RESULTS

Patients

For our initial analysis, we enrolled 2300 patients admitted to our department. A total of 807 patients were excluded due to diabetes or prediabetes diagnosed prior to admission; 153 patients were excluded because of incomplete medical records. In addition, among those excluded, there were 18 cases of acute heart failure on admission. Finally, we analyzed data collected from 1340 patients at a median age of 67 years, among whom 70.4% were male. Most of the patients were overweight, with a median BMI of 26 kg/m². For 66% of them, MI was the first manifestation of CAD. Baseline characteristics of the study population are shown in Table 1.

Analysis of MACE

MACE occurred in 8.13 % (n = 109) of the study group. There were 35 cases of MI, 19 cases of in-stent restenosis, 49 cases of unstable angina, 4 cases of stroke or transient ischemic attack, and 13 hospitalizations due to heart failure. Furthermore, among these cases, there were 12 patients who developed 2 incidents of MACE at a 1-year follow-up, and there were 5 cases of unstable angina and in-stent restenosis, 5 cases of MI, and in-stent restenosis, and 2 of myocardial infarction and hospitalization due to heart failure.

Analysis of the groups of patients, with and without incidence of MACE at a 1-year follow-up, revealed that there were no statistically significant differences in median age, ejection fraction, BMI, Gensini score, glucose, high-density lipoprotein cholesterol (HDL-C), the TyG index value, the occurrence of hypertension, or sex, and lipid-lowering therapy prior to admission. The patients with incidence of MACE had lower low-density lipoprotein cholesterol (LDL-C), non-high-density lipoprotein cholesterol (non-HDL-C), total cholesterol (TC), triglycerides (TG), and estimated glomerular filtration rate (eGFR). Moreover, 46.8% of them had been diagnosed with CAD prior to admission, whereas in the second group this was only 32.6%. Detailed demographic and clinical characteristics of those groups are presented in Table 2.

Univariate and multivariable regression analysis of MACE

Univariate regression analysis showed that previously diagnosed CAD, eGFR, LDL-C, and TC were significant predictors of MACE. However, in the multivariable model, only previously diagnosed CAD proved to be an independent predictor (odds ratio [OR], 1.54; 95% CI, 1.02–2.32; P = 0.03). The TyG index was not an indicator of MACE in the study group (P = 0.06). The significant predictors of MACE are presented in Table 3.

Patients with potential glucose metabolism disorders

To deepen our analysis, we divided the patients according to their glycemic control status into 2 groups: one with a lower fasting glucose level (FGL <7.8 mmol/l) and the other with potential, previously undiagnosed, glucose metabolic disorder or stress hyperglycemia caused by MI (FGL ≥7.8 mmol/l). Hyperglycemia occurred in 25.2% of the patients (n = 338). There was a difference in medians of the TyG index between those 2 groups: 8.7 (8.4–9) in the lower FGL group, versus 9.17 (8.86–9.5) in the higher one (P <0.01). There was, however, no significant difference between glycemic control status during hospitalization and incidence of MACE at a 1-year follow-up. MACE occurred in 7.68% (n = 77) of patients with lower FGL and in 9.47% (n = 32) of those with potential glucose metabolic disorder (P = 0.29). Additionally, after excluding from the analysis the patients with higher FGL, only CAD diagnosed prior to admission, eGFR, LDL-C and TC were statistically significant predictors of MACE in univariate regression analysis (Table 4). The TyG index value was insignificant (P = 0.12).

Univariate and multivariable regression analysis of one-year mortality

The all-cause mortality rate at a 1-year follow-up was 14.5% (n = 195) for the whole study group, whereas in-hospital mortality was 1.6% (n = 22). In univariate regression analysis, the TyG index value appeared to be an irrelevant indicator of all-cause mortality (P = 0.25), whereas age, BMI, Gensini score, eGFR, LDL-C, and TC were statistically significant. Finally, multivariable regression analysis showed that only age was an independent predictor of all-cause mortality at a 1-year follow-up (OR, 1.1; 95% CI, 1.06–1.13; P <0.01). Predictors of all-cause mortality are shown in Table 5.

