Introduction
Diabetes is strongly associated with cardiovascular (CV) disease, particularly coronary artery disease (CAD) [
1]. Chronic hyperglycemia is closely associated with an increased risk of adverse CV complications, thus reaching low levels of glycemia might improve the clinical outcomes for diabetes patients [
2]. However, it was still controversial about glucose control management in diabetes patients with CAD. Several clinical trials indicated that strict glucose control may result in unfavorable clinical outcomes with an increased risk of severe hypoglycemia [
1,
3,
4]. Therefore, it is important to identify the specific population that would benefit from glucose control management.
Insulin resistance has been considered as one important determinant of CV risk, which could predict future CV events directly [
5]. Triglyceride-glucose (TyG) index, calculated from fasting triglycerides (TG) and fasting blood glucose (FBG) levels, has recently been proposed as a simple and reliable indicator of insulin resistance [
6]. TyG index has been identified as a biomarker in predicting the prevalence and prognosis of CAD in the cohorts of CAD primary and secondary prevention population [
7‐
11].
However, the role of TyG in the glycemic control for diabetes patients with CAD remained unclear. Therefore, this study aimed to explore the relationship among TyG index, glycemic control status, and adverse CV events in diabetes patients with angiography-proven CAD.
Methods
Study design and population
This was a prospective cohort study conducted at Fuwai Hospital, Chinese Academy of Medical Sciences. The study protocol was complied with the Declaration of Helsinki and approved by the central ethics committee of Fuwai Hospital. Informed consent was obtained from all participants before the study was initiated.
Overall, from January 2017 to December 2018, 13,506 diabetes patients with angiography-proven CAD were consecutively recruited. Diabetes was recorded if the patient had a history of diabetes, received glucose-lowering therapy, had an FBG ≥ 7.0 mmol/L, glycosylated hemoglobin Alc (HbA1c) ≥ 6.5%, or 2 h plasma glucose ≥ 11.1 mmol/L in an oral glucose tolerance test [
12]. Angiography-proven CAD was defined as the presence of coronary stenosis ≥ 50% at least one major artery segment assessed by two experienced physicians according to the results of coronary angiography. Major exclusion criteria included missing detailed laboratory data (fasting TG and FBG), age < 18 or ≥ 80 years, severe hepatic or kidney dysfunction, decompensated heart failure, systemic inflammatory disease, malignant tumor, or acute infection.
Laboratory tests, echocardiography, and definition
On admission, blood samples were obtained from the cubital vein of each participation after at least 12 h of fasting. The concentrations of TG, total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), FBG, and creatinine were analyzed in an enzymatic assay by automated biochemical analyzer (Hitachi 7150, Tokyo, Japan). Low-density lipoprotein cholesterol (LDL-C) was calculated via the Friedewald method [
13]. HbA1c was measured with high-performance liquid chromatography (Tosoh G8 HPLC Analyzer; Tosoh Bioscience, Tokyo, Japan). The hsCRP was examined with standard biochemical techniques at the core laboratory of Fuwai Hospital. The estimated glomerular filtration rate (eGFR) was calculated using the Chinese-modified MDRD (Modification of Diet in Renal Disease) equation [
14]. The modified biplane Simpson rule was used to assess left ventricular ejection fraction (LVEF) at rest [
15].
The TyG index was calculated according to the following formula: Ln [fasting TG (mg/dL) × FBG (mg/dL)/2] [
16], and patients would be categorized according to baseline TyG tertiles (tertile 1 [T1]: <8.895; T2: 8.895-9.400 and T3: ≥9.400). According to the latest American Diabetes Association guideline, controlled glycemia was defined as targeting HbA1c levels less than 7% [
17]. Hypertension was defined as systolic blood pressure ≥ 140 mmHg, diastolic blood pressure ≥ 90 mmHg, or the use of antihypertensive therapy [
18].
