Introduction
Diabetes mellitus (DM) is a major cause of morbidity and mortality worldwide. (1) Type I DM is characterized by loss of β cell mass and reduced pancreatic insulin secretion with an average age of onset of 7-15 years. Conversely, type II is characterized by β cell dysfunction and peripheral insulin resistance with an average age of onset of 45-65 years. (2) As per the American Diabetes Association, glycated hemoglobin (HbA1c) levels ≥6.5%, random blood sugar levels 11.1 mmol/L, fasting blood sugar levels ≥7 mmol/L, or 2-hour, 75-gram oral glucose tolerance test reading 11.1 mmol/L is diagnostic of DM. (3) Macrovascular complications such as cerebrovascular disease, coronary artery disease (CAD), and peripheral vascular disease, and microvascular complications such as nephropathy, neuropathy, and retinopathy are common in patients with DM. Patients with DM have a higher risk of developing CAD and heart failure (HF) compared with non-diabetic patients. A Finnish population-based study by Haffner et al. showed that type II diabetic patients have a significant high risk of myocardial infarction. The Reykjavik population-based cohort study and Studies of Left Ventricular Dysfunction (SOLVD) trials and registry showed DM to be an independent risk factor for HF and an independent predictor of morbidity and mortality in HF, respectively. (4-6) Diabetic cardiomyopathy is DM-induced myocardial remodeling, fibrosis, and stiffness in the absence of cardiovascular risk factors as CAD, hypertension, and valvular heart disease. (7) In 2005, Fang et al. demonstrated a 27% prevalence of diabetic cardiomyopathy in a cohort of 120 type II diabetic patients without known cardiac disease or left ventricular (LV) hypertrophy. (8) Studies on the prevalence and cardiovascular outcomes of diabetic cardiomyopathy in patients with DM are mainly North American or European. The aim of this study was to explore the prevalence and cardiovascular outcomes of diabetic cardiomyopathy in an Egyptian type II diabetic patient population.
Patients and Methods
Study design
Single group study conducted at four cardiac centers in four tertiary care hospitals located in Cairo, Giza and El Fayoum governerates in Egypt. The study design was approved by the ethics committee, and all participants signed written informed consents.
Study participants
We recruited 200 patients from four hospitals in one country in 2018 and 2019. Study participants were patients referred to the Cardiology Departments at Agouza Police Hospital, National Heart Institute, Cairo University Hospital, and Fayoum University Hospital, Egypt. They were subjected to history taking and data collection for age, gender, body mass index (BMI), smoking, dyslipidemia, medications, DM, CAD, Framingham diagnostic criteria of HF, comprehensive clinical examination, 12-leads resting electrocardiogram, transthoracic echocardiography (TTE), one of the following laboratory investigations: HbA1c level, random blood sugar level, fasting blood sugar level, or 2-hour, 75-gram oral glucose tolerance test, and one of the following medical imaging investigations: cardiac catheterization and coronary angiography, computerized tomography coronary angiography, thallium stress myocardial perfusion imaging, dobutamine stress echocardiography, or cardiac magnetic resonance angiography. Screened participants were enrolled if they had type II DM. Screened participants with CAD, hypertension, congenital heart disease, cerebrovascular disease, valvular heart disease, pacemaker or defibrillator implantation, alcohol, amphetamine or anabolic steroids drug abuse, and/or cancer chemotherapy or radiotherapy were excluded from the study.
Study procedures
Two hundred participants were enrolled, consecutively assigned to a single group, and underwent TTE to detect diabetic cardiomyopathy. In our study, we defined diabetic cardiomyopathy by all of the following criteria: presence of type II diabetes mellitus, myocardial abnormalities in the form of LV dysfunction or hypertrophy, and absence of other causes of cardiomyopathy. Data documented with TTE included ratio of early diastolic mitral inflow velocity (E wave) to late diastolic mitral inflow velocity (A wave) (E/A) by transmitral flow, ratio of E wave to early diastolic mitral annulus velocity (E′ wave) (E/E′) by tissue Doppler imaging, ejection fraction (EF) by modified Simpson’s method, and LV mass indexed to body surface area (LVMi) by 2-dimensional (2D) echocardiography.