DISCUSSION

To the best of our knowledge, this study is the first that assesses the TyG index, measured during acute MI among non-diabetic patients, as a potential predictor of MACE and all-cause mortality at a 1-year follow-up. Previously, this metabolic marker was used as an easily accessible indicator of insulin resistance [9], a predictor of diabetes [10], and a biomarker of glycemic control in type 2 diabetes mellitus [11]. Since the TyG index is a quite novel IR marker, there is no internationally recognized cut-off value. Unger et al. [12] suggested that this value for metabolic syndrome in the general population was 8.8 in men and 8.7 in women, and in the study by Lee et al. [13], where the cut-off value for the TyG index was set at 8.8, this marker was a statistically significant predictor for incidental diabetes in 4-year follow-up. For the current study, the population’s median TyG index value was 8.8, which may suggest a high incidence of IR among patients with MI.

In many patients with MI, the level of fasting glucose is elevated and called “stress hyperglycemia”. This condition usually occurs in critically ill patients without diabetes mellitus diagnosed prior to admission [14, 15]. It appears to be connected with a stress mechanism, which is associated with steroid hormones, temporary IR, and a high level of free fatty acids [16]. According to the American Diabetes Association, stress hyperglycemia in hospitalized patients is related to a random glucose level greater than 7.8 mmol/l at any time [17]. In our research, this condition occurred in 25.2% of patients. There was no correlation between higher glucose level and incidence of MACE at 1-year follow-up. Furthermore, even after excluding from the analysis the patients with higher FGL, the TyG index, which is directly related to levels of TG and glucose, was not a predictor of MACE.

The usefulness of the TyG index as a predictor of cardiovascular events has previously been investigated in several studies, mostly among healthy individuals or patients with stable CAD.

A recent Chinese retrospective study [18] among 6076 healthy individuals aged over 60 years showed in a 6-year follow-up that a higher risk of CAD events was associated with an increasing value of the TyG index. Another study on that subject, conducted by Park et al. [19] and performed among healthy individuals with no traditional cardiovascular risk factors, showed that a TyG index value over 8.48 was a predictor of CAD. Finally, in an Iranian study [20], the risk of developing CAD increased with increasing quintiles of the TyG index in a long-term follow-up period (16 years).

To the best of our knowledge, little is known about the predictive value of the TyG index in patients with MI. Luo et al. [21] conducted a study on patients with STEMI, undergoing percutaneous coronary intervention, to assess the clinical outcomes of that marker during a follow-up period of 1 year. Those clinical outcomes were defined as major adverse cardiac and cerebrovascular events (MACCE) and included all-cause death, target vessel revascularization, MI, unstable angina pectoris, heart failure, stroke, and transient cerebral ischemia. In that study, patients were divided into 4 groups according to TyG-index quartiles. The incidence of MACCE and all-cause mortality was higher among patients with TyG index values in the highest quartile. Analysis of the predictors of MACCE showed statistical significance for a TyG index value ≥9.098, age, hypertension, diabetes, eGFR, number of implanted stents, and multivessel CAD in univariate analysis. In multivariable analysis, however, only a TyG index value ≥9.608 and the number of implanted stents were significant.

In our analysis, on the other hand, the TyG index value was not significant in univariate regression analysis (P = 0.06). In addition, in the multivariable model, only CAD diagnosed prior to admission was relevant (OR, 1.54; 95% CI, 1.02–2.32; P = 0.03).

In the study by Luo et al. [21], the percentage of patients with incidence of MACCE was higher than in our study — 34.3% vs. 8.13%. Moreover, patients in the MACCE group had higher mean values of FGL (9, standard deviation [SD] = 4.2 mmol/l vs. median value of 6.9 [5.9–8] mmol/l in our study), and 31.2% of them had diabetes. Furthermore, those patients had higher values of LDL-C, TC, and TG (mean value –1.9 [SD = 1.6] mmol/l vs. median value of 1.13 [0.9–1.44] mmol/l). Consequently, their TyG index value was higher, with a mean value of 10.076 (SD = 0.483) in the highest quartile group. Additionally, in that research, only 2.4% of patients with incidence of MACE had been diagnosed with CAD prior to admission, whereas in our study, this was 46.8%. Both study populations were similar concerning age, BMI, and the proportion of males. Finally, no correlation between 1-year mortality and the TyG index was found in our report, whereas in the study by Luo et al. [21] that correlation was statistically significant.

In another Chinese study presented by Mao et al. [22], patients with NSTEMI were initially divided into 2 groups according to their TyG index value, these being low (<8.8) and high (>8.8) scores. In that study, more than half of the patients had diabetes or glucose metabolism disorder. Additionally, the incidence of MACE. including cardiac death, nonfatal myocardial infarction, target vessel revascularization, congestive heart failure, and nonfatal stroke, was higher in the high TyG index group at a 1-year follow-up (12.8% vs. 22.8%; P <0.01).