Evaluation of CAD characteristics and management
Coronary angiogram was performed according to standard techniques by experienced interventional cardiologists. Two independent experienced interventional cardiologists reviewed angiographic data from the catheter laboratory of Fuwai Hospital and recorded the characteristics of CAD, including unique types of coronary stenosis, and the SYNergy between percutaneous coronary intervention with TAXus and cardiac surgery (SYNTAX) score.
Follow-up and clinical endpoints
Patients were followed up at 6-month intervals until 3-year duration after discharge from medical records, clinical visits, and/or telephone interviews by trained investigators who were blinded to the clinical data. The primary endpoint was CV events (a composite of CV death, nonfatal myocardial infarction [MI], and nonfatal stroke), and the major adverse cardiovascular event (MACE, a composite of CV death and nonfatal MI). The secondary endpoint was CV death, nonfatal MI, and nonfatal stroke. Death was considered CV-caused unless unequivocal non-CV cause could be established. Nonfatal MI was defined as positive cardiac troponins with typical chest pain, typical electrocardiogram serial changes, identification of an intracoronary thrombus by angiography or autopsy, or imaging evidence of new loss of viable myocardium or a new regional wall-motion abnormality [
19]. Nonfatal stroke was defined as a new focal neurological deficit lasting > 24 h confirmed by imaging evidence. The endpoints were confirmed by at least two professional physicians.
Statistical analysis
Continuous variables were expressed as mean ± SD. Categorical variables were presented as number (percentage). The Kolmogorov-Smirnov test was used to test the distribution pattern. The differences of baseline characteristics between groups were analyzed with the Student’s t-test, Mann-Whitney U test, Kruskal-Wallis H test, χ2-test, or Fisher exact test where appropriate. The cumulative incidence of clinical endpoints among groups was illustrated by the Kaplan-Meier curves and compared by the log-rank test. Restricted cubic spline (RCS) plots adjusted for age and sex were created to assess linearity assumptions of the relationship between TyG index and clinical endpoints. Univariable and multivariable Cox regression models were used to calculate the hazard ratios (HRs) and 95% confidence intervals (CIs). Potential confounding factors included in the multivariable Cox regression model were age, male sex, BMI, acute coronary syndrome (ACS) presentation, family history of CAD, previous MI, previous revascularization, hypertension, previous stroke, peripheral artery disease, current smoker, LVEF, serum creatinine, TC, HDL-C, LDL-C, hsCRP, SYNTAX score, chronic total occlusion lesion, aspirin use, statins use and insulin use. Subsequently, the relationships between the glycemic control status (controlled or uncontrolled glycemia) and CV events were evaluated according to TyG tertiles to explore the potential effect of TyG index on this association. These above analyses were made for the first subsequent event for all participants. Two tailed P values < 0.05 were considered as statistically significant. All statistical analyses were performed using R version 4.0.2 (The R Foundation).
Discussion
In this study, the association among TyG index, glycemic control, and adverse CV events in diabetes patients with angiography-proven CAD was evaluated, revealing that uncontrolled glycemia was significantly associated with CV events and MACEs in high TyG (T3) patients, while those association could not be seen in patients with low TyG index (T1 or T2). Our study demonstrated, for the first time, that the association between glycemic control and adverse CV events was more pronounced in high TyG patients, suggesting TyG could help making risk stratification when considering glycemic control for diabetes patients combined with CAD.
It still remained controversies about the impact of glycemic control on CV events. Several cardiovascular outcome trials, such as VADT [
20], ACCORD [
21], and ADVANCE [
22], failed to find a significant reduction of CV events risk when comparing more strict glycemic control with the standard care of diabetes. However, certain studies, such as the DCCT/EDIC [
23] and UKPDS [
24] study, have demonstrated that strict glycemic control might reduce the incidence of CV events. Identifying patients who are more likely to benefit from glycemic control management might help to resolve this problem.
