Triglyceride-glucose Index Predicts 2-year Adverse Cardiovascular and Cerebrovascular Events in Patients With Type 2 Diabetes Mellitus and Non-st-segment Elevation Acute Coronary Syndrome Who Underwent Off-pump Coronary Artery Bypass Grafting

Background: Data on the relationship between the triglyceride glucose (TyG) index and prognosis after off-pump coronary artery bypass grafting (OPCABG) are limited. This retrospective observational study evaluated the association of the TyG index with prognosis in patients with type 2 diabetes mellitus (T2DM) and non-ST-segment elevation acute coronary syndrome (NSTE-ACS) who underwent OPCABG. Methods: A total of 1534 patients with T2DM who underwent OPCABG for NSTE-ACS were enrolled in this study. Patients were divided into lower and higher TyG index groups according to the optimal cutoff value of the TyG index for predicting the primary endpoint. The primary outcomes included the occurrence of major adverse cardiovascular and cerebrovascular events (MACCEs), which were dened as all-cause death, nonfatal myocardial infarction, nonfatal stroke and symptomatic graft stenosis or occlusion. The TyG index was calculated using the following equation: TyG index = ln (fasting triglyceride level [mg/dL] × fasting glucose level [mg/dL]/2). Results: The overall incidence of MACCEs was 10.7% with a median follow-up of 23.1 months. The optimal TyG index cutoff for predicting MACCEs was 9.42 (sensitivity 49.4%; specicity 71.2%; area under the curve 0.631; P < 0.001). The incidence of MACCEs increased with a 1-unit increase in the TyG index [HR 2.218, 95% CI (1.733-2.839), P < 0.001]. A TyG index greater than 9.42 was independently associated with a higher risk of MACCEs [HR 2.092, 95% CI (1.573-2.784), P < 0.001] and each component. Furthermore, the effect of TyG in increasing MACCEs was consistent across groups of age, sex, BMI, hypertension, LDL-C, oral hypoglycemic agents, LIMA, number of grafts and IABP groups, except in patients diagnosed with NSTEMI, treated with statins and prescribed insulin at discharge. Conclusion: The TyG index predicts future MACCEs independently of known cardiovascular risk factors, suggesting that the TyG index may be a useful marker for prognosis in patients with T2DM and NSTE-ACS who underwent OPCABG. BMI body mass index, SBP systolic blood pressure, DBP diastolic blood pressure, MI myocardial infarction, PCI percutaneous coronary intervention, PVD peripheral vascular disease, TGs triglycerides, TC total cholesterol, LDL-C low-density lipoprotein cholesterol, HDL-C high-density lipoprotein cholesterol, hs-CRP high-sensitivity C-reactive protein, eGFR estimated glomerular ltration rate, FPG fasting plasma glucose, GA glycated albumin, TyG triglyceride glucose, LVEF left ventricular ejection fraction, UA unstable angina, NSTEMI non-ST-segment elevation myocardial infarction, LIMA left internal mammary artery, SVG saphenous vein grafts, ACEI angiotensin converting enzyme inhibitor, ARB angiotensin receptor blocker, DAPT dual antiplatelet therapy, DPP-4 Dipeptidyl peptidase-4, IABP intra-aortic balloon pump BMI body mass index, SBP systolic blood pressure, DBP diastolic blood pressure, MI myocardial infarction, PCI percutaneous coronary intervention, PVD peripheral vascular disease, TGs triglycerides, TC total cholesterol, LDL-C low-density lipoprotein cholesterol, HDL-C high-density lipoprotein cholesterol, hs-CRP high-sensitivity C-reactive protein, eGFR estimated glomerular ltration rate, FPG fasting plasma glucose, GA glycated albumin, TyG triglyceride glucose, LVEF left ventricular ejection fraction, UA unstable angina, NSTEMI non-ST-segment elevation myocardial infarction, LIMA left internal mammary artery, SVG saphenous vein grafts, ACEI angiotensin converting enzyme inhibitor, ARB angiotensin receptor blocker, DAPT dual antiplatelet therapy, DPP-4 Dipeptidyl peptidase-4, IABP intra-aortic balloon pump Italic values indicate statistically signicant associations blood DBP diastolic blood pressure, PCI percutaneous coronary intervention, PVD peripheral vascular disease, TGs triglycerides, TC total cholesterol, LDL-C low-density lipoprotein cholesterol, HDL-C high-density lipoprotein cholesterol, hs-CRP high-sensitivity C-reactive protein, eGFR estimated glomerular ltration rate, FPG fasting plasma glucose, GA glycated albumin, TyG triglyceride glucose, LVEF left ventricular ejection fraction, UA unstable angina, NSTEMI non-ST-segment elevation myocardial infarction, IABP intra-aortic balloon pump Primary and secondary outcomes are all signicantly different between the lower and higher TyG index groups


