Association between triglyceride glucose index and subclinical left ventricular systolic dysfunction in patients with type 2 diabetes

doi:10.21203/rs.3.rs-2310247/v1

Background

Triglyceride glucose (TyG) index has been considered a new biomarker for diagnosis of angiocardiopathy and insulin resistance. However, the association of TyG index with subclinical left ventricular (LV) systolic dysfunction still lacks a comprehensive exploration. The study was carried out to examine this relationship in the asymptomatic with type 2 diabetes mellitus (T2DM).

Methods

150 T2DM cases with preserved LV ejection fraction (LVEF ≥ 50%) from June 2021 to December 2021 were enrolled in this study. The subclinical LV function was evaluated through global longitudinal strain (GLS), with the pre-defined GLS < 18% as the cutoff for subclinical LV systolic dysfunction. The TyG index calculation was achieved according to ln (fasting triglycerides (mg/dL) × fasting glucose (mg/dL)/2), which was then stratified into 4 quartiles (TyG-Q).

Results

The analyses of baseline characteristics in the four TyG-Q (Q1 (TyG ≤ 8.89) n = 38, Q2 (8.89 < TyG ≤ 9.44) n = 37, Q3 (9.44 < TyG ≤ 9.83) n = 38, and Q4 (TyG > 9.83) n = 37) were conducted. A negative correlation of TyG index with GLS (r=-0.307, P < 0.001) was revealed according to correlation analysis. After the gender and age adjusted in multi-model logistic regression analysis, the higher TyG index (OR 6.86; 95% CI 2.44 to 19.30; P < 0.001, Q4 vs Q1) showed a significant association with GLS < 18%, which was still maintained after further adjustment for related clinical confounding factors (OR 5.15, 95%CI 1.13 to 23.39, p = 0.034, Q4 vs Q1). Receiver operator characteristic analysis indicated a diagnostic capacity of TyG index for GLS < 18% (area under curve: 0.678; P < 0.001).

Conclusions

Higher TyG index had a significant association with the subclinical LV systolic dysfunction in asymptomatic T2DM patients, with the potential to exert prognostic value for the progression of myocardial damage.

Type 2 diabetes mellitus

Triglyceride glucose index

Global longitudinal strain

Left ventricular systolic function

Insulin resistance

A causal relation is lying between diabetes mellitus and heart failure. For instance, diabetic cardiomyopathy as a diffuse cardiomyopathy resulting from glucose metabolism disorder has received increasing interest in recent years, which generally manifests as pathological cardiac remodeling, systolic and diastolic dysfunction, and may eventually develop to be overt heart failure[1, 2]. Studies have demonstrated the insulin resistance and/or hyperinsulinaemia as the source of the cascade that contributes to diabetic cardiomyopathy[3, 4]. The early stages of diabetic cardiomyopathy are usually ignored and underestimated in clinics. Some studies have shown the existence of asymptomatic systolic LV dysfunction in cases of type 2 diabetes mellitus (T2DM) with hidden cardiac disease manifestations[5], and more suggest the emergence of reduced global longitudinal strain (GLS) at an early stage of this disorder process prior to the detectability of ejection fraction (EF) transformations[6]. Therefore, the early identification followed by the punctual intervention is of vital significance for individuals at high risk of diabetic cardiomyopathy. TyG index is indicated to serve as an alternative biomarker of insulin resistance which is convenient and reliable[7–9], as is closely related with cardiovascular disease[10]. However, there is not yet enough evidence on evaluation of the clinical value of TyG index on subclinical LV dysfunction in diabetes. Herein, we explored the correlation of TyG index with LV longitudinal systolic function in T2DM patients without experiences of heart disease.

Study subjects

A total of 150 type 2 diabetic cases categorized by World Health Organization criteria at the Department of Endocrinology of Xijing Hospital of Air Force Medical University during the period of June 2021 to December 2021[11] were enrolled. According to the clinical records, subjects with (1) type 1 diabetes and other specific types of diabetes were excluded; (2) LVEF < 50%; (3) serious valvular disorders; (4) a coronary artery disease history or other heart disease; (5) arrhythmia; (6) too poor speck tracking image quality for analysis.

Data Collection

Demographic data were provided by the electronic medical record system, with diabetes duration, gender, age, systolic and diastolic pressures, and medication covered. All patients were measured for height and weight through specially-assigned personnel on the admission day, with the body mass index (BMI) defined as weight/height² (kg/m²). Hypertension was described as systolic blood pressure (SBP) ≥ 140 mmHg and/or diastolic blood pressure (DBP) ≥ 90 mmHg without antihypertensive medication, or with a previous hypertension physician identification. After an at least 8-hour overnight fast, an evaluation on the total cholesterol, triglyceride, high and low-density lipoprotein cholesterol, and uric acid were performed with an automatic biochemical analyzer. Fasting glucose index was detected with the glycosylated hemoglobin (HbA1c) checked by high-performance liquid chromatography. Immunoturbidimetry depending on a COBAS INTEGRA 400 plus autoanalyzer (Germany) was carried out to determine the urinary albumin-to-creatinine ratio (UACR). The ln[fasting triglycerides (mg/dL) × fasting glucose (mg/dL)/2][7] was used to identify the TyG index, and the patients were grouped into Q1 (TyG index ≤ 8.89), Q2 (8.89 < TyG index ≤ 9.44), Q3 (9.44 < TyG index ≤ 9.83), and Q4 (TyG index > 9.83) based on the TyG index levels.

