Abstract
Background
Postoperative cerebral infarction following coronary artery bypass grafting (CABG) for multivessel coronary artery disease (CAD) is a major complication and is associated with insulin resistance (IR). This study used the Triglyceride-Glucose (TyG) Index, a robust indicator of IR, to assess its association with cerebral infarction and other adverse events in patients with off-pump CABG (OPCABG).
Methods
This retrospective observational study included 3654 CAD cases from eight centres across China. The primary outcome was postoperative cerebral infarction. The predictive role of the TyG Index was evaluated using multivariate logistic regression and restricted cubic spline regression. Receiver operating characteristics analysis was conducted to assess its impact on model performance.
Results
A total of 89 patients experienced postoperative cerebral infarction. After adjusting for confounding factors, the TyG Index, whether treated as a categorical variable (OR=2.23, 95% CI 1.24 to 4.02) or a continuous variable (OR=1.80, 95% CI 1.29 to 2.51), was found to be a significant independent risk factor for postoperative cerebral infarction (both p<0.001). The restricted cubic splines regression model revealed a linear dose-response association between the TyG Index and the risk of postoperative cerebral infarction (p for non-linearity=0.861). Subgroup analysis did not indicate any interactions among subgroups (p for interaction >0.05). Incorporating the TyG Index yielded a modest but statistically significant improvement in discrimination for postoperative cerebral infarction (area under the receiver operating characteristics curve 0.724 vs 0.708; p<0.001).
Conclusions
IR reflected by an elevated TyG Index predicts the risk of postoperative cerebral infarction in patients undergoing OPCABG.
Trial registration number
Chinese Clinical Trial Registry: Chictr2400085741.
Keywords: Acute Coronary Syndrome, Coronary Artery Bypass, Stroke, Metabolic Syndrome
WHAT IS ALREADY KNOWN ON THIS TOPIC
Postoperative cerebral infarction is a severe complication in patients with coronary artery disease (CAD) undergoing off-pump coronary artery bypass grafting (OPCABG).
Insulin resistance (IR) significantly increases the incidence of cerebrovascular events.
However, its accurate evaluation remains a clinical challenge.
WHAT THIS STUDY ADDS
This study provides an early preoperative clinical lipid-glucose metabolism marker for identifying high-risk patients for postoperative cerebral infarction following OPCABG.
We employed the clinically easy-to-calculate Triglyceride-Glucose (TyG) Index as an indicator of IR levels and demonstrated, based on a nationwide CAD patient cohort from nine cardiovascular centres, the predictive value of the TyG Index for postoperative cerebral infarction following OPCABG.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
The findings of this study will enable surgeons to identify patients with IR before OPCABG.
These findings may also reduce the risk of postoperative cerebral infarction through targeted treatments to control IR.
Introduction
Coronary artery bypass grafting (CABG), which can provide complete revascularisation, is the preferred treatment strategy for complex coronary artery disease (CAD).1 Perioperative complications, including post-CABG cerebral infarction, could substantially impact the prognosis of patients undergoing CABG.2 Notably, the incidence of postoperative cerebral infarction is approximately 1.2%–2%, and the mortality rate of affected patients could reach nearly 20% in the perioperative period.2 3
Insulin resistance (IR) is a common metabolic disorder among patients preparing for CABG. IR is defined as a decreased sensitivity to insulin at normal plasma glucose levels.4 Consequently, it is associated with the occurrence and progression of cardiovascular and cerebrovascular diseases as well as poor prognosis following revascularisation procedures.5 6 Notably, the novel lipid-glucose metabolism marker, the Triglyceride-Glucose (TyG) Index, is a simple and clinically applicable measure that accurately reflects the degree of IR in patients. Furthermore, it demonstrates strong consistency with the gold standards used to assess IR.7,9
IR assessed by the TyG Index is associated with adverse outcomes in patients with cardiovascular diseases and following various cardiac surgical procedures.10 11 Additionally, a 10-year follow-up study indicated that patients with IR have an approximately 1.5-fold increased risk of cardiovascular adverse events after undergoing CABG compared with patients without IR.12 A meta-analysis has also demonstrated a significant, heterogeneous association between the TyG Index and cerebral infarction.13 To date, no research has elucidated the relationship between the TyG Index, as a measure of IR, and the incidence of cerebrovascular accidents and other perioperative complications in patients undergoing off-pump CABG (OPCABG) surgery.14 In this study, we used the TyG Index to investigate the relationship between the degree of IR and the incidence of adverse perioperative events such as cerebral infarction in OPCABG patients.
Methods
Study cohort
This study was a multicentre, retrospective, observational cohort study. We retrospectively included a total of 8150 patients who were aged ≥18 years, diagnosed with CAD and underwent OPCABG surgery at one of eight tertiary class 3 clinical centres (detailed in online supplemental table S1) in Mainland China from January 2021 to October 2022. A total of 3655 patients were included in the analysis after excluding individuals (1) Without complete fasting blood glucose (FBG) and triglyceride (TG) data, (2) Undergoing concurrent surgeries, (3) With a history of CABG, and (4) With severe infections or malignant tumours. Moreover, patients were stratified into three tertiles according to the TyG Index. The cohort design is shown in figure 1.
Figure 1. The flow diagram of the cohort design. CAD, coronary artery disease; OPCABG, off-pump coronary artery bypass grafting; FBG, fasting blood glucose; TG, triglycerides.
This research was conducted following the Declaration of Helsinki. Patients or the public were not involved in the design, conduct, reporting or dissemination plans of our research. Using our multicentre retrospective registry study database, we investigated the predictive value of preoperative clinical information, laboratory tests, and inspection data for the incidence and prognosis of postoperative complications in patients with CAD. These clinical data were obtained by reviewing electronic medical record systems. Furthermore, the study was registered with the Chinese Clinical Trials Registry (registration number: Chictr2400085741).