E/A and E/E′ cut-off values of 0.75 and 10.0, EF cut-off value of 50%, or LVMi cut-off value of 125 g/m2 for men or 110 g/m2 for women were used to categorize participants into 2 categories. Participants with E/A >0.75 and E/E′ ≥10, EF <50%, or LVMi >125 g/m2 for men or >110 g/m2 for women were classified as diabetic cardiomyopathy participants, while participants with E/A ≤0.75 and E/E′ <10, EF ≥50%, or LVMi ≤125 g/m2 for men or ≤110 g/m2 for women were classified as no diabetic cardiomyopathy participants.
End points
The study evaluated the prevalence and cardiovascular outcomes of diabetic cardiomyopathy in an Egyptian type II diabetic patient population.
Statistical analysis
The echocardiographic assessment data were coded, and the data were analysed with the Statistical Package for the Social Sciences software (SPSS®) version 25. Quantitative data was expressed as means and standard deviations, while qualitative data was expressed as medians and ranges. Parametrically and non-parametrically distributed quantitative variables were compared with the Independent t-test and Mann-Whitney test, respectively. Qualitative variables were compared with the Chi-square test or Fisher exact test. (9,10) The confidence interval and accepted margin of error were set to 95% and 5%, respectively. Any comparison considered statistically significant was at P<0.05 or less, while P<0.01 was considered highly significant. Final data analysis was as per protocol analysis.
Results
We recruited 200 patients in the study. There was male gender predominance in the study (71% were men).
Diabetic cardiomyopathy
Among the 200 participants studied (57 women and 143 men; mean age was 51 ± 7.02 years), the prevalence of LV diastolic dysfunction grade II or III (E/A >0.75 and E/E′ ≥10), LV systolic dysfunction (EF <50%), and LV hypertrophy (LVMi >125 g/m2 for men or >110 g/m2 for women) in the study population was 18.5%, 5.0%, and 8.0%, and the percentages of the diabetic cardiomyopathy vs no diabetic cardiomyopathy participants were 23% vs 77%, respectively (Figure 1). The mean age for the diabetic cardiomyopathy group was 7.8 years older than the average age for the no diabetic cardiomyopathy group (P = 0.008), and the mean BMI for the diabetic cardiomyopathy group was 1.6 kg/m2 higher than the mean BMI for the no diabetic cardiomyopathy group (P = 0.010), respectively. Dyslipidemia (Χ2(1) = 5.860, P = 0.015, V = 0.171), duration of DM (t(198) = -3.440, P = 0.0007), and HbA1c (t(198) = -9.415, P = <0.0001) were significantly different between both groups (Table 1). There was a highly significant statistical difference between the mean EF in the diabetic cardiomyopathy group vs no diabetic cardiomyopathy group (t(198) = -4.963, P = <0.0001). The mean EF for the diabetic cardiomyopathy group was 5.5% lower than the mean EF for the no diabetic cardiomyopathy group. There was a highly significant strong association between diabetic cardiomyopathy and LV diastolic dysfunction grade II and III (Χ2(1) = 151.987, P = <0.0001, V = 0.872), a highly significant moderate association between diabetic cardiomyopathy and LV systolic dysfunction (Χ2(1) = 35.240, P = <0.0001, V = 0.420), and a highly significant strong association between diabetic cardiomyopathy and LV hypertrophy (Χ2(1) = 58.223, P = <0.0001, V = 0.540), with an absolute risk increase of 80%, 22%, and 35% in the diabetic cardiomyopathy group, respectively (Figure 2).