Conventional Echocardiography

The ultrasound measurements of all subjects were collected referring to the guidelines of the American Society of Echocardiography[12]. Two-dimensional (2D) echocardiography (Philips Healthcare, iE33 system, X5-1 probe) was performed on each subject accompanied with electrocardiogram. LV fractional shortening (LVFS), LVEF, heart rate, and stroke volume of each subject were analyzed. Afterwards, the Pulse Doppler sampling volume was located under the mitral valve to perform the measurement on the early diastolic blood flow velocity (E peak) and late diastole (A peak) to calculate E/A ratio, then to measure the early diastolic mitral annulus motion velocity (E' peak) to calculate E/E' ratio by placed at the septal side of the mitral annulus.

Speck-tracking Echocardiography

Echocardiography images for 2D speck-tracking echocardiography (STE) with the rate of 60–90 frames/s in three cardiac cycles were acquired. Then the LV view of apical four-chamber, apical two-chamber and apical three-chamber were analyzed on a software of QLAB 8.1 2D imaging system. The trace of endocardial border at end-diastole was performed in a manual manner, to obtain the 2D strain-time curve and bull's-eye plot of 17 segments of LV (Fig. 1). The trace was adjusted based on a visual assessment of the tracking quality by observer and the untraceable images of spots elicited by atrial fibrillation were excluded from the subsequent analysis. The LV GLS for the average value of the three peak strains in systole was calculated for subclinical LV systolic function evaluation. Based on the previous studies and the latest guidelines of the European Association of Cardiovascular Imaging, the pre-defined cutoff of GLS < 18% was adopted to evaluate subclinical LV systolic dysfunction [12–14].

Statistical analysis

Kolmogorov-Smirnov Test was performed for the normal distribution and nonparametric test for the skewed data. Categorical variables were described in the form of percentages n (%) and continuous variables of means ± SD, or median and interquartile range. Inter-group comparisons were carried out using One-way ANOVA or Mann-Whitney U tests. Least Significant Difference (LSD) was adopted for Post hoc comparisons. Pearson's Chi-square test was performed on categorical variable comparisons. The correlation between TyG index and GLS was evaluated according to Pearson correlation coefficients. Three forced-entry logistic regression models were performed to determine the independent association of GLS < 18% with TyG index. The potential confounders referring to age, gender, diabetes duration, systolic pressure, HbA1c, BMI, hypertension, heart rate, microalbuminuria and LVEF were all involved into the multivariate regression models. To confirm an alternative index for identifying the reduced GLS < 18%, the models covering TyG index and HbA1c were established, followed by the comparison on both according to the area under the receiver operating characteristic (ROC) curves (AUC). The measurement of GLS was performed by one professional physician to avoid the inter-observer and intra-observer variability. To prevent the ambiguity of negative size to a value, the GLS was given in the absolute value form. All statistical analysis was carried out on IBM SPSS statistics 26.0, with P -value < 0.05 considered as statistically significant.

Baseline characteristics

A total of 150 cases (mean age: 53.43 ± 13.81 years, diabetes duration: 10.25 ± 7.22 years, male sex: 64.0%) were enrolled, with the clinical features across quartiles of the TyG index listed in Table 1, which indicated an increased hypertension prevalence on subjects with the higher TyG index, accompanied with the higher levels of systolic pressure, fasting glucose, HbA1c, heart rate, triglyceride, lipoprotein cholesterol with low-density, and total cholesterol (all P < 0.05). Moreover, the insulin treatment statistically varied across the TyG index quartiles (p = 0.025), while without obvious difference in gender, age, diabetes duration and BMI among groups (P > 0.05).