Data collection and definitions
The demographic data of patients, including age, gender and smoking status, as well as medical history, such as hypertension, type 2 diabetes mellitus (T2DM) and history of myocardial infarction, were self-reported on admission and confirmed through clinical examinations following admission. Imaging was conducted and reported according to international standards. Laboratory tests were performed using internationally standardised and qualified instruments. Moreover, the primary exposure variable in this study was the TyG Index. To ensure consistency with the established calculation method, we first converted TG and FBG values from mmol/L to mg/dL (TG (mg/dL) = TG (mmol/L) × 88.5; FBG (mg/dL) = FBG (mmol/L) × 18). The TyG Index was then calculated as ln(TG (mg/dL) × FBG (mg/dL)/2).8 Information during and after surgery was obtained from electronic medical records, including surgical nursing records, surgeons’ operation notes and postoperative intensive care unit nursing records.
The primary outcome observed in this study was postoperative symptomatic cerebral infarction in patients, defined as new-onset or worsening neurological focal signs persisting for over 24 hours postoperatively. Instances of cerebral infarction were subsequently corroborated by lesions detected in follow-up cranial CT scans.15 Secondary outcomes included the following: postoperative myocardial infarction defined as chest pain unrelated to the surgical incision and accompanied by ECG, ultrasonic and high-sensitivity troponin evidence;16 duration of the OPCABG procedure, duration of intensive care unit stay, time of hospitalisation and in-hospital mortality according to surgical nursing records and other electronic medical records.
Statistical analysis
We first quantified variable-wise missingness and applied a 20% threshold to screen out variables with excessive missingness. To minimise selection bias, we did not rely solely on complete-case analysis. Instead, we performed multivariate imputation by chained equations (MICE) for continuous variables and mode imputation for categorical variables. The imputed variables and their missingness rates are presented in online supplemental table S4. Continuous variables were summarised as means±SD or medians with IQRs and compared using analysis of variance. Categorical variables were presented as counts and percentages, with group differences assessed by Pearson’s χ2 test or Fisher’s exact test. Baseline characteristics were analysed across TyG Index tertiles. We performed univariate linear and logistic regression analyses to identify the risk factors for postoperative cerebral infarction following OPCABG. Additionally, we constructed multivariate logistic regression models to validate whether the TyG Index is an independent risk factor for postoperative cerebral infarction in OPCABG patients. Two models were employed to adjust for confounding variables: Model 1 adjusted for sex and age, while Model 2 further adjusted for variables with a value of p<0.05 in the univariate analysis, including a history of cerebral infarction, hypertension, T2DM, atrial fibrillation and smoking. The TyG Index was analysed as both a continuous variable and tertile variable (T1: 7.06–8.56; T2: 8.56–9.06; T3: 9.06–12.04). Effect size was reported using ORs and 95% CIs. We evaluated multicollinearity among predictors in the multivariate logistic regression by calculating variance inflation factors (VIFs) for each covariate and by inspecting pairwise correlations. We considered VIF ≥2.5 and absolute pairwise correlation coefficients (|r|) ≥0.6 as indicative of potential collinearity. The restricted cubic spline (RCS) model was constructed to evaluate the dose-response association between the TyG Index and the risk of postoperative ischaemic stroke, using the 5th, 35th, 65th and 95th percentiles as four knots to achieve an optimal balance between flexibility and stability. Then, we conducted prespecified subgroup analyses to test the consistency of the predictive effect of the TyG Index across clinically relevant strata. Cut-offs were defined to mirror routine clinical thresholds: age ≥70 years to represent older, higher-risk patients; body mass index (BMI) ≥24 kg/m² to denote overweight status in Chinese adults; and physician-diagnosed hypertension, T2DM and hyperlipidaemia as binary risk-factor states; sex was analysed as male versus female. These thresholds were selected to maximise clinical interpretability, and align with contemporary prevention guidance and national criteria, given that age, adiposity, blood pressure, glycaemic status and lipid abnormalities are established determinants of cerebrovascular risk.17 By calculating the value of p for interaction, we assessed the impact of each subgroup on the primary outcome. Finally, we assessed whether adding the TyG Index improved prediction of postoperative cerebral infarction using receiver operating characteristics curve analysis based on multivariate logistic regression (Model 2 variables). Model discrimination was compared via area under the receiver operating characteristics curve (AUC) and the DeLong test, while net reclassification improvement (NRI) and integrated discrimination improvement (IDI) were used to evaluate improvements in risk reclassification. We simultaneously evaluated calibration for both the conventional model (Model 2 variables) and the TyG-augmented model using decilewise calibration plots based on quantile binning of predicted risks and the Hosmer–Lemeshow goodness-of-fit test (10 groups; df=8). All statistical analyses were performed using R (V.4.4.2, released on 31 October 2024) and Python (V.3.12.6, released on 6 September 2024).
Results
Baseline characteristics
In this study, 3654 patients were included in the research cohort. The average age of the cohort was 62.5±8.9 years. Male patients accounted for 75.7% of the cohort. The baseline characteristics of these groups are presented in table 1.