Introduction
Coronary artery bypass grafting (CABG) remains the preferred treatment option for patients with severe multiple vessel coronary artery disease [1]. Although surgical techniques and medical treatment have been developed, the risk of recurrent adverse cardiovascular and cerebrovascular events (MACCEs) remains relatively high in patients who have had acute coronary syndrome (NSTE-ACS) [2].
The triglyceride-glucose (TyG) index, a product of triglycerides and fasting plasma glucose (FPG), is a good tool that correlates with surrogate and direct measures of insulin resistance (IR) [3,4]. Previous studies have reported that the TyG index is signi cantly associated with an increased risk of developing T2DM, hypertension and adverse cardiovascular events [5][6][7][8]. Moreover, several cross-sectional studies have reported a strong relation between the TyG index and atherosclerosis in different clinical conditions [9][10][11]. However, whether TyG may be associated with the adverse CV events of patients with T2DM and NSTE-ACS who were treated with OPCABG has not yet been assessed. Therefore, in the present study, we determined by appropriate guidelines [14]. NSTEMI was de ned as having symptoms of ischemia and elevated cardiac troponin I (cTnI) without an elevation of the ST segment. UA was diagnosed as ischemic symptoms at rest and exacerbated or new-onset symptoms with transient ischemic ST-segment shifts without release of myocardial enzymes related to myocardial necrosis. Hypertension was de ned as systolic blood pressure ≥ 140 mmHg or diastolic blood pressure ≥ 90 mmHg, a previous diagnosis of hypertension or the use of antihypertensive medication. Peripheral vascular disease (PVD) was de ned as aorta and arteries other than coronary arteries with exercise-related claudication, reduced or absent pulsation or angiographic stenosis of more than 50%.

Follow-up and endpoint event
In January 2021, all patients were followed up by trained professionals who were blinded to the baseline information. Information about adverse prognostic events was obtained from patients or their family members by telephone questionnaire. The information was further con rmed by careful veri cation of corresponding medical records if necessary. The primary observational endpoint was de ned as MACCE consisting of four events as follows: all-cause death, nonfatal myocardial infarction (MI), nonfatal stroke and symptomatic graft stenosis or occlusion. The secondary observational endpoint was each component of the composite primary endpoint. MI was de ned as elevated cardiac troponin higher than the upper reference limit with ischemia indicated by symptoms and/or electrocardiographic changes with or without an elevation of the ST segment. Symptomatic graft stenosis or occlusion was de ned as having typical angina symptoms again after discharge and identi cation of graft stenosis or occlusion by coronary CTA scan. The rst primary endpoint event that occurred during the follow-up was used for analysis in the current study. For patients with multiple adverse outcomes occurring almost simultaneously during the follow-up, only the most severe event (all-cause death > nonfatal MI > nonfatal stroke > symptomatic graft stenosis or occlusion) was selected for the analyses. If the same event occurred multiple times, only the rst occurrence was used for the analyses.