Table 1

Baseline characteristics of participants by quartiles of TyG index
Characteristic	Q1 (TyG index ≤ 8.89) n = 38	Q2 (8.89 < TyG index ≤ 9.44) n = 37	Q3 (9.44 < TyG index ≤ 9.83) n = 38	Q4(TyG index > 9.83) n = 37	P value
Male, n (%)	23(60.5)	26 (70.3)	26 (68.4)	21(56.8)	0.573
Age, years	52.46 ± 15.31	55.84 ± 13.30	52.61 ± 11.18	53.05 ± 15.28	0.692
BMI, kg/m²	23.29 ± 3.88	23.87 ± 3.54	24.82 ± 3.10	24.51 ± 4.11	0.273
Diabetes duration, years	10.79 ± 7.08	8.57 ± 6.97	9.17 ± 6.78	12.29 ± 7.79^e	0.115
Heart rate, bpm	72.45 ± 8.73	73.65 ± 12.12	77.74 ± 2.71	81.31 ± 4.02^ce	0.008
SBP, mmHg	127.08 ± 15.29	135.08 ± 14.53	131.26 ± 14.62	139.59 ± 24.44^cf	0.019
DBP, mmHg	76.16 ± 9.30	78.97 ± 9.53	77.39 ± 9.20	80.43 ± 14.86	0.359
HbA1c, (%)	7.93 ± 1.64	8.49 ± 1.84	9.05 ± 1.96^b	9.75 ± 2.46^ce	0.001
FPG, mmol/l	8.06 ± 1.92	11.10 ± 4.22^a	12.51 ± 3.92^b	15.25 ± 6.04^cef	< 0.001
Total cholesterol, mmol/l	3.57 ± 1.05	3.72 ± 1.10	4.07 ± 0.90	4.83 ± 1.57 ^cef	< 0.001
HDL-C, mmol/l	1.19 ± 0.32	1.10 ± 0.53	1.09 ± 0.47	1.03 ± 0.36	0.426
LDL-C, mmol/l	1.95 ± 0.91	2.17 ± 0.97	2.61 ± 1.27^b	2.76 ± 1.22^ce	0.006
Triglyceride, mmol/l	0.79 ± 0.28	1.19 ± 0.42	1.66 ± 0.56^bd	3.18 ± 1.81^cef	< 0.001
Apolipoproteins A1, g/l	1.27 ± 0.22	1.14 ± 0.18^a	1.16 ± 0.18^b	1.23 ± 0.21	0.028
Apolipoproteins B, g/l	0.56 ± 0.17	0.63 ± 0.21	0.81 ± 0.49^bd	0.82 ± 0.31^ce	0.001
Albuminuria, mg/l	11.30(8.13–15.18)	12.85(8.83–20.08)	9.20(8.00–15.00)	14.90(8.30–105.00)	0.109
UACR, mg/mmol	1.45(0.77–2.42)	1.45 (1.03–3.55)	1.70 (0.75–2.73)	2.41(1.12–16.47)	0.172
Uric acid, umol/l	301.43 ± 66.61	315.00 ± 80.30	327.30 ± 87.99	341.92 ± 88.45^c	0.177
hypertension, n (%)	8 (21.1)	20 (54.1)^a	19 (50.0)^b	21(56.8)^c	0.006
Diabetic nephropathy, n (%)	5 (13.2)	8(21.6)	8(21.1)	14 (37.8)	0.081
Diabetic retinopathy, n (%)	3 (7.9)	3 (8.1)	5 (13.2)	10 (27.0)	0.058
Medical treatment
ACEI/ARB, n (%)	4 (10.5)	8 (21.6)	11 (28.9)	11 (29.7)	0.163
CCB, n (%)	5 (13.2)	10(27.0)	8 (21.1)	8 (21.6)	0.523
Statin, n (%)	8 (21.1)	7(18.9)	10 (26.3)	10 (27.0)	0.805
Insulin, n (%)	23 (60.5)	12 (32.4)^a	24 (63.2)	22 (59.5)	0.025
SGLT-2I, n (%)	4(10.5)	3(8.1)	4 (10.5)	3 (8.1)	0.968
GLP-1RA, n (%)	5 (13.2)	2 (5.4)	4(10.5)	4 (10.8)	0.723
Metformin, n (%)	27(71.1)	26 (70.3)	27 (71.1)	25(67.6)	0.986
αGI, n (%)	15(39.5)	13 (35.1)	12 (31.6)	13(35.1)	0.914
Note: Values are presented as mean ± SD, median (interquartile range) or n (%). Bold indicated value of P < 0.05. Abbreviations: TyG, triglyceride glucose; BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; FPG, fasting plasma glucos; TC, total cholesterol; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; UACR, urinary albumin-to-creatinine ratio; CCB, calcium channel blocker; ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin II receptor blocker; SGLT-2I, sodium glucose cotransporter 2 inhibitor; GLP-1RA, glucagon like peptide-1receptor agonist; α-GI, α-glucosidase inhibitor.
a P < 0.05 between 1st quartile and 2nd quartile.
b P < 0.05 between 1st quartile and 3rd quartile.
c P < 0.05 between 1st quartile and 4th quartile.
d P < 0.05 between 2nd quartile and 3rd quartile.
e P < 0.05 between 2nd quartile and 4th quartile.
f P < 0.05 between 3rd quartile and 4th quartile.

Association Of Tyg Index With Gls

Table 2 displayed the characteristics of the LV function stratified by quartiles of TyG index. No statistically significant differences were observed in traditional echocardiographic parameters, that were, LVEF, LVFS, stroke volume, E, E', E/A and E/E' across TyG index quartiles (all P > 0.05). However, the GLS exhibited a stepwise decrease in line with the increase of TyG index quartile (19.2 ± 3.3 vs 18.9 ± 2.8 vs 17.2 ± 3.6 vs 16.7 ± 3.3, P-trend = 0.001). Pearson correlation analysis indicated the strong and negative correlation of TyG index (r=-0.307, P < 0.001) and HbA1c (r=-0.470, P < 0.001) with GLS.