Table 1. Baseline characteristics of participants stratified by TyG Index.
| Variables | Total | Tertile 1 | Tertile 2 | Tertile 3 | P value |
|---|---|---|---|---|---|
| (n=1218) | (n=1218) | (n=1218) | |||
| TyG Index | 8.9±0.6 | 8.2±0.3 | 8.8±0.1 | 9.6±0.4 | <0.001 |
| Demographic characteristics | |||||
| Age (years) | 62.5±8.9 | 63.7±8.5 | 62.4±8.9 | 61.4±9.0 | <0.001 |
| BMI (kg/m2) | 25.8±3.3 | 25.2±3.4 | 25.9±3.1 | 26.3±3.3 | <0.001 |
| Male, n (%) | 2786 (75.7%) | 996 (81.8%) | 938 (77.0%) | 852 (69.9%) | <0.001 |
| Smoking, n (%) | 1556 (42.6%) | 527 (43.3%) | 528 (43.3%) | 501 (41.1%) | 0.444 |
| Medical history | |||||
| Hypertension, n (%) | 2245 (61.4%) | 701 (57.6%) | 735 (60.3%) | 809 (66.4%) | <0.001 |
| T2DM, n (%) | 1368 (37.4%) | 298 (24.5%) | 389 (31.9%) | 681 (55.9%) | <0.001 |
| Hyperlipidaemia, n (%) | 1779 (48.7%) | 555 (45.6%) | 577 (47.4%) | 647 (53.1%) | <0.001 |
| Ischaemic stroke | 848 (23.9%) | 272 (23.3%) | 298 (25.1%) | 278 (23.3%) | 0.487 |
| Myocardial infarction, n (%) | 503 (13.8%) | 154 (12.6%) | 202 (16.6%) | 147 (12.1%) | 0.002 |
| Carotid revascularisation, n (%) | 32 (0.9%) | 16 (1.3%) | 11 (0.9%) | 5 (0.4%) | 0.057 |
| Preoperative investigations | |||||
| TG (mg/dL) | 148.4±91.4 | 88.2±20.9 | 135.3±31.8 | 221.7±120.1 | <0.001 |
| FBG (mg/dL) | 114.5±44.8 | 89.3±15.9 | 105.1±24.8 | 149.1±57.0 | <0.001 |
| HDL-C (mmol/L) | 1.0±0.2 | 1.1±0.2 | 1.0±0.2 | 0.9±0.2 | <0.001 |
| LDL-C (mmol/L) | 2.2±0.9 | 2.0±0.8 | 2.3±0.8 | 2.4±0.9 | <0.001 |
| HbA1c (%) | 6.6±1.3 | 6.0±0.9 | 6.4±1.2 | 7.3±1.5 | <0.001 |
| LVEF | 59.5±8.3 | 59.9±8.0 | 59.6±8.1 | 58.9±8.7 | 0.027 |
| Intraoperative measurements | |||||
| Operation time (hours) | 4.2±0.8 | 4.2±0.8 | 4.2±0.8 | 4.3±0.8 | 0.029 |
| Number of target vascular (n) | 3.4±1.0 | 3.3±1.0 | 3.4±1.0 | 3.4±0.9 | 0.012 |
| Blood loss volume (ml) | 724.7±318.2 | 707.8±283.0 | 733.6±321.2 | 732.9±346.8 | 0.127 |
| RBC infusion (U) | 0.1±0.5 | 0.1±0.6 | 0.1±0.5 | 0.1±0.6 | 0.381 |
| Plasma infusion (U) | 10.0±62.6 | 11.5±65.9 | 9.2±60.5 | 9.4±61.3 | 0.537 |
| Lowest intraoperative SBP (mm Hg) | 94.7±13.5 | 94.3±13.6 | 94.8±13.5 | 95.1±13.6 | 0.389 |
| LIMA use, n (%) | 2228 (61.0%) | 724 (59.4%) | 727 (59.7%) | 777 (63.7%) | 0.051 |
| SVG use, n (%) | 3532 (96.6%) | 1170 (96.1%) | 1180 (96.9%) | 1182 (97.0%) | 0.392 |
| Postoperative events | |||||
| Cerebral infarction, n (%) | 89 (2.4%) | 18 (1.5%) | 26 (2.1%) | 45 (3.7%) | 0.001 |
| Ventilator time (hours) | 14.0±5.5 | 27.6±32.4 | 30.2±77.8 | 28.8±37.2 | 0.27 |
| Myocardial infarction, n (%) | 11 (0.3%) | 3 (0.2%) | 5 (0.4%) | 3 (0.2%) | 0.693 |
| IABP, n (%) | 157 (4.3%) | 46 (3.8%) | 49 (4.0%) | 62 (5.1%) | 0.239 |
| ECMO, n (%) | 14 (0.4%) | 6 (0.5%) | 3 (0.2%) | 5 (0.4%) | 0.607 |
| ICU stay (hours) | 35.3±59.3 | 33.3±37.7 | 36.5±83.0 | 36.3±47.4 | 0.064 |
| Length of hospitalisation (days) | 14.0±5.5 | 13.8±4.8 | 14.2±6.3 | 14.0±5.5 | 0.428 |
| In-hospital death, n (%) | 68 (1.9%) | 23 (1.9%) | 21 (1.7%) | 24 (2.0%) | 0.904 |
Values of p in bold are <0.05.
BMI, body mass index; ECMO, extracorporeal membrane oxygenation; FBG, fasting blood glucose; HbA1c, glycated haemoglobin A1c; HDL-C, high-density lipoprotein cholesterol; IABP, intra-aortic balloon pump; ICU, intensive care unit; TyG Index, Triglyceride-Glucose Index; LDL-C, low-density lipoprotein cholesterol; LVEF, left ventricular ejection fraction; RBC, red blood cell; SBP, systolic blood pressure; T2DM, type 2 diabetes mellitus; TG, triglycerides; SVG usage, saphenous vein graft usage; LIMA usage, left internal mammary artery usage.