Insulin secretion and resistance play important roles in glycemic control and might further influence the CV outcomes of diabetic patients [
25]. The higher degree of insulin resistance might induce insufficient insulin secretion after treatment with glycemic control agents, especially those targeting in improving insulin secretion, and result in a poor response in glucose control management, which might lead to persistent hyperinsulinemia and hyperglycemia [
26]. Hyperinsulinemia might continuously active the growth factor receptor-bound protein-2 signal pathway inactive the insulin receptor substrate pathway, and increase the level of plasminogen activator inhibitor-1, vascular cell adhesion molecule 1, and endothelin-1, which might induce the vasoconstriction, proliferation, migration of endothelium, promote atherosclerotic plaque formation and instability and increase the risk the adverse CV events [
27]. On the population level, however, few studies reported the impact of insulin resistance on the association between glycemic control and CV events. To our knowledge, only one small-sample study achieving glycemic control and improving insulin resistance might slightly but not significantly reduce the incidence of CV events for early type 2 diabetes patients [
28]. Suitable insulin resistance indices and larger population with higher CV risk might help clarify this question.
The TyG index has been extensively validated as a dependable indicator for evaluating insulin resistance, exhibiting notable sensitivity and specificity. Consequently, it has found widespread application in clinical settings due to its practicality, affordability, and versatile utility [
29]. The relationship between TyG index and CAD has been thoroughly examined, and previous cohort studies and meta-analyses have extensively demonstrated its predictive value for a high incidence of CAD [
30,
31]. For CAD patients, multiple cohort studies have indicated that individuals with a high TyG index are independently associated with an increased risk of repeated revascularization and in-hospital mortality [
32,
33]. For diabetic patients, one previous study has indicated TyG index was associated with the all-cause mortality risk in patients with diabetes or pre-diabetes [
34]. However, when it comes to the glycemic control for diabetic patients combined with established CAD, there was no research elucidating the role of the TyG with adverse CV events in this population. Thus, we focused on this point for the first time and revealed that uncontrolled glycemia was associated with an increase in CV risk only in those with high TyG patients.
This study has several limitations. Firstly, this is a single-center observational study. Therefore, it was not feasible to establish a definitive causal relationship between the TyG in conjunction with glycemia control status and the incidence of CV events. Secondly, dynamic changes in the TyG and glycemia control status during follow-up were not presented in our study. It was still unknown about the association between changes of TyG and glycemia control status and prognosis for diabetic CAD population. Thirdly, despite controlling for potential confounders as covariates in multivariable regression models, it is important to acknowledge that the impact of uncollected confounders cannot be completely disregarded. Fourthly, since there were no new hypoglycemic agents (such as glucagon-like peptide-1 receptor agonist, odium-dependent glucose transporters 2 inhibitors) data available, we could not estimate the effect of those new hypoglycemic agents on glycemic control management in this study. Fifthly, the post-hoc analysis of the NID-2 trial demonstrated that the number of main CV risk factors well controlled by drug therapy significantly influences the clinical outcome for diabetic patients with very high CV risk [
35]. Although we had tried our best to adjust CV risk factors (including hypertension, TC, HDL-C, LDL-C, aspirin use, statins use, and insulin use) in our study, follow-up data regarding medications (such as antihypertensive agents, lipid‑modulating agents, anticoagulants and anti-diabetic drugs) were not available in this study, which possibly had impacts on CV outcomes. Sixthly, our study only retrospectively analyzed the results of glycemic control and was unable to explore the specific process of glycemic management and the specific changes in HbA1c levels during glycemic management. Detailed glycemic control strategy and HbA1C target assisted by TyG index need to be further confirmed in larger prospective studies. Seventh, although we adjusted all available baseline data of insulin and lipid-lowering drugs, levels of TyG index might be affected by the above medication. Further studies might be needed to confirm the association of the TyG index with lipid-lowering agents and insulin therapy in CAD patients with diabetes. Finally, it was unclear whether glycemic control could improve clinical outcomes by improving insulin resistance. Future prospective and longitudinal studies are needed to explore the impact and mechanism of controlling blood glucose and improving insulin resistance on the clinical outcomes of diabetes patients with established CAD.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.