The effect of type 2 diabetes mellitus on multiple obstructive coronary artery disease

Although type 2 diabetes mellitus (T2DM) individuals easily develop three‐vessel disease (3VD) coronary artery disease (CAD), there is very little information available about their left ventricle (LV) functions. The purpose of this study is to evaluate the LV function using two‐dimensional speckle tracking echocardiography (2‐D STE) in T2DM patients with 3VD.


INTRODUCTION
The prevalence of type 2 diabetes mellitus (T2DM) is increasing rapidly worldwide, as a result, the number of persons with the disease has doubled over the past three decades. 1 T2DM can deteriorate the structure and function of the left ventricular (LV). 2 Myocardial fibrosis, microangiopathy, altered myocardial metabolism, and elevated oxidative stress all contribute to T2DM's myocardial dysfunction. [3][4][5] One of the most common causes of morbidity and mortality in people with T2DM is coronary artery disease (CAD). 6 Moreover, T2DM is known as a significant risk factor leading to multiple vessels CAD, especially three-vessel disease (3VD). 7 According to reports, individuals with non-ST segment elevation myocardial infarctions had a 30.2% overall frequency of 3VD, while patients with DM comorbidity had a substantially higher frequency. 8 DM patients are more likely to develop 3VD. However, there is a dearth of information on its cardiac function.
Two-dimensional speckle tracking echocardiography (2-D STE), commonly referred to as deformation imaging, is a recent innovation that has been introduced as a quantitative method to impartially evaluate regional cardiac function. 9 Myocardial strain has superior diagnostic power than conventional approaches for identifying clinically undetectable LV dysfunction brought on by DM or CAD. 10 Previous researches showed that patients with uncomplicated DM, as well as those with pre-diabetes and poor glucose tolerance, have considerably lower global systolic longitudinal strain (GLS). 11 Additionally, GLS was significantly lower in DM patients with obstructive CAD than that of patients without DM. 12 Furthermore, researchers have claimed that GLS could be used to predict DM or CAD, whereas the changes of GLS in patients with both DM and three-vessel CAD are often not detected by resting echocardiography.
In this study, we used 2-D speckle tracking to examine the global and segmental longitudinal strain in 3VD patients with DM. In particular, the effects of DM duration or glucose control state on patients' LV function were assessed. We sought to determine if global and/or segmental longitudinal strain evaluated by 2-D STE could be a viable approach for identifying LV dysfunction in diabetic patients with 3VD.

Study population
Patients admitted to Tongji Hospital, Huazhong University of Science and Technology for evaluation of chest pain or the presence of CAD from November 1, 2018 and November 30, 2020 were analyzed. These patients who received echocardiography and subsequently coronary angiography (CAG) within 7 days for initial screening were included.
In total, 103 patients with 3VDproved by CAG were prospectively enrolled in this study. 13  DM group (3VD-DM) or without type 2 DM group (3VD-non-DM). We retrospectively registered the available previous HbA1c from each of the 5 years before the admission ( Figure S1). When the subjects had more than one HbA1c available during one of the yearly intervals, we registered the HbA1c that was the closest to 12 months before the next. A control group was randomly taken from our database by the observers who were not involved in echocardiographic analysis.

Echocardiographic examination
All echocardiographic studies were performed using a commercially

Speckle-tracking strain analysis
Speckle-tracking strain analysis was performed on each patient with the aid of a single dedicated software to evaluate LV longitudinal function, in terms of GLS. Two-dimension gray-scale harmonic images were obtained from each of three standard apical (apical long-axis, four-chamber, and two-chamber) views, and three consecutive heartbeat cycles of each view were saved in digital format for offline strain analysis. All images were captured at a frame rate of between 40 and 60 fps with a single focus. LV global and segmental longitudinal strains were quantified using Echo Pac (GE Vingmed) by automated function imaging (AFI). 15 The longitudinal speckle-tracking strain was calculated using an automated contouring detection algorithm and allowed the operator further adjustment of the region of interest to improve the tracking quality. The automated algorithm provided quantitative measurements of global and segmental peak systolic longitudinal strains (PSLSs) in a single bull's-eye summary. GLS was determined as the averaged peak strain of 18 segments from the three standard apical views and was expressed as an absolute value in accordance with current guidelines.