Table 2

Characteristics of the LV function stratified by quartiles of TyG index
Echocardiographic indexes	Q1 (TyG index ≤ 8.89) n = 38	Q2 (8.89 < TyG index ≤ 9.44) n = 37	Q3 (9.44 < TyG index ≤ 9.83) n = 38	Q4(TyG index > 9.83) n = 37	P value
LV EF, %	60.14 ± 3.89	60.19 ± 5.55	60.13 ± 5.01	59.91 ± 4.30	0.995
LV FS, %	31.94 ± 3.36	32.09 ± 4.34	32.41 ± 3.92	31.73 ± 3.86	0.912
Stroke volume, ml	47.22 ± 8.59	48.47 ± 9.64	48.22 ± 9.91	45.61 ± 7.50	0.561
E	67.12 ± 19.01	67.90 ± 14.12	65.97 ± 13.74	64.55 ± 15.03	0.850
A	73.48 ± 19.17	80.40 ± 21.18	75.07 ± 19.81	83.26 ± 19.40	0.181
E'	7.92 ± 3.01	7.01 ± 2.41	7.74 ± 2.75	6.77 ± 2.44	0.303
E/A ratio	1.00 ± 0.38	1.36 ± 2.82	0.94 ± 0.36	0.83 ± 0.39	0.508
E/E' ratio	10.36 ± 4.23	10.22 ± 3.49	9.52 ± 3.01	10.14 ± 3.11	0.825
GLS, %	19.23 ± 3.28	18.87 ± 2.79	17.22 ± 3.56^bd	16.67 ± 3.31^ce	0.001
Note: Bold indicated value of P < 0.05. Abbreviations: TyG, triglyceride glucose; LVEF, left ventricular ejection fraction; LVFS, LV fractional shortening; E, peak early diastolic mitral flow velocity; A, peak late diastolic mitral flow velocity; E/A, peak early diastolic (E-wave) and late diastolic (A-wave) velocities ratio; E/E', mitral inflow E and mitral E' annular velocities ratio; GLS, global longitudinal strain.
a P < 0.05 between 1st quartile and 2nd quartile.
b P < 0.05 between 1st quartile and 3rd quartile.
c P < 0.05 between 1st quartile and 4th quartile.
d P < 0.05 between 2nd quartile and 3rd quartile.
e P < 0.05 between 2nd quartile and 4th quartile.
f P < 0.05 between 3rd quartile and 4th quartile

Subsequently, Table 3 displayed the logistic regression results based on the TyG index quartile. Taking Q1 as the reference, the risks in the Q3 and Q4 group of GLS < 18% was found significantly higher in comparison to the Q1 group in the univariate model (Q3 vs Q1: OR 4.29, 95% CI 1.63 to 11.35, P = 0.003; Q4 vs Q1: OR 5.83, 95% CI 2.15 to 15.82, P < 0.001, respectively). After the gender and age adjusted, the relation of TyG index of the 3rd quartile and 4th quartile with GLS < 18% still existed (Q3 vs Q1: OR 4.87, 95%CI 1.78 to 13.28, p = 0.002; Q4 vs Q1: OR 6.86, 95%CI 2.44 to 19.30, p < 0.001, respectively). With the further adjustment for confounders of age, gender, diabetes duration, systolic pressure, HbA1c, BMI, hypertension, heart rate, logarithmic microalbuminuria and LVEF, the higher quartile of TyG index remained an independent risk indicator related to GLS < 18% (Q4 vs Q1: OR 5.15, 95%CI 1.13 to 23.39, p = 0.034).

Table 3

Locgistic regression analysis of GLS < 18% by TyG index quartiles
	Model 1		Model 2		Model 3
	OR (95%CI)	P value	OR (95%CI)	P value	OR (95%CI)	P value
Q1 (TyG index ≤ 8.89)	Reference	—	Reference	—	Reference	—
Q2 (8.89 < TyG index ≤ 9.44)	1.34 (0.50, 3.64)	0.561	1.59 (0.57, 4.48)	0.380	1.26(0.30, 5.23)	0.753
Q3 (9.44 < TyG index ≤ 9.83)	4.29 (1.63, 11.35)	0.003	4.87 (1.78, 13.28)	0.002	4.55(1.13, 18.30)	0.033
Q4 (TyG index > 9.83)	5.83 (2.15, 15.82)	< 0.001	6.86 (2.44, 19.30)	< 0.001	5.15(1.13, 23.39)	0.034
Note: Bold indicated value of P < 0.05. Abbreviations: TyG, triglyceride glucose; OR, odds ratio; CI, confidential interval. Mode 1: unadjusted; Mode 2: adjusted age and sex; Mode 3: further adjusting for confounders of age, gender, diabetes duration, systolic pressure, HbA1c, BMI, hypertension, heart rate, log transformed microalbuminuria and LVEF