Statistical analysis
Continuous variables are presented as the mean ± standard deviation (SD) or median (25th and 75th percentiles: P25, P75) in the case of normal or nonnormal distribution, and differences between the two groups were examined by independent-sample t-test or Mann-Whitney U test correspondingly.
Categorical variables are described as counts (percentages) and compared by Pearson chi-square test (Pearson χ 2 test) with Yates' continuity correction. Spearman's rank correlation test or Pearson correlation test was used to evaluate the correlations between the TyG index and cardiovascular and cerebrovascular risk factors when appropriate. The Pearson correlation test was used to evaluate the correlation between two continuous variables with a normal distribution, while Spearman's rank correlation test was applied in cases where one or more of the variables being analyzed were nonnormally distributed continuous variables or categorical variables. Receiver operating characteristic (ROC) curve analysis was performed to determine the optimal cutoff point value of the TyG index for predicting the primary endpoint. Kaplan-Meier survival analyses were performed to evaluate the incidence rate of adverse events between groups according to the optimal cutoff point of the TyG index, and discrepancies between groups were evaluated by the log-rank test. The predictive value of the variables for the primary endpoint was evaluated by univariate and multivariate Cox proportional hazards analyses. The TyG index was analyzed in the following two ways: (1) as a categorical variable and (2) as a continuous variable. In the multivariate Cox proportional hazards analyses, four models were established to evaluate the predictive value of the TyG index for primary endpoint. Among these models, confounders were selected according to statistical signi cance (P < 0.2) in the univariate analysis and clinical importance. The models were as follows: Model 1, adjusted for age, sex (female), BMI, current smoking status, current drinking status, dyslipidemia, prior MI, prior PCI, prior stroke and PVD; Model 2, adjusted for variables included in Model 1 in addition to diagnosis (NSTEMI), TC, HDL-C, hs-CRP, creatinine, eGFR, uric acid, GA and LVEF; Model 3, adjusted for variables included in Model 2 in addition to multiple grafts, LIMA use and IABP use; and Model 4, adjusted for variables included in Model 3 in addition to DAPT at discharge, statin at discharge, oral hypoglycemic agents (metformin, alpha-glucosidase inhibitor, sulfonylurea or dipeptidyl peptidase 4 inhibitor) at discharge and insulin at discharge. Further strati ed analyses according to age (≤ 65 and > 65 years), sex, BMI (≤ 28 and > 28 kg/m 2 ), hypertension, initial diagnosis (UA and NSTEMI), LDL-C (≤ 70 and > 70 mg/dL), LVEF (≤ 40% and > 40%), medication at discharge including statins, oral hypoglycemic agents, insulin, type of grafts (LIMA-and LIMA+), number of grafts (one or two grafts and three or more grafts) and IAPB use were employed to examine the consistency of the prognostic impact of TyG index for primary endpoint. The model used in the strati ed analyses consisted of all covariates used in Model 4, except for the variables that were used for strati cation. The interaction of the TyG index and variables used for strati cation was examined by likelihood ratio tests.
The prognostic impact of the TyG index for each component of the primary endpoint was also assessed by using Model 4. FPG and TGs were not included in the multivariate analysis because they were used to calculate the TyG index. The results of Cox proportional hazards analyses are presented as hazard ratios (HRs) and 95% con dence intervals (CIs).
Statistical tests were performed with the R Programming Language (version 4.0.5). A two-tailed P value < 0.05 was regarded as statistically signi cant.

Results
Baseline characteristics of patients A total of 1534 admitted hospital patients were included with a median follow-up time of 23.1 months (interquartile range, 17.4-29.5 years) and a mean age of 62.7 ± 8.2 years old. Patients who experienced events had the following characteristics: older; higher levels of TGs, TC, LDL-C, eGFR, GA, Creatinine, FPG and TyG index; more likely to use multiple SVGs rather than LIMA + SVGs; and more likely to have lipidregulating agents than those who had no events (Table 1). Left ventricular ejection fraction, IABP support and medication at discharge were similar between these two groups. ROC curve analysis showed that the AUC of the TyG index for predicting the primary endpoint was 0.631 (95% CI 0.584-0.677, P < 0.001). The optimal cutoff point of the TyG index was 9.42 with a sensitivity of 49.4% and a speci city of 71.2%. Patients with a higher TyG index were younger, had signi cantly higher BMI levels and had a higher proportion of dyslipidemia. Laboratory indexes, including TG, TC, LDL-C, hs-CRP, uric acid, FPG and GA, were signi cantly higher in patients with a higher TyG index, while HDL-C levels were relatively lower in these patients. However, Creatinine and eGFR, which were different between patients with and without events, were not signi cant between lower and higher TyG groups. In the higher TyG index group, more patients were diagnosed with NSTEMI (Table 2). Left ventricular ejection fraction, IABP support and medication at discharge were also similar between these two groups. The TyG index was positively correlated with sex (female), BMI, DBP, TGs, TC, LDL-C, hs-CRP, creatinine, uric acid, FPG, GA, diagnosis (NSTEMI) and dyslipidemia but negatively correlated with age, HDL-C and eGFR (Table 3).  Table 4). Kaplan-Meier curves for the incidence of the primary endpoint and each component according to the optimal cutoff point of the TyG index are shown in Fig. 2. Primary endpoint, all-cause death, nonfatal MI, nonfatal stroke and symptomatic graft stenosis and occlusion all showed a signi cant difference between the lower and higher TyG index groups ( Fig. 2; primary endpoint log-rank P < 0.001; all secondary endpoints log-rank P < 0.05).

Cox proportional hazard analyses of the prognostic implication of the TyG index
In the multivariate Cox proportional hazard analysis, four models (Models 1-4 as described above) comprised of variables that had statistical signi cance (P < 0.2) and/or clinical importance were constructed to evaluate the predictive potential of the TyG index for the primary endpoint. After adjusting for confounding variables, a higher TyG index remained an independent risk predictor of the primary endpoint despite regarding the TyG index as a nominal or continuous variable (all P < 0.001 in Models 1-4) ( Table 5). Detailed information on Model 4 is shown in Additional le 1: Table S1.  (Table 6).  (Fig. 3).