Heart failure
The prevalence of HF (where the echocardiographic findings met the diagnostic criteria of diabetic cardiomyopathy and the clinical findings met the Framingham diagnostic criteria of heart failure) and pre-clinical HF (where the echocardiographic findings met the diagnostic criteria of diabetic cardiomyopathy but the clinical findings did not meet the Framingham diagnostic criteria of heart failure) in the diabetic cardiomyopathy group was 65% and 35%, respectively. There was a highly significant strong association between diabetic cardiomyopathy and HF (Χ2 (1) = 118.159, P = <0.0001, V = 0.769) (Figure 3). The mean age for HF was 4.1 years older than the mean age for pre-clinical HF in the diabetic cardiomyopathy group (58.1 vs 54 years for the preclinical HF) (P = 0.033). Smoking was significantly and strongly associated with HF in the diabetic cardiomyopathy group (Χ2 (1) = 41.851, P = <0.0001, V = 0.954) (Figure 4). Smokers in the diabetic cardiomyopathy group had 20% increased risk for HF (RR= 1.17, 95 CI: 0.7352 to 1.8696). In contrast, there was a non-significant association between dyslipidemia and HF in the diabetic cardiomyopathy group (Χ2 (1) = 4.403, P = 0.078, V = 0.309).
Discussion
Structural alterations of the myocardium have been suggested as probable mechanisms for LV dysfunction in patients with diabetic cardiomyopathy. A nationwide case-control study by Bertoni et al. demonstrated a significant association between DM and diabetic cardiomyopathy (OR = 1.58, 95% CI: 1.55-1.62), after adjustment for hypertension, gender, race, and median income. (11) A retrospective cohort study published in 2010 demonstrated a significant association of LV diastolic dysfunction with subsequent HF in diabetic patients after adjustment for age, gender, BMI, hypertension, CAD, EF, left atrial volume, deceleration time, LVMi and relative wall thickness. (12) Dandamudi et al. cross-sectional study investigated the risk of developing LV dysfunction in patients with diabetic cardiomyopathy. They found a significant 16.9% prevalence of diabetic cardiomyopathy in the Olmsted County community diabetic patient population. The prevalence of LV diastolic and systolic dysfunction in patients with diabetic cardiomyopathy was 54.4% and 7.3%, respectively. Diabetic cardiomyopathy nearly doubled the risk of LV dysfunction, diastolic dysfunction, and systolic dysfunction after adjustment for age and gender. (13) Our cross-sectional study results were consistent with the results of the Dandamudi et al., and we demonstrated high prevalence of diabetic cardiomyopathy in an Egyptian type II diabetic patient population. In addition, we demonstrated significantly strong associations between diabetic cardiomyopathy and LV diastolic dysfunction grade II and III and LV hypertrophy, significant association between diabetic cardiomyopathy and LV systolic dysfunction, high prevalence of HF in the diabetic cardiomyopathy patients, and a significantly strong association between smoking and HF in patients with diabetic cardiomyopathy.
Our study did not have missing data, allowing robust per protocol analysis; the investigators who analysed and reported the TTE assessment outcomes were blinded to the laboratory results; and, to the best of our knowledge, the echocardiographic features of diabetic cardiomyopathy in the Egyptian type II diabetic patients had not been previously studied.
The study has some limitations that need to be acknowledged. It was a multicentered study with a small sample size. Being a cross-sectional study, did not allow us to investigate the chronological relationship between the TTE assessment outcomes and the HF timeline in patients with diabetic cardiomyopathy. Newer techniques such as the strain rate method, speckle tracking image, and 3D echocardiography are more accurate for evaluation of the LV wall function and thickness compared to transmitral flow, tissue Doppler imaging, modified Simpson’s method, and 2D echocardiography.
Conclusions
Diabetic cardiomyopathy was prevalent in an Egyptian type II diabetic patient population, significantly and strongly associated with LV diastolic dysfunction grade II and III and LV hypertrophy, significantly associated with LV systolic dysfunction, and could be considered a primary myocardial disease predisposing to HF. Patients with diabetic cardiomyopathy who smoke increase their risk of developing HF by 20%, an observation which warrants further research.