The ROC curves depicted in Fig. 2 demonstrated the diagnostic validity of TyG index in identifying subclinical LV systolic dysfunction (GLS < 18%). Notably, the TyG index with the cut-off value of 9.6 (AUC: 0.678; P < 0.001) displayed a sensitivity of 73.8% with a specificity of 54.3% for predicting GLS < 18%. The HbA1c also exhibited high AUC of 0.742, a sensitivity of 62.9% with a specificity of 76.2% for reduced GLS < 18% (P < 0.001). Subsequently, the composite variable with TyG index and HbA1c combined showed the increased AUC and diagnostic values (AUC: 0.770; sensitivity: 65.7%, specificity: 80.0%, P < 0.001).

To our best knowledge, the present study is the first to explore the association of TyG index to LV longitudinal myocardial function in asymptomatic T2DM patients with preserved LVEF. The results demonstrated the close relation of the increased TyG index with the elevated risk of LV longitudinal myocardial dysfunction. Moreover, it was found that the composite parameters of TyG index and HbA1c exhibited a certain value for the identification of the reduced GLS < 18%.

The correlation between heart failure and diabetes has been notably confirmed by epidemiological and clinical studies[15–17]. However, among diabetes-related complications, diabetic cardiomyopathy, as an "entity", remains poorly understood. Increasing studies have indicated a hidden subclinical period existing in diabetic cardiomyopathy, featured by the subtle abnormalities in function and structure[18]. In this context, the asymptomatic LV dysfunction, defined as the abnormal diastolic or systolic function without clinically detectable heart disease, is frequently reported in T2DM patients, which is expected to be between 50% and 70% [19] and presented as LV systolic dysfunction in one-third of patients[20]. Recently, GLS has been adopted as a preferred indicator to evaluate global LV systolic function considering the longitudinal sub-endocardium fibers as the most vulnerable that is first to be damaged by metabolic disorders in the early stage of diabetic heart disease[21, 22]. Ernande et al. also proposed the presence of LV longitudinal dysfunction in T2DM cases with preserved LVEF but normal LV diastolic function, which was defined as GLS < 18%[23]. The specific mechanism of this disorder is remaining to be uncovered. Metabolic characterizations indicated the impaired insulin metabolic signaling as a contributing pathophysiological abnormality associated with diabetic cardiomyopathy[3, 4].

Up to now, the standard estimation of insulin resistance, hyperinsulinemic-euglycemic clamp (HEGC) test, still requires diagnostic technology, which is highly costing and is not available for basic-level hospitals utilization. TyG index as an ideal surrogate of insulin resistance regardless of insulin treatment status has been widely validated to be robustly related to cardiovascular events[24–26]. However, evidence on the validity of TyG index on LV longitudinal myocardial function in those without prominent symptoms of heart failure is still not sufficient. Indeed, individuals with insulin resistance tend to develop systematic metabolic disorders, including dyslipidaemia, hyperglycaemia, and hypertension that were also reported by the present study, as the highest quartile of TyG index tended to be associated with a high prevalence of hypertension, poor blood glucose control and lipid level. These interactions will significantly promote the insulin resistance. Notably, these patients were more prone to suffering from the reduced GLS compared to those in the lowest quartile. Subsequently, the multi-model logistics regression analysis demonstrated the independent association of higher TyG index (ORs: 4.55 and 5.15 in the Q3 and Q4 groups compared with Q1 group) with subclinical LV systolic dysfunction assessed by GLS < 18%. This result was consistent with the view of Ikonomidis et al, who reported that insulin resistance was related to GLS and resulted in LV longitudinal dysfunction in immediate family of T2DM patients[27]. In fact, the reduced coronary flow reserve has been evidenced as a crucial determinants of LV longitudinal subendocardial myocardial fiber deformation. In addition, the insulin resistance may induce myocardial injury through other various mechanisms, including oxidative stress, fibrosis, autonomic nervous dysfunction and inefficient energy metabolism[28, 29]. As a result, TyG index may serve as the contributing reference for detecting cardiac involvement at a relatively earlier stage of diabetic heart disease.

In a study with large cohort followed for 10 years, Sánchez-Íñigo et al. first proposed a positive correlation of TyG index (AUC: 0.708) with heart events[30]. Similarly, the plotted ROC here indicated the certain clinical validity of TyG index (AUC: 0.678) for reduced GLS. More interestingly, compared to the HbA1c and the composite index, the TyG index with cut-off value of 9.6 showed the highest sensitivity but the lowest specificity in predicting subclinical LV systolic dysfunction. This reached an agreement with the previous studies where insulin resistance has been recognized as both a pathogenic trigger and a predictor of cardiovascular events[10]. Moreover, the AUC of TyG index binding to HbA1c (AUC: 0.770) observed in this study provided an incremental prognostic value for poor cardiac outcomes. Despite the absence of absolute illustration on the underlying mechanisms of this relationship, it is determined that TyG index represents the combined effect of "glycotoxicity" and "lipid toxicity", which prominently contribute to the reduced endocardial collateral flow density and the impaired coronary microcirculation in patients with T2DM[4, 31]. Consequently, it is not unexpected to observe the systemic lipid disturbances covering the elevated total cholesterol, triglycerides, low-density lipoprotein cholesterol levels, and apolipoproteins in the present study that in turn evoke the oxidative stress and inflammation, with the potential to elicit lipotoxic cardiomyopathy[32]. These pathologies further support the triggering role of insulin resistance for early initiation of the myocardial function changes in diabetic patients, such as hyperglycemia and dysfunction of lipid oxidation and utilization[33].