Throughout the three tertiles, patients with higher TyG Indexes had a younger average age (tertile 1: 63.7±8.5 vs tertile 2: 62.4±8.9 vs tertile 3: 61.4±9.0, p<0.001), a lower proportion of male patients (tertile 1: 81.8% vs tertile 2: 77.0% vs tertile 3: 69.9%, p<0.001) and higher BMI levels (tertile 1: 25.2±3.4 vs tertile 2: 25.9±3.1 vs tertile 3: 26.3±3.3, p<0.001). These patients also had a higher proportion of hypertension, T2DM and hyperlipidaemia history (p<0.001). Compared with the lower TyG Index group, the higher tertiles also exhibited significantly higher levels of lipid profiles, glycated haemoglobin A1c and serum creatinine along with lower estimated glomerular filtration rate levels and left ventricular ejection fraction.
Surgical data and postoperative outcomes
The intraoperative parameters, including the duration of the operation and the number of target vessels, were incrementally higher as the TyG Index elevated. No significant differences across the tertiles were observed regarding intraoperative blood loss, red blood cell and plasma infusion.
Eighty-nine (2.4%) patients experienced cerebral infarction during the postoperative period. Additionally, there was a notable trend of increasing incidence of postoperative cerebral infarction across tertiles of the TyG Index, from 1.5% in the lowest tertile to 2.1% in the middle tertile, followed by an increase to 3.7% in the upper tertile (p=0.001). Yet no significant differences between the groups were observed in the incidences of postoperative myocardial infarction and in-hospital mortality, etc.
Then, we performed univariate analyses, which revealed that the TyG Index had predictive ability for postoperative cerebral infarction (OR=1.87; 95% CI 1.38 to 2.53, p<0.001) as detailed in table 2. Sex, age, history of cerebral infarction, T2DM, hypertension and atrial fibrillation were also significantly associated with the primary outcome.
Table 2. Univariate regression analysis for post-OPCABG cerebral infarction stratified by TyG Index.
| Characteristic | Cerebral infarction | |
|---|---|---|
| Univariate OR (95% CI) | P value | |
| TyG | 1.87 (1.38 to 2.53) | <0.001 |
| Age | 1.04 (1.01 to 1.06) | 0.005 |
| Female | 1.93 (1.25 to 2.99) | 0.003 |
| BMI | 0.96 (0.90 to 1.03) | 0.254 |
| Ischaemic stroke history | 2.36 (1.52 to 3.64) | <0.001 |
| Hyperlipidaemia history | 1.19 (0.78 to 1.81) | 0.428 |
| Hypertension history | 1.83 (1.13 to 2.95) | 0.014 |
| T2DM history | 1.90 (1.25 to 2.90) | 0.003 |
| AF history | 5.14 (1.52 to 17.40) | 0.009 |
| Myocardial infarction history | 0.98 (0.53 to 1.80) | 0.936 |
| Alcohol | 0.82 (0.50 to 1.36) | 0.445 |
| Smoking | 0.61 (0.39 to 0.96) | 0.033 |
P values in bold are <0.05.
AF, atrial fibrillation; BMI, body mass index; OPCABG, off-pump coronary artery bypass grafting; T2DM, type 2 diabetes mellitus; TyG Index, Triglyceride-Glucose Index.
Predictive value of TyG Index for perioperative stroke and other events
To further evaluate the robustness of the TyG Index in predicting postoperative cerebral infarction, we implemented multivariate regression analysis (table 3). After adjusting for all significant variables from the univariate analysis (history of cerebral infarction, hypertension, diabetes, atrial fibrillation and smoking), the TyG Index remained an independent risk factor for postoperative cerebral infarction following OPCABG. In Model 2, when the TyG Index was analysed as a continuous variable, each one-unit increase in the TyG Index was associated with a 1.80-fold increase in the incidence of the primary outcome (95% CI 1.29 to 2.51, p<0.001). Among the three tertiles of the TyG Index, the risk of postoperative cerebral infarction in the third tertile was 2.23 times higher than in the first tertile (95% CI 1.24 to 4.02, p=0.007). The value of p for trend results showed that the association between the TyG Index categories and the risk of postoperative cerebral infarction was statistically significant. These associations were not affected by collinearity: all predictors exhibited low VIFs (range ≈1.0–1.3; none≥2.5), indicating an absence of harmful multicollinearity (online supplemental figure S1). The correlation between TyG and diabetes history was modest (r=0.2717), and correlations between TyG and other covariates were small in magnitude (online supplemental figure S2). In addition, a significant linear relationship was observed between the TyG Index and the duration of surgery (ß=0.05, 95% CI 0.01 to 0.09) (online supplemental table S2), suggesting that higher TyG Index values were associated with longer surgical times. However, no significant association was found between the TyG Index and other perioperative complications, myocardial infarction, intensive care unit stay, total hospitalisation duration or in-hospital mortality. Finally, we performed a restricted cubic spline regression analysis (figure 2), revealing a clear, linear, positive correlation between the TyG Index and the incidence of postoperative stroke (value of p for non-linearity=0.861).
Table 3. Multivariate logistic regression analysis for post-OPCABG cerebral infarction stratified by TyG Index.
| TyG Index | OR (95% CI) | |
|---|---|---|
| Model 1 | Model 2 | |
| Per unit increase | 1.93 (1.41 to 2.63)* | 1.80 (1.29 to 2.51)* |
| Tertile 1 | Reference (1) | Reference (1) |
| Tertile 2 | 1.42 (0.77 to 2.62) | 1.30 (0.70 to 2.41) |
| Tertile 3 | 2.59 (1.48 to 4.53)* | 2.23 (1.24 to 4.02)* |
| Value of p for trend | <0.001 | 0.005 |
Model 1: adjusted for age, sex.
