In the early stages of DC, echocardiography can be the prior method for assessing cardiac structural and functional abnormalities in diabetic patients due to its inexpensive and non-invasive advantages (16,17). Tissue Doppler imaging (TDI) measures left ventricular isovolumetric diastolic time, early mitral annular diastolic motion velocity (Ve), late diastolic motion velocity (Va) and the Ve/Va ratio (18). By comparing these data, cardiac diastolic function images can be divided into 2 types: Restricted mitral flow pattern and unrestricted mitral flow pattern. Unrestricted mitral flow patterns include normal, impaired and false normalization (16–18). This technique has high sensitivity and specificity, and is currently considered to be a reliable method for evaluating left ventricular diastolic function in patients with coronary heart disease (19,20). Intravenous contrast echocardiography is based on the elevated sonic reflectivity of contrast agents; therefore, the contrast injection can aid in the accurate monitoring of the endocardial borders, and the measuring of ventricular size and ventricular motion. The complement of intra-cardiac echocardiography (ICE) into 2D-echocardiography provides an improved evaluation of left ventricular function (21). A previous study demonstrated that alterations in the contractility, systolic function and microcirculation can be well observed via ICE in T1DM murine models (22).

Speckle tracking echocardiography (STE) is a novel imaging technique for achieving mechanical deformation of the myocardia (23). It uses the images acquired from echocardiography to quantify the inter-pixel distance during the cardiac cycle. Therefore, morphological alterations of myocardial deformation can be detected (23,24). It has been demonstrated that the diabetic microangiopathy in T2DM patients is closely associated with the myocardial deformation through 3D-STE (25). In this regard, ~24% of diabetic patients are diagnosed with systolic functional impairment by 2D-STE, which cannot be detected through conventional methods (26).

MRI is a highly sensitive tool that not only detects abnormal wall motion and cardiac hypertrophy, but also provides information concerning arrhythmias and cardiomyopathy. Rijzewijk et al (27) believe that contrast-enhanced MRI can be used to predict the progression of ventricular arrhythmias, myocardial steatosis and heart failure in patients with a history of ischemic heart disease. Cardiac MRI can detect myocardial metabolic abnormalities through different radioactive elements and positron emission tomography (PET) (28).

Cardiac catheterization is currently the best technique for assessing intracardiac hemodynamics and is considered to be the gold standard for the diagnosis of diastolic cardiac dysfunction, but it is not widely used due to its invasiveness (29,30). Left and right ventricular catheterization can be employed to assess left ventricular end-diastolic pressure and mean pulmonary arterial wedge pressure, respectively (29,30). The left ventricular end-diastolic pressure >16 mmHg or a pulmonary capillary wedge pressure >12 mmHg can be used as criteria for the diagnosis of diastolic dysfunction (31). Coronary angiography is mainly used to determine coronary artery stenosis, which tends to occur in late DC (32–34).

Previous studies have found that extracellular matrix proteins are associated with persistent ventricular remodeling (35,36). MMPs are endogenous proteolytic enzymes that degrade type I and type III collagen during the development of heart failure, thereby promoting myocardial fibrosis; they can also affect the expression of microRNAs (miRNAs) and cause myocardial contractile dysfunction (37–39). During the process of myocardial fibrosis, serum MMPs (especially MMP-9) are increased, while tissue inhibitory factors of MMPs are decreased (40). Ban et al (41) measured serum MMP-7 in patients with DC via enzyme-linked immunosorbent assay (ELISA), and found that MMP-7 was significantly higher in patients with diastolic dysfunction than in patients with normal diastolic functions.

Hcy is a type of sulfur-containing amino acid that is an intermediate metabolite of methionine. It has been reported that plasma Hcy levels are positively correlated with the risk of cardiovascular events (42). For every 5 µmol/l increase in plasma Hcy, the risk of ischemic heart disease is increased by 32%, and for each 5 µmol/l decrease, the risk is reduced by 16% (43). Previous studies have shown that high Hcy (HHcy) and hyperglycemia can both lead to DC by inducing oxidative stress and reducing the level of peroxisome proliferator-activated receptor γ (PPAR-γ) (44–46), but whether there is a synergistic mechanism between them is not yet clear. Mishra et al (44) found that in HHcy and hyperglycemic murine models, the ventricular end-diastolic diameter was increased, and PPAR-γ, tissue metalloproteinase-4, and thioredoxin inhibitor levels were decreased, whereas those of MMP-9 were increased, suggesting that endogenous Hcy reduces the expression of PPAR-γ and induces the decoupling of endothelial-myocytes (EMs), resulting in diastolic dysfunction and further progression of DC.

The serum procollagen type III aminopeptide (PIIINP) level reflects the metabolism of type III collagen. A previous study proposed that PIIINP is an important indicator of early left ventricular dysfunction in obese patients with insulin resistance (47). Epshteyn et al (48) found that 96% of diabetic patients with left ventricular dysfunction had increased levels of brain natriuretic peptides (>90 pg/ml). Cardiac troponin is a type of protein released from ischemic or inflammatory disease-damaged cardiomyocytes. Russell et al (49) reported that the levels of troponin T were elevated in the blood of neonates with congenital cardiomyopathy or cardiac insufficiency in mothers with diabetes. However, in adult DC patients, its clinical significance remains uncertain.