Study Strengths And Limitations

Previous studies are mostly focused on patients with existing symptoms of heart failure. In contrast, more emphasis was enriched in early attention to the LV subclinical phase of patients in the present study. The relatively time-consuming requirement, high cost, and professionals feature of speckle tracking echocardiography, and the accumulated training to perform effective measurement and analysis, may limit its application in daily clinical practice for general diabetes physicians without enough experience on this technique. As stated above, the hyperglycemia elicited by insulin resistance switch on the cascade of diabetic heart dysfunction, which induces metabolic disorders, followed by the endothelial dysfunction, cardiac hypertrophy and fibrosiss[34, 35]. HEGC has been confirmed closely related to the poor prognosis in type 2 diabetes, exerting a praisable validity for the prevention and treatment of those group of patients. However, the complex operation and uneconomical test limit its wide availability in clinical practice. The homeostasis model assessment of insulin resistance (HOMA-IR) is considered another preferential index, which requires the measurement of fasting insulin. Since the absence of standardized method for insulin measurement or the recognized cut-off value, it is relatively difficult to be applied in primary hospitals. Promisingly, the TyG index has been validated as a more accurate novel measurement of insulin resistance compared to HOMA⁃IR[36]. More prominently, this index is available and inexpensive in reality. The outcomes of this study demonstrated the validity of TyG index to assist clinicians to screen out people at high risk of cardiovascular events, exerting a more prominent role in the prevention and intervention of diabetic cardiomyopathy. Thus, it is recommended to perform a punctual monitoring of TyG index as soon as possible. For people in T2DM population with high TyG index in particular, metabolic disorders are suggested controlled earlier, and the advanced hypoglycemic medicines for cardiovascular protection should be administrated to effectively avoid and delay the occurrence and development of diabetic heart disease and ultimately bring clinical benefits to patients.

The limitations of this study should also be considered. Firstly, resulting from the cross-sectional study, the specific causal relationship of TyG index with the reduced GLS remains unclear. Secondly, not all patients without coronary artery disease have undergone the invasive coronary angiography. Thirdly, despite the analysis on the medication of the patients, the underlying contributions of medications for lipid-lowering and LV function improvements were failed to be controlled in this study. Fourthly, the insulin resistance parameters were not analyzed in this study. Lastly, covariates involved in the multivariable regression models were taken as the potential confounders based on previous studies[37, 38] or their biological plausibility, which partly limits the further application of the research findings.

In conclusion, the present study demonstrated a close relation of an elevated TyG index to the decreased GLS, which could be adopted as a sensitive and practical index to predict the subclinical LV systolic dysfunction. Therefore, TyG index should be punctually monitored for asymptomatic T2DM patients with normal LVEF, which is expected to effectively identify the occurrence of diabetic heart disease and delay its development.

TyG: Triglyceride glucose; LVEF: Left ventricular ejection fraction; T2DM: Type 2 diabetes mellitus; GLS: Global longitudinal strain; BMI: Body mass index; SBP: Systolic blood pressure; DBP: Diastolic blood pressure; HbA1c: Glycosylated hemoglobin; UACR: Urinary albumin-to-creatinine ratio; LVFS: Left ventricular fractional shortening; E peak: Early diastolic blood flow velocity; A peak: Late diastolic blood flow velocity; E' peak: Early diastolic mitral annulus motion velocity; 2D STE: Two-dimensional speck-tracking echocardiography; AUC: Area under the curve; ROC: Receiver operating characteristic. HEGC: Hyperinsulinemic-euglycemic clamp; HOMA-IR: Homeostasis model assessment of insulin resistance

Ethics approval and consent to participate

The present study protocol was reviewed and approved by the Institutional Review Board of Xijing Hospital of Air Force Medical University (No. XJLL -KY20222107).

Availability of data and materials

The datasets analysed during the current study are available from the corresponding author on reasonable request.

Competing interests

The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.