Model 2: adjusted for history of cerebral infarction, hypertension, type 2 diabetes mellitus, atrial fibrillation, smoking.
Values of p in bold are <0.05.
Value of p<0.05.
OPCABG, off-pump coronary artery bypass grafting; TyG Index, Triglyceride-Glucose Index.
Figure 2. Restricted cubic spline regression for the impact of TyG Index on postoperative cerebral infarction. TyG Index, Triglyceride-Glucose Index.
Subgroup analysis
We systematically divided the research cohort into six subgroups based on the history of hyperlipidaemia, T2DM, hypertension, age (>70), BMI (≥24) and gender. According to the stratification results in figure 3, no interactions were observed among the subgroups (value of p for interaction >0.05). The absence of interaction is crucial as it underscores the broad applicability of the TyG Index in predicting perioperative neurological complications across different patient populations and medical histories.
Figure 3. Subgroup analysis of the impact of Triglyceride-Glucose Index (TyG Index) on postoperative cerebral infarction. BMI, body mass index. Values of p in bold are <0.05.
Incremental predictive value of the TyG Index
The predictive model for postoperative stroke demonstrated improved discriminatory performance by including the TyG Index. As shown in figure 4, the AUC for the model incorporating the TyG Index was higher than that without the TyG Index (0.724 vs 0.708, p<0.001). Visual inspection of calibration plots suggested good agreement between predicted and observed risks for both models across deciles (online supplemental figure S3A,B). The Hosmer–Lemeshow goodness-of-fit test did not indicate miscalibration for either the conventional model (Model 2: df=8, p=0.642) or the TyG-augmented model (TyG+M2: df=8, p=0.277), consistent with acceptable calibration in our cohort (online supplemental table S3). However, evaluation of NRI and IDI observed no significant improvement in the models’ reclassification ability (table 4).
Figure 4. Receiver operating characteristics curves for the prediction of postoperative cerebral infarction. Conventional model: sex, age, ischaemic stroke history, hypertension history, type 2 diabetes mellitus history, atrial fibrillation history. TyG+Conventional Model: sex, age, ischaemic stroke history, hypertension history, type 2 diabetes mellitus history, atrial fibrillation history, TyG Index. AUC, area under the receiver operating characteristics curve; TyG Index, Triglyceride-Glucose Index.
Table 4. The incremental predictive ability of the TyG Index.
| Conventional risk model | TyG Index + conventional risk model | P value | |
|---|---|---|---|
| C-statistics | 0.708 (0.630 to 0.753) | 0.724 (0.630 to 0.746) | <0.001 |
| NRI | Reference | 0.000 (−0.007 to 0.007) | 0.520 |
| IDI | Reference | 0.000 (−0.007 to 0.007) | 0.519 |
P values in bold are <0.05.
IDI, integrated discrimination improvement; NRI, net reclassification improvement.
Discussion
In this multicentre cohort study, which included 3654 patients with severe CAD, the TyG Index was used to assess IR in patients undergoing OPCABG and to explore its association with the incidence of postoperative cerebral infarction and other complications. The main findings are as follows: the elevated TyG Index might be an independent risk factor for postoperative cerebral infarction following OPCABG. After adjusting for confounding factors, each one-unit increase in the TyG Index was associated with nearly a twofold increase in the risk of post-OPCABG cerebral infarction. This association remained significant across all subgroups. Second, regarding other perioperative adverse events, we observed a significant linear correlation between the TyG Index and the duration of OPCABG surgery. Furthermore, incorporating the TyG Index produced a modest improvement in discrimination, although risk reclassification did not significantly improve. Importantly, these findings underscore the potential value of the TyG Index in predicting perioperative complications in OPCABG patients, providing surgeons with a simple and cost-effective tool to assess IR and predict the risk of adverse outcomes.
IR, which is defined as a decreased sensitivity to insulin at normal plasma glucose levels, could initiate its detrimental pathological changes in the cardiovascular system well before the development of T2DM.18 Correspondingly, in previous studies, IR has been proven to be significantly associated with poor prognosis in patients with CAD.5 However, conventional IR measurements (eg, the hyperinsulinaemic-euglycaemic clamp test and homoeostasis model) are generally not recommended to clinicians due to their complexity and high cost.8 19 20
The TyG Index could be considered an economical and practical tool for the early identification of IR in clinical practice, as it is simply calculated from TG, a major marker of lipid metabolism, and FBG, a key indicator of glucose metabolism, and could reliably indicate the extent of IR.8 18 Several studies have demonstrated that the TyG Index is a robust IR predictor. For instance, the study by Guerrero-Romero et al showed that the TyG Index, when used as a diagnostic tool for IR, had a sensitivity of 96% and a specificity of 85% compared with other gold standard methods.8
Recent studies have demonstrated a significant association between higher TyG Index values and the progression of coronary atherosclerosis and calcification, which could ultimately lead to a 1.44-fold increase in cardiovascular mortality.21 22 Additionally, IR has been linked to adverse outcomes following revascularisation, including elevated levels of troponin I and increased cardiovascular mortality.6 In the context of CABG, a study based on the US Department of Veterans Affairs found that patients with metabolic syndrome, which includes IR, had a 1.3-fold higher risk of myocardial infarction within 6 years after CABG compared with those without metabolic syndrome. However, this study focused on the broader spectrum of metabolic syndrome and did not specifically address the impact of IR on CABG prognosis.12 Furthermore, Zhang et al used the TyG Index as a marker for IR. They demonstrated that a predictive model incorporating the TyG Index achieved a high area under the curve of 0.93 in predicting the 5-year incidence of major adverse cardiovascular events after CABG.10 23 While these studies have focused on relatively long-term cardiovascular outcomes, there has been limited investigation into the predictive value of the TyG Index for perioperative complications following OPCABG, such as postoperative cerebral infarction.14 The current research answers the need to further explore the potential role of the TyG Index in predicting postoperative complications in OPCABG patients.