Discussion
The primary nding of the present study was that an increased TyG index is a strong indicator of MACCEs, including all-cause death, nonfatal MI and stroke, and symptomatic graft stenosis or occlusion, in patients with NSTE-ACS and T2DM who underwent OPCABG, even after adjustment for confounding risk factors. Compared to participants with a lower TyG index, those with a higher TyG index had an apparently higher incidence of MACCEs and each component.
Insulin resistance (IR) is traditionally measured by the homeostatic model assessment of IR (HOMA-IR) [15].  [21,24], while the cutoff value in the present study was 9.42. A TyG index higher than 9.42 indicates a higher risk of MACCEs after OPCABG in 2 years.
We also analyzed the predictive value by strati cation according to the use of statins, oral hypoglycemic agents and insulin, but no signi cant difference was found. Unexpectedly, the predictive value of the TyG index was more signi cant in patients with perioperative IABP insertion compared to patients without IABP insertion, indicating that patients using IABPs are more sensitive to an increase in the TyG index.
Thus, patients with IABP may have had more severe coronary disease and less coronary ow reserve [25], suggesting that little progress in treating coronary atherosclerosis may have a very large effect on heart function in the future.
The potential mechanism in the association of the  [27]. In the present study, participants with a higher TyG index tended to have more severe and complex clinical conditions in terms of higher lipid pro les and lower eGFR. The correlation analysis also showed that the TyG index is positively related to multiple risk factors for LDL-C and negative for eGFR. (2) Many studies have shown that patients with diabetes who underwent CABG have a higher rate of myocardial infarction than those without diabetes. In the present study, among diabetic patients, the higher TyG group had a higher incidence of acute myocardial infarction and symptomatic graft stenosis or occlusion. A higher TyG index may affect native coronary atherosclerosis because the TyG index has been demonstrated to be related to coronary artery calci cation [28] and arterial stiffness [29][30][31]. In addition, recent studies have found that IR is associated with coronary plaque vulnerability [32,33]. In contrast, a higher level of TyG may accelerate the atherosclerosis of grafts and accelerate the process of graft failure because IR has a proin ammatory and procoagulatory effect and is associated with endothelial dysfunction [34,35]. (3) Low S et al.
reported that FPG and TGs mainly re ect IR from the liver and adipose cells, respectively [7]. Therefore, the TyG index may re ect IR from these two aspects and thus be closely related to IR, which has been widely demonstrated to have a signi cant relationship with endothelial dysfunction, oxidative stress, cardiovascular remodeling, coagulation imbalance and in ammatory response, leading to late graft failure.
This study had several limitations. (1) The study was a single-center, retrospective, observational study in a highly selected cohort with strict exclusion criteria, and the sample size was relatively small, which may have weakened the power of the results. Further prospective, multicenter studies in a more extensive population with a larger sample size are needed to verify the present ndings. (2) The TyG index was assessed only once at admission. The changes in the TyG index during the follow-up period, which may have better predictive value for adverse prognosis, were not assessed in our study. (3) A certain proportion of participants received lipid-regulating therapy and antidiabetic treatment at admission, which may have had a potential impact on the TyG index and the study results. (4) Nearly all of the study population was Chinese patients. The results should be cautiously interpreted and expanded to Western populations as differences in metabolic levels exist among different races. (5) It is di cult to rule out that some patients may be complicated with undiagnosed systemic diseases, such as occult malignancies, which may have an impact on the assessment of prognosis. (6) We did not calculate the SYNTAX score and did not know whether complete revascularization occurred because most of the coronary angiograms were in optical discs from other hospitals, which prevented retrospective analysis. (7) The follow-up time was relatively short.

Conclusions
Based on retrospective studies, we concluded that the TyG index predicts future MACCEs in patients with T2DM and NSTE-ACS who underwent OPCABG independently of known cardiovascular risk factors. Importantly, this relationship remained signi cant after adjustment for other confounders, suggesting that the TyG index may be a useful marker for prognosis in patients with T2DM and NSTE-ACS who underwent OPCABG. Further prospective, large studies are required to con rm our ndings.

Declarations
Ethics approval and consent to participate This study was approved by the institutional review board of Beijing Anzhen Hospital, Capital Medical University. Given the retrospective nature of this study, the requirement for informed consent was waived.

Availability of data and materials
The datasets used during the current study are available from the corresponding author on reasonable request.

Con ict of Interest
The authors declare that the research was conducted in the absence of any commercial or nancial relationships that could be construed as a potential con ict of interest. This is a list of supplementary les associated with this preprint. Click to download. TableS1.docx