Funding

This study was funded by the National Natural Science Foundation of China (82070839 to JZ) and Natural Science Basic Research Program of Shaanxi (2020JZ-31 to JZ)

Authors’ contributions

Yanyan Chen analyzed and wrote the manuscript. Jianfang Fu, Yi Wang, Ying Zhang and Min Shi performed the research. Cheng Wang, Mengying Li, Li Wang and Xiangyang Liu were responsible for data collection. Shengjun Ta, Liwen Liu and Zeping Li were responsible for statistical analysis. Xiaomiao Li was responsible for the overall guidance and revisions of the article. Xiaomiao Li and Jie Zhou reviewed the manuscript and were responsible for the quality control and administration of the article as a whole. All authors gave their final approval of the final version of the manuscript. Xiaomiao Li and Jie Zhou contributed equally to this work as cocorresponding authors.

Acknowledgements

Not applicable

Tan Y, Zhang Z, Zheng C, Wintergerst KA, Keller BB, Cai L. Mechanisms of diabetic cardiomyopathy and potential therapeutic strategies: preclinical and clinical evidence. Nat Rev Cardiol. 2020;17:585–607.
Peterson LR, Gropler RJ. Metabolic and Molecular Imaging of the Diabetic Cardiomyopathy. Circ Res. 2020;126:1628–45.
Jia G, Whaley-Connell A, Sowers JR. Diabetic cardiomyopathy: a hyperglycaemia- and insulin-resistance-induced heart disease. Diabetologia. 2018;61:21–8.
Jia G, Demarco VG, Sowers JR. Insulin resistance and hyperinsulinaemia in diabetic cardiomyopathy. Nat Rev Endocrinol. 2016;12:144–53.
Faden G, Faganello G, De Feo S, Berlinghieri N, Tarantini L, Di Lenarda A, et al. The increasing detection of asymptomatic left ventricular dysfunction in patients with type 2 diabetes mellitus without overt cardiac disease: data from the SHORTWAVE study. Diabetes Res Clin Pract. 2013;101:309–16.
Minciună IA, Hilda Orășan O, Minciună I, Lazar AL, Sitar-Tăut AV, Oltean M, et al. Assessment of subclinical diabetic cardiomyopathy by speckle-tracking imaging. Eur J Clin Invest. 2021;51:e13475.
Guerrero-Romero F, Simental-Mendía LE, González-Ortiz M, Martínez-Abundis E, Ramos-Zavala MG, Hernández-González SO, 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.
Dikaiakou E, Vlachopapadopoulou EA, Paschou SA, Athanasouli F, Panagiotopoulos Ι, Kafetzi M, et al. Τriglycerides-glucose (TyG) index is a sensitive marker of insulin resistance in Greek children and adolescents. Endocrine. 2020;70:58–64.
Sánchez-García A, Rodríguez-Gutiérrez R, Mancillas-Adame L, González-Nava V, Díaz González-Colmenero A, Solis RC, et al. Diagnostic Accuracy of the Triglyceride and Glucose Index for Insulin Resistance: A Systematic Review. Int J Endocrinol. 2020; 2020: 4678526.
Tao LC, Xu JN, Wang TT, Hua F, Li JJ. Triglyceride-glucose index as a marker in cardiovascular diseases: landscape and limitations. Cardiovasc Diabetol. 2022;21:68.
Alberti KG, Zimmet PZ. Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus provisional report of a WHO consultation. Diabet Med. 1998;15:539–53.
Lang RM, Badano LP, Mor-Avi V, Afilalo J, Armstrong A, Ernande L, et al. Recommendations for cardiac chamber quantification by echocardiography in adults: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiogr. 2015;28:1–39.e14.
Ernande L, Bergerot C, Girerd N, Thibault H, Davidsen ES, Gautier Pignon-Blanc P, et al. Longitudinal myocardial strain alteration is associated with left ventricular remodeling in asymptomatic patients with type 2 diabetes mellitus. J Am Soc Echocardiogr. 2014;27:479–88.
Marwick TH, Leano RL, Brown J, Sun JP, Hoffmann R, Lysyansky P, et al. Myocardial strain measurement with 2-dimensional speckle-tracking echocardiography: definition of normal range. JACC Cardiovasc Imaging. 2009;2:80–4.
Kenny HC, Abel ED. Heart Failure in Type 2 Diabetes Mellitus. Circ Res. 2019;124:121–41.
Mchugh K, Devore AD, Wu J, Matsouaka RA, Fonarow GC, Heidenreich PA, et al. Heart Failure With Preserved Ejection Fraction and Diabetes: JACC State-of-the-Art Review. J Am Coll Cardiol. 2019;73:602–11.
Jankauskas SS, Kansakar U, Varzideh F, Wilson S, Mone P, Lombardi A, et al. Heart failure in diabetes. Metabolism. 2021;125:154910.
Falcão-Pires I, Leite-Moreira AF. Diabetic cardiomyopathy: understanding the molecular and cellular basis to progress in diagnosis and treatment. Heart Fail Rev. 2012;17:325–44.