In the current study, we employed various methods to analyse the relationship between the TyG Index and the incidence of perioperative complications in patients undergoing OPCABG, revealing a significant positive correlation between the TyG Index and the incidence of postoperative cerebral infarction. In OPCABG—without the CPB circuit—the leading pathways to perioperative ischaemic stroke are aortic manipulation-related atheroembolism and haemodynamic instability-related watershed injury, with additional contributions from vessel thrombosis.24 25 Our observation that higher TyG—a surrogate of IR—associates with increased postoperative stroke risk accords with an IR-related milieu of hypercoagulability, endothelial dysfunction and systemic inflammation that may amplify these mechanism-specific hazards. This perspective aligns with known pathophysiological links between IR and cerebrovascular risk. For instance, carotid stenosis is a significant risk factor for cerebral infarction, and the TyG Index was significantly associated with carotid atherosclerosis, intima-media thickness, plaque and the severity of stenosis.22 26 This suggests that IR may indirectly increase the risk of stroke by affecting the extent of carotid lesions. The progression of aortic plaque is also promoted by IR27 28 because IR exacerbates inflammatory responses and oxidative stress processes, thereby accelerating endothelial damage.29 Thus, IR significantly increases the risk of cerebral infarction due to the elevated embolisation risk of aortic plaque during surgery. Additionally, particularly in patients with old myocardial infarction lesions, IR promotes the formation of left ventricular mural thrombus by facilitating thrombin synthesis and platelet aggregation, leading to a disorder in the coagulation-fibrinolysis system.30 31 Future studies are warranted to elucidate how IR increases the risk of perioperative cerebral infarction in OPCABG patients. Additionally, we found that an increase in the TyG Index was associated with a longer surgery duration. IR may lengthen the operative time by increasing technical complexity, vascular fragility and metabolic vulnerability.32 Longer procedures are likely to extend patients’ exposure to haemodynamic instability, thereby potentially increasing the risk of postoperative ischaemic stroke.33 However, no differences were observed in other intraoperative variables, such as blood loss or the need for transfusion. Therefore, whether IR leads to more intraoperative uncertainties remains to be investigated by further research.
This study also discovered a linear relationship between the TyG Index and the incidence of postoperative stroke, highlighting the importance of assessing the TyG Index before OPCABG surgery. By evaluating this marker, surgeons could adopt more targeted strategies to manage IR in high-risk patients, reducing the likelihood of adverse cerebral and cardiovascular events.34 35 For instance, medications such as glucagon-like peptide-1 receptor agonists and sodium-glucose cotransporter-2 inhibitors can significantly improve insulin sensitivity and thus may reduce the occurrence of perioperative adverse events in CABG patients. Our study also revealed that individuals with elevated TyG Indexes exhibit a higher prevalence of cardiovascular disease risk factors, necessitating the implementation of pharmacotherapies for secondary prevention (eg, antiplatelet therapy) and the optimisation of blood pressure and lipid profiles.
Strengths and limitations
This study has notable strengths and limitations. Unlike previous research involving mixed cardiac surgeries,11 36 this is the first to focus solely on OPCABG, thereby avoiding the confounding effects of cardiopulmonary bypass on postoperative stroke. It used a multicentre Chinese database, enhancing the generalisability of findings across diverse regions. However, due to its retrospective nature, key confounders—such as intraoperative aortic manipulation and anaesthesia duration—could not be fully controlled, and gold standard IR testing was not feasible. In addition, we could not definitively separate mediation from confounding regarding operative time; therefore, residual pathway bias is possible. Missingness was low and handled with MICE and mode imputation; nevertheless, residual bias may persist if the data were not missing completely at random. Moreover, the exclusive use of Chinese patient data may limit global applicability. Future prospective, multicentre studies with larger cohorts and precise IR assessment are warranted.
Conclusions
This is the first study to reveal that an elevated TyG Index predicts the risk of postoperative cerebral infarction and other perioperative complications in patients undergoing OPCABG. These findings imply that interventions targeting IR, as evaluated by the TyG Index, may decrease the occurrence of perioperative cerebral infarction and other unfavourable outcomes in patients undergoing OPCABG, ultimately improving patient outcomes.
Supplementary material
Acknowledgements
The authors thank AiMi Academic Services (www.aimieditor.com) for English language editing and review services.
Footnotes
Funding: This study was supported by the Beijing Municipal Science and Technology Commission, Administrative Commission of Zhongguancun Science Park (No. Z221100007422015 and Z241100007724008), Beijing Municipal Administration of Hospitals Clinical Medicine Development of Special Funding Support (No. ZLRL202317), Beijing Natural Science Foundation (No. 7232037 and L232030), Beijing Advanced Innovation Centre for Big Data-based Precision Medicine (No. PXM2021_014226_000026), Science and Technology Foundation of Beijing Anzhen Hospital (No. KCGY2023), National Natural Science Foundation of China (No.82470495) and Beijing Anzhen Hospital High Level Research Funding (No.2024AZB2001). HYL received funding from the Beijing Municipal Science and Technology Commission, Administrative Commission of Zhongguancun Science Park, Beijing Natural Science Foundation, Science and Technology Foundation of Beijing Anzhen Hospital, National Natural Science Foundation of China, and the High-Level Research Funding of Beijing Anzhen Hospital. HYL provided critical support in the conception and design of the study, its implementation, and the decision to submit the manuscript for publication. HJZ received support from the Beijing Municipal Science and Technology Commission, Administrative Commission of Zhongguancun Science Park, the Clinical Medicine Development Special Funding Support from the Beijing Municipal Administration of Hospitals, and the Beijing Advanced Innovation Centre for Big Data-based Precision Medicine. HJZ provided important administrative support and contributed study materials and patient data.