Statistical analysis
Continuous variables were expressed as mean values with standard deviation for normally distributed data and median with interquartile range for non-normally distributed data, while categorical variables were expressed as frequencies and percentages. One-way analysis of variance with post hoc analysis by Bonferroni was used to compare continuous variables and the Chi-square test or Fisher's exact test for categorical variables.
Independent associations of GLS with clinical and echocardiographic parameters in 3VD with DM were evaluated through multiple regression analysis.
The intraclass correlation coefficient was used to determine interand intra-observer reproducibility for GLS from 20 randomly selected subjects. For all steps, a p-value of <.05 was regarded as statistically significant. All analyses were performed with SPSS version 20.0 software.

Baseline characteristics
The baseline clinical characteristics of the patients are summarized in Table 1. The patients of 3VD-DM were younger and had higher incidences of risk factors of hypertension than those in the 3VD-non-DM group. In addition, some medications, such as angiotensin-converting enzyme inhibitors and angiotensin II receptor blockers, were more frequently taken by the 3VD-DM patients. There were no significant differences in baseline characteristics of male, weight or body mass index among the 3VD-DM, 3VD-non-DM, and control groups.

Angiography data of the study population
The data of CAG are shown in Table 2. The 3VD-DM group had more incidence of high-grade stenotic lesions (HSL) than that of 3VD-non-DM group (59% vs. 30%; p = .004). Stenosis was found more frequently in the left circumflex artery in the 3VD-DM group than that in the 3VD-non-DM group (25% vs. 8%; p = .017). However, no significant difference of stenosis was observed in the left anterior descending artery and right coronary artery between the 3VD-DM and 3VD-non-DM.

Echocardiography characteristics of the study population
The baseline echocardiographic parameters are presented in Table 3.
The relative ventricular thickness (IVS and LVPW) were significantly thicker, and the peak velocities of late (A) diastolic flow and E/e' ratio were higher in the 3VD-DM and 3VD-non-DM groups, compared with the control group. The E/A ratio and peak velocities of early (E) patients with 3VD-non-DM were significantly lower than those of control group. However, there was no significant difference in these parameters between the 3VD-DM and the 3VD-non-DM groups.

Comparison of global and segmental peak systolic longitudinal strains of the study population
Global and segmental PSLSs of all subjects are shown in Table 4.

The effect of the duration of DM on the GLS in the 3VD patients
The effect of the duration of DM on the 3VD patients was evaluated by strain analysis. According to the duration years of DM, 3VD-DM patients were divided into two subgroups: DM < 5 years and DM ≥5 years. In Figure 2, GLS was significantly lower in the diabetic group with longer disease duration (DM ≥ 5 years), than that of the shorter diabetic duration group (DM < 5 years) (14.25 ± 2.31% vs. 16.65 ± 1.96%, p = .007). In the two subgroups, the global, mid and apical PSLSs of those with DM ≥5 years were significantly lower than of those with DM < 5 years. These findings indicate that worse GLS and segmental PSLSs were observed in the 3VD patients with longer DM duration.  Note: P-value using Chi-square test or one-way analysis of variance. Abbreviations: ACE, angiotensin converting enzyme; ARB, angiotensin II receptor antagonist; DM, diabetes mellitus; HDL, high density lipoprotein; LDL, low density lipoprotein; 3VD, three-vessel disease. *P < .05 compared with the control group. **P < .05 compared with the 3VD-non-DM group.

Reproducibility
The good intra-observer and inter-observer correlations for strain measurement was shown in Table 5. The intraclass correlation coefficients for inter-observer reproducibility of GLS were .979 (95% confidential interval: .946-.989), and the corresponding coefficients for intra-observer reproducibility were .970 (95% confidential interval: .922-.986).

TA B L E 2
Angiographic findings of study patients by group.

