, inflammation, and metabolic pathways may partially obscure their mechanistic associations. Secondly, this retrospective study represents a single-center analysis utilizing observational data derived from the MIMIC-IV database, which may limit the universality of the conclusions. In addition, our research has a limited sample size, and the conclusions drawn from our study would benefit from validation through larger cohort analyses. In the future, it is necessary to verify it through a large sample and multicenter cohort and explore the improvement effect of targeted regulation of stress hyperglycemia, such as short-term preoperative insulin intensive therapy, on electrophysiological remodeling. Furthermore, longitudinal studies still need to clarify the dynamic impact of the time and frequency of SHR measurement on predictive efficiency. In addition, we cannot determine the time and leading cause of AF after CABG surgery, which will reduce the clinical relevance of this analysis.

Future studies need to verify the clinical value of SHR in randomized controlled trials and explore its combined application with emerging biomarkers such as left atrial diameter. Develop a multidimensional prediction model integrating SHR, conduct intervention studies to verify the preventive effect of blood sugar stress regulation on POAF, and reveal the molecular mechanisms of SHR-related metabolic pathways through omics technology.