Natali A, Nesti L, Fabiani I, Calogero E, Di Bello V. Impact of empagliflozin on subclinical left ventricular dysfunctions and on the mechanisms involved in myocardial disease progression in type 2 diabetes: rationale and design of the EMPA-HEART trial. Cardiovasc Diabetol. 2017;16:130.
Cioffi G, Giorda CB, Chinali M, Di Lenarda A, Faggiano P, Lucci D, et al. Analysis of midwall shortening reveals high prevalence of left ventricular myocardial dysfunction in patients with diabetes mellitus: the DYDA study. Eur J Prev Cardiol. 2012;19:935–43.
Ikonomidis I, Tzortzis S, Triantafyllidi H, Parissis J, Papadopoulos C, Venetsanou K, et al. Association of impaired left ventricular twisting-untwisting with vascular dysfunction, neurohumoral activation and impaired exercise capacity in hypertensive heart disease. Eur J Heart Fail. 2015;17:1240–51.
Dillmann WH. Diabetic Cardiomyopathy. Circ Res. 2019;124:1160–62.
Ernande L, Bergerot C, Rietzschel ER, De Buyzere ML, Thibault H, Pignonblanc PG, et al. Diastolic dysfunction in patients with type 2 diabetes mellitus: is it really the first marker of diabetic cardiomyopathy? J Am Soc Echocardiogr. 2011;24:1268-75.e1.
Garofolo M, Gualdani E, Scarale MG, Bianchi C, Aragona M, Campi F, et al. Insulin Resistance and Risk of Major Vascular Events and All-Cause Mortality in Type 1 Diabetes: A 10-Year Follow-up Study. Diabetes Care. 2020;43:e139-e41.
Laakso M, Kuusisto J. Insulin resistance and hyperglycaemia in cardiovascular disease development. Nat Rev Endocrinol. 2014;10:293–302.
Lebovitz HE. Insulin resistance–a common link between type 2 diabetes and cardiovascular disease. Diabetes Obes Metab. 2006;8:237–49.
Ikonomidis I, Pavlidis G, Lambadiari V, Kousathana F, Varoudi M, Spanoudi F, et al. Early detection of left ventricular dysfunction in first-degree relatives of diabetic patients by myocardial deformation imaging: The role of endothelial glycocalyx damage. Int J Cardiol. 2017;233:105–12.
Zhao CT, Wang M, Siu CW, Hou YL, Wang T, Tse HF, et al. Myocardial dysfunction in patients with type 2 diabetes mellitus: role of endothelial progenitor cells and oxidative stress. Cardiovasc Diabetol. 2012;11:147.
Witteles RM, Fowler MB. Insulin-resistant cardiomyopathy clinical evidence, mechanisms, and treatment options. J Am Coll Cardiol. 2008;51:93–102.
Sánchez-Íñigo L, Navarro-González D, Fernández-Montero A, Pastrana-Delgado J, Martínez JA. The TyG index may predict the development of cardiovascular events. Eur J Clin Invest. 2016;46:189–97.
Adameova A, Dhalla NS. Role of microangiopathy in diabetic cardiomyopathy. Heart Fail Rev. 2014;19:25–33.
Nakamura M, Sadoshima J. Cardiomyopathy in obesity, insulin resistance and diabetes. J Physiol. 2020;598:2977–93.
Pan Y, Zhong S, Zhou K, Tian Z, Chen F, Liu Z, et al. Association between Diabetes Complications and the Triglyceride-Glucose Index in Hospitalized Patients with Type 2 Diabetes. J Diabetes Res. 2021; 2021: 8757996.
El Hayek MS, Ernande L, Benitah JP, Gomez AM, Pereira L. The role of hyperglycaemia in the development of diabetic cardiomyopathy. Arch Cardiovasc Dis. 2021;114:748–60.
Avagimyan A. The Pathophysiological Basis of Diabetic Cardiomyopathy Development. Curr Probl Cardiol. 2022: 101156.
Son DH, Lee HS, Lee YJ, Lee JH, Han JH. Comparison of triglyceride-glucose index and HOMA-IR for predicting prevalence and incidence of metabolic syndrome. Nutr Metab Cardiovasc Dis. 2022;32:596–604.
Mochizuki Y, Tanaka H, Matsumoto K, Sano H, Toki H, Shimoura H, et al. Clinical features of subclinical left ventricular systolic dysfunction in patients with diabetes mellitus. Cardiovasc Diabetol. 2015;14:37.
Yamauchi Y, Tanaka H, Yokota S, Mochizuki Y, Yoshigai Y, Shiraki H, et al. Effect of heart rate on left ventricular longitudinal myocardial function in type 2 diabetes mellitus. Cardiovasc Diabetol. 2021;20:87.

No competing interests reported.

Proofreport.pdf

Association between triglyceride glucose index and subclinical left ventricular systolic dysfunction in patients with type 2 diabetes

Status:

Journal Publication

Version 1

Abstract

Background

Methods

Results

Conclusions

Figures

Background

Methods

Study subjects

Data Collection

Conventional Echocardiography

Speck-tracking Echocardiography

Statistical analysis

Results

Baseline characteristics

Association Of Tyg Index With Gls

Discussion

Study Strengths And Limitations

Conclusions

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1