DISCUSSION
The present study illustrates an assessment of LV myocardial defor-

LV longitudinal function in T2DM
In the past three decades, several experimental, pathological, epidemiological, and clinical data confirmed the association of DM with cardiac dysfunction. According to the standard echocardiography in our study, diastolic dysfunction was found in the 3VD-DM and the 3VD-non-DM patients, compared with the control group. However, these differences weren't been found between the 3VD-DM and the 3VD-non-DM patients.
Echocardiography-based speckle-tracking strain imaging is an emerging modality to assess LV function. The use of noninvasive strain imaging may provide additional information to aid in perioperative risk stratify caption and management for these high-risk patients, such as DM. 1617 DM patients are more likely to display LV longitudinal myocardial systolic dysfunction, as evidenced by lower GLS. 18 It has been reported that DM patients have lower longitudinal myocardial mechanics, circumferential and rotational mechanics. 19 And diabetic patients might have dysfunction apparent in the longitudinal direction. 20

LV longitudinal function in 3VD patients with T2DM
Diffuse multi-vessel atherosclerosis is frequently present in T2DM patients, before symptoms of ischemia emerges. 22 The frequency of 3VD in patients with DM comorbidity is much higher. Sometimes, the The reasons for the worse GLS in 3VD-DM patients were prone to impair cardiac systolic and diastolic function. Some mechanisms have been reported to play an important role in DM patients' LV longitudinal dysfunction. 24 These changes are observed as changes in fatty acid metabolism, increased apoptosis, activation of the renin-angiotensin system, autonomic neuropathy, and increased oxidative stress. 7,25 All TA B L E 5 Intra-and inter-observer variability for echocardiographic characteristics.  Abbreviation: ICC, intra-class correlation coefficients; PSLS, peak systolic longitudinal strain.
these underlying pathogenetic conditions change the cardiac structure and increase the possibility of myocardial ischemia. 26 Since the 3VD leads to the more seriously myocardial ischemia, the decreased longitudinal strain could be observed more obviously in both 3VD and DM patients.

The effect of the duration and the glucose control level of DM on the GLS of 3VD patients
Several other studies found that the impairment of the GLS in asymptomatic DM patients was progressed with time. 27 The effect of the DM duration on the 3VD patients is unknown. Our study investigated the GLS changes in 3VD patients with different DM duration. It showed the 3VD patients with DM ≥ 5 years had significantly lower longitudinal strain values compared with the duration of DM < 5 years, especially in global, mid, and apical PSLSs. Other researches demonstrated the duration of diabetic disease was the only independent predictor for the decrease in GLS, 20 similar to our results. It seems that the longer DM duration causes worse GLS, even in the 3VD patients. It is postulated that prolonged exposure to hyperglycemia can epigenetically modify gene expression profiles in human cells and that hyperglycemic memory is sustained even after hyperglycemic control is therapeutically achieved. 27 That may be one of the explanation of the longer DM duration causes worse GLS in the 3VD patients.
Then, the effect of the glucose control state on the 3VD patients with DM was evaluated. A study showed that diabetes with poor blood glucose control, as defined by HbA1c ≥ 7%, leads to reductions of LV systolic strain. 28 Other clinical trials have shown that normalization of blood glucose failed to reduce cardiovascular outcomes in the diabetic population. 29 In our study, we found that there was no significant difference between the 3VD patients with HbA1C ≥ 7% or HbA1C < 7%.
This result demonstrates the glucose state has limited effect on the GLS in the 3VD patients. It may come from that myocardial ischemia was severe because of 3VD, which can cause the decreased GLS. Relatively, the damage of uncontrolled glucose level on the cardiac function in the 3VD-DM patients may be limited.

Limitations
There are several limitations to our study. First, the LV strain in the radial and circumferential directions was not evaluated. The automated algorithm used in the present study only permits the assessment of longitudinal LV strain. Then, patients with and without DM differed in some clinical characteristics, including age, con-founding comorbid conditions such as hypertension, and some medications treatment as outlined in Table 1. Although these differences were mostly due to DM, it is also possible that the differences observed in longitudinal strain might be related to the aforementioned differences in clinical characteristics. In our study, patients in the DM group had a higher incidence of hypertension than those in the non-DM group, which may influence the strain difference between groups with and without DM. 30 However, the exclusion of hypertensive patients from the study is unlikely because the incidence of hypertension is very high in patients with CAD. Therefore, the impairment of strains may not be specific to CAD or DM, and further investigations are needed to validate our findings.

CONCLUSIONS
The present study found that cardiac function is impaired mostly in the patients with both 3VD and DM compared with patients with either condition alone, indicating the significant effect of 3VD with DM on the impairment of myocardium. The duration of DM is a significant risk of cardiac dysfunction damage; on the contrary, the glucose control state of diabetics has limited effect on 3VD patients. GLS at rest might be a useful parameter in the early identification of cardiac dysfunction of three-vessel CAD in patients with DM.