To deepen the analysis, Mao [22] divided patients into 4 groups, depending on the TyG index value and the occurrence of glucose metabolism disorder. There was a statistically significant difference between the incidence of MACE among the patients without glucose metabolism disorder with low (10.7%) and high (33.8%) TyG index values. Finally, in univariate analysis, the TyG index was significantly associated with MACE (hazard ratio [HR], 1.951; 95% CI, 1.416–2.688; P <0.01). Furthermore, in the multivariable model, the TyG index also remained an independent predictor of MACE. In the Mao [22] research, the study group was not divided according to the incidence of MACE, so a simple comparison with our study is difficult to perform. In a relatively small population of 438 patients, the incidence of MACE was 17.8%, whereas in our population of 1340 patients, MACE occurred only in 8.13%.

On the other hand, a simple correlation of the value of the TyG Index with the incidence of atherosclerosis, its severity, and incidence of MACE, is questionable. Alizargar et al. [23], in their article assessing the practical value of the TyG index, emphasize that using this marker can be easily biased by hyperlipidemia, diabetes, or other glucose metabolic disorders, as the TyG index has a direct relationship with levels of TG and glucose (based on the TyG index formula). In conclusion, these factors should be carefully considered to justify the use of the TyG index as a biomarker. In the Polish population, we can still observe insufficient adherence to guidelines concerning the proper level of glucose, lipid profile, blood pressure, BMI, physical activity, and smoking [24]; therefore, the potential use of the TyG index might be limited.

Dziedzic et al. [25] in their study concerning educational programs among Polish elderly patients showed that several training meetings performed to change lifestyle in that group had an impact on the lipid profile of the participants, particularly concerning the level of TG (P = 0.02). Finally, Wybraniec et al. [26] revealed that patients enrolled on a similar program (Managed Care After Acute Myocardial Infarction Program) had a lower rate of MACE at a 1-year follow-up (11.3% vs. 19.1%; P = 0.0006). We believe that similar programs should be implemented in order to properly manage basic cardiological risk factors, reduce the rate of MACE, and improve patient survival.

Vega et al. [27] also presented concerns concerning the predictive value of the TyG index. In that study, this index was a positive predictor of coronary heart disease, cardiovascular disease, and all-cause mortality, but only unadjusted and, after adjustment, for age, smoking, BMI, and systolic blood pressure. After an additional adjustment for non-HDL-C level, the HR was lower — 0.83 for coronary heart disease, 0.89 for cardiovascular disease, and 0.89 for all-cause mortality. Our findings correspond with Vega’s conclusions that the TyG index does not predict all-cause mortality.

In our opinion, this metabolic biomarker should not be used as a predictor of clinical outcomes among non-diabetic patients with MI for several reasons. Firstly, the TyG index has a direct relationship with glycemia which can be labile in acute conditions such as MI. Secondly, in our study, the TyG index was not a predictor of MACE even after we excluded from the study the group of patients with FGL ≥7.8 mmol/l. Finally, there was no association between the TyG index value and all-cause mortality.

Study limitations

CONCLUSIONS

The TyG index does not appear to be a predictor of MACE among non-diabetic patients with MI. We believe that its potential use in acute conditions is limited by acute metabolic changes accompanying MI, and it does not help to identify non-diabetic individuals at a greater risk of poor clinical outcomes. Furthermore, no association between the TyG index value and all-cause mortality at a 1-year follow-up also reflects the questionable clinical value of that parameter. Moreover, comprehensive evaluation of cardiovascular risk factors should focus primarily on basic risk factors. Additional markers may be useful but after the effective management of these risk factors.

Article information

Acknowledgments: The study was supported by a grant from Jagiellonian University Medical College N41/DBS/000098; to EK).

Conflict of interest: None declared.

Open access: This article is available in open access under Creative Common Attribution-Non-Commercial-No Derivatives 4.0 International (CC BY-NC-ND 4.0) license, allowing to download articles and share them with others as long as they credit the authors and the publisher, but without permission to change them in any way or use them commercially. For commercial use, please contact the journal office at kardiologiapolska@ptkardio.pl.

How to cite: Drwiła D, Rostoff P, Gajos G, et al. Prognostic value of the triglyceride-glucose index among non-diabetic patients with acute myocardial infarction at one-year follow-up. Kardiol Pol. 2021; 79(10): 1116–1123, doi: 10.33963/KP.a2021.0104.

REFERENCES