Provenance and peer review: Not commissioned; internally peer reviewed.
Patient consent for publication: Not applicable.
Ethics approval: This study involves human participants and was approved by the Ethics Review Committee of Beijing Anzhen Hospital, Capital Medical University (approval number: KS2023090). This study was based on previously existing medical records, and the study subjects could not be contacted. Moreover, no personal information or commercial interests were involved in the data analysis or reporting process; therefore, informed consent could be waived.
Data availability free text: The data sets used and analysed during the current study are available from the corresponding author on reasonable request.
Data availability statement
Data are available upon reasonable request.
References
- 1.Lawton JS, Tamis-Holland JE, et al. Writing Committee Members ACC/AHA/SCAI Guideline for Coronary Artery Revascularization: A Report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. J Am Coll Cardiol. 2021;79:e21–129. doi: 10.1016/j.jacc.2021.09.006. [DOI] [PubMed] [Google Scholar]
- 2.Gaudino M, Angiolillo DJ, Di Franco A, et al. Stroke After Coronary Artery Bypass Grafting and Percutaneous Coronary Intervention: Incidence, Pathogenesis, and Outcomes. J Am Heart Assoc. 2019;8:e013032. doi: 10.1161/JAHA.119.013032. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.ElBardissi AW, Aranki SF, Sheng S, et al. Trends in isolated coronary artery bypass grafting: An analysis of the Society of Thoracic Surgeons adult cardiac surgery database. J Thorac Cardiovasc Surg. 2012;143:273–81. doi: 10.1016/j.jtcvs.2011.10.029. [DOI] [PubMed] [Google Scholar]
- 4.Reaven GM. Banting lecture 1988. Role of insulin resistance in human disease. Diabetes. 1988;37:1595–607. doi: 10.2337/diab.37.12.1595. [DOI] [PubMed] [Google Scholar]
- 5.Laakso M, Kuusisto J. Insulin resistance and hyperglycaemia in cardiovascular disease development. Nat Rev Endocrinol. 2014;10:293–302. doi: 10.1038/nrendo.2014.29. [DOI] [PubMed] [Google Scholar]
- 6.Uetani T, Amano T, Harada K, et al. Impact of insulin resistance on post-procedural myocardial injury and clinical outcomes in patients who underwent elective coronary interventions with drug-eluting stents. JACC Cardiovasc Interv. 2012;5:1159–67. doi: 10.1016/j.jcin.2012.07.008. [DOI] [PubMed] [Google Scholar]
- 7.Du T, Yuan G, Zhang M, et al. Clinical usefulness of lipid ratios, visceral adiposity indicators, and the triglycerides and glucose index as risk markers of insulin resistance. Cardiovasc Diabetol. 2014;13:146. doi: 10.1186/s12933-014-0146-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Guerrero-Romero F, Simental-Mendía LE, González-Ortiz M, et al. The product of triglycerides and glucose, a simple measure of insulin sensitivity. Comparison with the euglycemic-hyperinsulinemic clamp. J Clin Endocrinol Metab. 2010;95:3347–51. doi: 10.1210/jc.2010-0288. [DOI] [PubMed] [Google Scholar]
- 9.Wang S, Shi J, Peng Y, et al. Stronger association of triglyceride glucose index than the HOMA-IR with arterial stiffness in patients with type 2 diabetes: a real-world single-centre study. Cardiovasc Diabetol. 2021;20:82. doi: 10.1186/s12933-021-01274-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Chen L, Ding X-H, Fan K-J, et al. Association Between Triglyceride-Glucose Index and 2-Year Adverse Cardiovascular and Cerebrovascular Events in Patients with Type 2 Diabetes Mellitus Who Underwent Off-Pump Coronary Artery Bypass Grafting. Diabetes Metab Syndr Obes. 2022;15:439–50. doi: 10.2147/DMSO.S343374. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Wu Z, Xie L, Guo D, et al. Triglyceride-glucose index in the prediction of adverse cardiovascular events in patients without diabetes mellitus after coronary artery bypass grafting: a multicenter retrospective cohort study. Cardiovasc Diabetol. 2023;22:230. doi: 10.1186/s12933-023-01969-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Deo SV, Sundaram V, Wilson B, et al. Adverse events after coronary artery bypass grafting in patients with preoperative metabolic syndrome: A 10-year follow-up of the Veterans Affairs Database. J Thorac Cardiovasc Surg. 2022;163:2096–103. doi: 10.1016/j.jtcvs.2020.08.018. [DOI] [PubMed] [Google Scholar]
- 13.Yang Y, Huang X, Wang Y, et al. The impact of triglyceride-glucose index on ischemic stroke: a systematic review and meta-analysis. Cardiovasc Diabetol. 2023;22 doi: 10.1186/s12933-022-01732-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Norris P, Gow J, Arthur T, et al. Metabolic syndrome and surgical complications: a systematic review and meta-analysis of 13 million individuals. Int J Surg. 2024;110:541–53. doi: 10.1097/JS9.0000000000000834. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Sacco RL, Kasner SE, Broderick JP, et al. An updated definition of stroke for the 21st century: a statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke. 2013;44:2064–89. doi: 10.1161/STR.0b013e318296aeca. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Gaudino M, Dangas GD, Angiolillo DJ, et al. Considerations on the Management of Acute Postoperative Ischemia After Cardiac Surgery: A Scientific Statement From the American Heart Association. Circulation. 2023;148:442–54. doi: 10.1161/CIR.0000000000001154. [DOI] [PubMed] [Google Scholar]
