Diabetes mellitus (DM) is a global health problem. The prevalence of DM worldwide continues to increase, with a projected rise from 425 million in 2017 to 629 million by 2045 globally. The effects of this pathological entity on cardiovascular health (CV) create further public health challenges and remain responsible for the increased prevalence of cardiovascular disease (CVD) [1]. Beyond coronary artery disease (CAD) progression, DM can precipitate or worsen heart failure (HF) due to multifactorial mechanisms (i.e., the accumulation of advanced glycation end products (AGEs), oxidative stress, inflammatory status impairment, decay of intracellular calcium, changes in microRNAs expression) ^[2][3][4][2,3,4]. Patients with diabetes develop HF at more than two-times the rate of patients without diabetes [1]. The underlying mechanisms that interconnect DM to HF merit further investigation, although diabetic cardiomyopathy and the frequent concomitant presence of microvascular dysfunction have been implicated as the principal insults. Nowadays, DM and HF have been recognized as sharing similar metabolic pathways; thus, the clinical manifestation of heart failure can be noticed in the entire spectrum of ejection fraction [5]. Thus, HF approach should be through structural, hemodynamic and functional evaluation, while the need for the early detection of functional and morphological alterations in the cardiovascular system in diabetic patients has emerged. Therefore, the necessity comes forth for evidence-based cardioprotective tools, apart from drugs, such as the implementation of a Mediterranean diet [6].

Nowadays the concept of HF has been disconnected from the presence of reduced ejection fraction [7], while symptoms and signs of HF can be acknowledged in the whole spectrum of left ventricular (LV) volume exchange, which is prescribed as LV ejection fraction (LVEF). Dichotomizing function using LVEF is a major oversimplification, as patients who present with a small LV cavity size and ventricular hypertrophy, or significantly impaired longitudinal contraction may also develop low flow status. HF can be present with normal or reduced end-diastolic volume or even with increased end-diastolic volume but reduced stroke volume [8].

DM has been recognized as sharing similar pathophysiological pathways with HF, even in the absence of clinical recognition. In clinical diabetic cardiomyopathy, rwesearchers usually recognize two faces: the dilated and the restrictive one. The first one may be due to extended coronary atherosclerotic disease or due to the inflammatory process predisposing to dilatation ^[9][10][9,10], while the second one may be due to microvascular dysfunction, infiltration, fibrotic accumulation or even low metabolic state [11].

DM causes several alterations in cardiac energy handling and myocardial performance [12]. In the presence of any disturbance of glucose metabolism, cardiac adenosine triphosphate (ATP) production is affected. Under normal circumstances, the major energy-providing substrates for the heart are fats (triglycerides and fatty acids) and carbohydrates (glycogen, glucose and lactate). Depending on its environment, the heart selects the most efficient fuel for respiration, and continues likewise even when stressed (adaptation). The absence or the overabundance of any fuel may result in metabolic toxicity and contractile dysfunction [13]. The failing heart relies on an exogenous substrate for energy provision due to the lack of endogenous energy reserves. It has been presented that even if the heart is unable to synthesize creatinine, its contractile performance can be preserved under rest conditions. However, this performance is reduced when challenged with an inotropic agent; this can be attributed to increased susceptibility to ischemic injury and electrical instability, which in turn can also lead to a state of low metabolism, where the intrinsic myocardial performance is depressed [14].

DM rarely presents in isolation. Not infrequently, diabetic patients concomitantly present with several other cardiovascular risk factors such as obesity or arterial hypertension (HTN) [15].

Obesity affects myocardial structure and performance through systemic inflammation, epicardial adipose tissue (EAT) accumulation and consequently the development of the clinical condition described as “Cardiac steatosis”. Furthermore, myocardial fibrosis, augmented production of reactive oxygen species (ROS) and reduced production of nitric oxide (NO) by endothelium and increased myocyte stiffness are several mechanisms behind cardiac dysfunction in diabetes [12]. In obese type 2 diabetes mellitus (T2DM) patients, the unfavorable metabolic environment characterized by hyperglycemia, lipotoxicity, abundance of AGEs and hyperinsulinemia can induce coronary microvascular dysfunction, and lead to the development of heart failure with preserved ejection fraction (HFpEF) [8]. Hyperglycemia causes impairment in endothelial NO generation and a reduction in the production of cyclic guanosine monophosphate (cGMP), which in turn down-regulates protein kinase G (PKG) activity in cardiomyocytes and consequently titin protein function, causing diastolic distensibility. Similarly, AGEs impair endothelial NO production and predisposes to concentric LV remodeling and myocardial stiffness as observed in diabetic cardiomyopathy patients with HFpEF ^[8][12][8,12]. In addition, in T2DM there is an increase in glucose autooxidation and free fatty acid concentrations, which can create oxidative stress in the myocardium and consequently induce concentric ventricular remodeling [3].

The EAT exhibits a direct interplay with the heart, in metabolic and mechanica aspects. The lack of muscle fascia between the EAT and the myocardium makes the two tissues dependent on the same microvasculature, whereas, on the other hand, it allows direct paracrine and vasocrine interactions. In fact, it has been shown that in obese patients, the EAT secretes several proinflammatory chemokines and cytokines, collectively called adipokines (including among others, tumor necrosis factor alpha (TNF-α), monocyte chemoattractant protein-1 (MCP-1), interleukin (IL)-6, IL-1β, plasminogen activator inhibitor-1 (PAI-1), resistin and S100 calcium-binding protein A9 (S100A9). All together, they create a proinflammatory state in the myocardium associated with cardiomyocyte stiffness, coronary endothelial dysfunction and fibrosis, which are all implicated in the development of HFpEF [16].