- 17.Bushnell C, Kernan WN, Sharrief AZ, et al. 2024 Guideline for the Primary Prevention of Stroke: A Guideline From the American Heart Association/American Stroke Association. Stroke . 2024;55 doi: 10.1161/STR.0000000000000475. [DOI] [PubMed] [Google Scholar]
- 18.Laakso M. Cardiovascular Disease in Type 2 Diabetes From Population to Man to Mechanisms. Diabetes Care. 2010;33:442–9. doi: 10.2337/dc09-0749. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Pacini G, Mari A. Methods for clinical assessment of insulin sensitivity and beta-cell function. Best Pract Res Clin Endocrinol Metab. 2003;17:305–22. doi: 10.1016/s1521-690x(03)00042-3. [DOI] [PubMed] [Google Scholar]
- 20.Rudvik A, Månsson M. Evaluation of surrogate measures of insulin sensitivity - correlation with gold standard is not enough. BMC Med Res Methodol. 2018;18:64. doi: 10.1186/s12874-018-0521-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Won K-B, Park EJ, Han D, et al. Triglyceride glucose index is an independent predictor for the progression of coronary artery calcification in the absence of heavy coronary artery calcification at baseline. Cardiovasc Diabetol. 2020;19:34. doi: 10.1186/s12933-020-01008-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Li W, Chen D, Tao Y, et al. Association between triglyceride-glucose index and carotid atherosclerosis detected by ultrasonography. Cardiovasc Diabetol. 2022;21:137. doi: 10.1186/s12933-022-01570-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Zhang H, Chong H, Li Z, et al. Triglyceride−glucose index in the prediction of major adverse cardiovascular events in patients with type 2 diabetes mellitus after coronary artery bypass surgery: A retrospective cohort study. Front Endocrinol. 2022;13:1015747. doi: 10.3389/fendo.2022.1015747. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Likosky DS, Marrin CAS, Caplan LR, et al. Determination of etiologic mechanisms of strokes secondary to coronary artery bypass graft surgery. Stroke. 2003;34:2830–4. doi: 10.1161/01.STR.0000098650.12386.B3. [DOI] [PubMed] [Google Scholar]
- 25.Baker AJ, Naser B, Benaroia M, et al. Cerebral microemboli during coronary artery bypass using different cardioplegia techniques. Ann Thorac Surg. 1995;59:1187–91. doi: 10.1016/0003-4975(95)00128-8. [DOI] [PubMed] [Google Scholar]
- 26.Miao M, Zhou G, Bao A, et al. Triglyceride-glucose index and common carotid artery intima-media thickness in patients with ischemic stroke. Cardiovasc Diabetol. 2022;21:43. doi: 10.1186/s12933-022-01472-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Semenkovich CF. Insulin resistance and atherosclerosis. J Clin Invest. 2006;116:1813–22. doi: 10.1172/JCI29024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Ormazabal V, Nair S, Elfeky O, et al. Association between insulin resistance and the development of cardiovascular disease. Cardiovasc Diabetol. 2018;17:122. doi: 10.1186/s12933-018-0762-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Beverly JK, Budoff MJ. Atherosclerosis: Pathophysiology of insulin resistance, hyperglycemia, hyperlipidemia, and inflammation. J Diabetes. 2020;12:102–4. doi: 10.1111/1753-0407.12970. [DOI] [PubMed] [Google Scholar]
- 30.Stegenga ME, van der Crabben SN, Levi M, et al. Hyperglycemia stimulates coagulation, whereas hyperinsulinemia impairs fibrinolysis in healthy humans. Diabetes. 2006;55:1807–12. doi: 10.2337/db05-1543. [DOI] [PubMed] [Google Scholar]
- 31.Undas A, Wiek I, Stêpien E, et al. Hyperglycemia is associated with enhanced thrombin formation, platelet activation, and fibrin clot resistance to lysis in patients with acute coronary syndrome. Diabetes Care. 2008;31:1590–5. doi: 10.2337/dc08-0282. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Armstrong EJ, Rutledge JC, Rogers JH. Coronary artery revascularization in patients with diabetes mellitus. Circulation . 2013;128:1675–85. doi: 10.1161/CIRCULATIONAHA.113.002114. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Gregory A, Stapelfeldt WH, Khanna AK, et al. Intraoperative Hypotension Is Associated With Adverse Clinical Outcomes After Noncardiac Surgery. Anesth Analg. 2021;132:1654–65. doi: 10.1213/ANE.0000000000005250. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Andrikou E, Tsioufis C, Andrikou I, et al. GLP-1 receptor agonists and cardiovascular outcome trials: An update. Hellenic J Cardiol. 2019;60:347–51. doi: 10.1016/j.hjc.2018.11.008. [DOI] [PubMed] [Google Scholar]
- 35.Abu-Qaoud MR, Kumar A, Tarun T, et al. Impact of SGLT2 Inhibitors on AF Recurrence After Catheter Ablation in Patients With Type 2 Diabetes. JACC Clin Electrophysiol. 2023;9:2109–18. doi: 10.1016/j.jacep.2023.06.008. [DOI] [PubMed] [Google Scholar]
- 36.Huang R, Xu X, Xu C, et al. Association between the insulin resistance and all-cause mortality in patients with moderate and severe aortic stenosis: a retrospective cohort study. Cardiovasc Diabetol. 2023;22:238. doi: 10.1186/s12933-023-01975-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data are available upon reasonable request.