Moreover, high levels of ROS produced by the EAT are responsible for the oxidative stress in the myocardium and the coronary vasculature. As a source of angiotensin II, the EAT triggers coronary vasoconstriction leading to ischemia, especially in patients with DM due to coexisting vasculopathies [17]. Metabolically, EAT expansion is associated with intramyocardial accumulation of triglycerides causing cardiac steatosis. It seems that myocardial triglyceride content is independently associated with reduced pumping function and impaired ventricular strain parameters. Cardiac steatosis induces fetal gene transcription that favors the utilization of myocardial glucose instead of free fatty acids, which further aggravates lipid accumulation, even under physiological conditions ^[16][17][18][16,17,18].

Diabetic cardiomyopathy can represent an important solitary etiology of HFpEF in patients without coronary artery disease, arterial hypertension or other forms of structural heart disease [11]. Despite not being fully elucidated yet, it seems to share similar molecular pathways of impaired energy utilization leading to low performance status.

Those patients diagnosed may have the dilated or restrictive phenotype with concentric LV remodeling and diastolic LV dysfunction and even low levels of natriuretic peptides (NP). Patients with HFpEF and normal NP levels usually display mild diastolic dysfunction and preserved cardiac output reserve during exercise, despite the marked elevation in filling pressures. While clinical outcomes are not as poor compared with patients with high NP, patients with normal NP HFpEF exhibit increased risk of death or HF readmissions compared with patients without HF. These patients also reveal worse right ventricular function, which is associated with low levels of NP, and, more often, secondary valvular regurgitation. The increased morbidity and mortality that they show in comparison with patients without heart failure, emphasize the importance of this phenotype ^[19][20][21][19,20,21].

The pathophysiology of ventricular arrhythmias in non-ischemic cardiomyopathy (non-ICM) involves complex remodeling of both ventricles with interstitial and perivascular patchy fibrosis (extensive endocardial scarring is detected by magnetic resonance imaging (MRI)) in 33% of DCM patients, with a propensity for epicardial and mid-myocardial layers, intertwined with surviving myocytes, localized necrosis and cellular infiltrates [22]. In addition to the altered electrophysiological properties of hibernating/stunned myocytes around the scar, fibroblasts themselves have been found to affect ionic currents by providing mechanoelectrical feedback via stretch-activated ion channels, establishing connexin-based gap junctions with surrounding myocytes and altering the orientation of cardiac lamellae. Thus, reentry is the most common (89%) mechanism for sustained Ventricular Tachycardia (VT) in DCM. Other mechanisms include increased automaticity and afterdepolarizations caused by abnormalities in ventricular myocytes’ resting potential (stretch-induced reduction in SERCA activity and backwards NCX function ZIPES) and assisted by the increased levels of circulation catecholamines ^[23][24][23,24].

5. How to Follow up the Progression of Diabetic Cardiomyopathy

Ejection fraction is used to assess LV function, and in the majority of situations, dictates the treatment of cardiovascular diseases [5]. However, LVEF as a measure of LV function has some important limitations. In particular, in DM patients, it loses its prognostic capacity, as there is an altered relationship between LVEF and death or HF in patients with diabetes. The risk for death or HF in a diabetic patient with an LVEF of 40% is equivalent to the risk in a non-diabetic patient with an LVEF of 25% [25]. Deformation imaging (strain and strain rate) using speckle-tracking echocardiography has been shown to be more sensitive than EF in detecting myocardial contractility [26]. A significant advantage over classical Doppler-based measurements of these parameters is the independence of speckle tracking from angle-related effects on measurements (whereas proper alignment is necessary in Doppler techniques). Following the introduction of deformation imaging, mainly including the study of tissue strain and strain rate, a more precise characterization of myocardial systolic and diastolic function has been made possible. More specifically, myocardial strain, longitudinal, circumferential and radial, has appeared to be relatively load-independent, given that they compensate for the initial condition of the myocardium [27]. The capability for real time three-dimensional imaging has allowed the inclusion of transplanar segment motion, occurring because of myofibrillar orientation that would have otherwise not been taken into account, simultaneously reducing examination time. Three-dimensional speckle-tracking-based deformation measurements have been shown to assess LV volume, function, dysynchrony and rotation, even when only subclinical derangements exist and can give more sensitive information of the effects of fibrosis than the estimation of EF. Thus, it has important prognostic value in the case of HFpEF [28]. However, a wide range of correlation has recently been noted between three-dimensional real-time speckle tracking voxel and tagged cardiac MRI (gold standard) circumferential strain measurements, emphasizing the need for further improvement [29]. The most important limitation in applying the above measurements is the requirement for regular cardiac cycles. Following combination of the aforementioned advances, echocardiography has been shown to carry significant prognostic value in DCM patients, regarding not only myocardial function but also arrhythmogenesis and sudden cardiac death [30]. Although segmental strain curves are synchronized in normal myocardium, in DCM, alterations in the curves’ form (signifying abnormal stretching in systole and post-systolic shortening in diastole), peak systolic values (result of reduced contractility) and dispersion of their peak values (highlighting the, often subclinical, dysynchrony) are evident [31]. The diagnosis of EAT expansion is warranted for the identification of EAT-related HFpEF phenotypes. Transthoracic echocardiography can assess EAT thickness, measured as the echo-lucent area between the epicardial surface and parietal pericardium, although it cannot be used to estimate EAT volume, having also relatively poor inter-observer and intra-observer variability among other limitations. The gold standard method for the assessment of EAT volume is cardiac magnetic resonance (CMR). In line with this, the European society of cardiology consensus recommends the use of a stepwise score-based algorithm to diagnose HFpEF [32].