Early diagnosis and intervention of DC can slow disease progression, and provide an improved prognosis. Therefore, a combination of multiple techniques for the diagnosis of DC is recommended. For example, a combination of NT-proBNP quantification and echocardiography has more reliable diagnostic value for DC diagnosis than either technique alone (50). In addition, further identification of novel biomarkers is needed, such as microRNAs and long non-coding RNAs (lncRNAs).

Hyperglycemia plays a key role in the pathogenesis of DC. Hyperglycemia has been shown to directly or indirectly damage cardiomyocytes, fibroblasts and endothelial cells via the accumulation of reactive oxygen species (ROS) (51). Hyperglycemia can increase the production of ROS through the electron chain, which can induce apoptosis of cardiomyocytes (52). ROS activate polyADP-ribose polymerase (PARP), consequently causing an increase in glycosylation and inhibition of glyceraldehyde phosphate dehydrogenase (GAPDH), which transforms the glycolysis process into a cascade of cardiomyocyte injury (53). This process involves increased production of glycation end products, activation of hexosamine pathways, and increased production of protein kinases (PKCs). High levels of ROS, PARP, glycation end products, and aldose reductase production induced by hyperglycemia can lead to cell apoptosis (54,55).

In the diabetic state, increased hepatic fat synthesis and fat lysis in adipocytes promote the formation of circulating fatty acids and triglycerides (56,57). The transportation of fatty acids into cardiomyocytes requires the intermediation of insulin, hyperinsulinemia and hyperlipidemia can increase the fatty acid transportation, and thus induce lipotoxicity in cardiomyocytes (58). The following three mechanisms are included in lipotoxicity: i) Increased production of ROS [increased levels of oxidized fatty acids can increase the mitochondrial membrane potential and thus increase production of ROS (55,59)]; ii) increased production of ceramide [ceramide is a sphingomyelin that induces apoptosis of cardiomyocytes by inhibiting the mitochondrial respiratory chain (59)]; and iii) effects on myocardial contractile function. Elevated fatty acid levels in cardiomyocytes can cause the opening of potassium channels, resulting in shortening of the action potential duration and opening of L-type calcium channels, ultimately affecting the calcium storage of calcium pump of sarcoplasmic reticulum and impairing myocardial contractile function (60,61). Excessive fatty acid uptake and metabolism not only cause the accumulation of intermediate metabolites of fatty acids, increase the oxygen demand and uncoupling of mitochondria, but also increase the production of ROS, reduce the synthesis of ATP, cause mitochondrial dysfunction, and increase apoptosis (62). Therefore, hyperlipidemia plays a core role in the pathogenesis of DC.

Insulin resistance and subsequent hyperinsulinemia are typical pathophysiological characteristics of diabetes mellitus. In the normal heart, 2/3 of the energy provided for myocardial contraction is derived from fatty acid oxidation, and the remaining 1/3 is derived from glucose and lactic acid metabolism (63). Glucose utilization is significantly limited in patients with hyperinsulinemia or insulin resistance, therefore myocardial energy metabolism is more dependent on fatty acid oxidation (64). Moreover, hyperinsulinemia induces cardiomyocyte hypertrophy through a variety of mechanisms. Multiple epigenetic and genetic changes caused by hyperinsulinemia lead to the dysregulation of extracellular and intracellular protein expression, especially extracellular matrix proteins, leading to cardiomyocyte hypertrophy and myocardial fibrosis (65).

A balanced regulation of calcium metabolism in cardiomyocytes guarantees normal cardiac contractility (66). Oxidative stress, aggregation of long-chain acetylcarnitine, and changes in lipid membrane components can affect myocardial calcium homeostasis (66–68). Recent reports have shown that calcium pump ATPase, sodium-potassium ATPase and sodium-calcium exchange, and changes in ryanodine receptor function are involved in the pathogenesis of DC (69,70).

The mechanism of the renin-angiotensin-aldosterone system (RAAS) in the development of diabetes to heart failure has long been known (71,72). Research has confirmed that upregulation of the RAAS system in diabetic patients is closely related to cardiac hypertrophy and fibrosis (71–75). Angiotensin directly acts on cardiomyocytes and cardiac fibroblasts via angiotensin receptor-1, causing cardiac hypertrophy and fibrosis (76).

An antioxidant response element (ARE) is a promoter sequence located at the 5′end of the protective gene such as superoxide dismutase (SOD) or glutathione S-transferase (GST), and is a specific DNA-promoter-binding sequence that can be activated by a variety of oxidative and electrophilic compounds, thereby activating phase II detoxification enzymes and antioxidant enzyme gene expression, and protecting normal tissues (91). NF-E2 related factor 2 (Nrf2) is a transcriptional factor that regulates the transcriptional activity of its gene by binding to the 5′ end ARE sequence of its own gene (92). It is believed that Nrf2 plays a key role in cellular oxidative injury protection and the regulation of cell sensitivity to stress. Nrf2 upregulates the expression of antioxidant genes; thus, activation of Nrf2 is beneficial for cells to resist oxidative stress (93). Yoh et al (94) found that hyperglycemia and oxidative stress can accelerate renal injury, and Nrf2 is an important factor in preventing the occurrence of diabetic complications. He et al (95) found that hyperglycemia damages cardiomyocytes through the ROS system and induces DC. Nrf2-depleted rats can quickly develop DC and exhibit increased levels of glucose-induced apoptosis at lower concentrations, therefore, Nrf2 has been suggested to be a key regulator of ARE's anti-oxidative stress and can prevent the progression of DC.

Hemeoxygenase 1 (HO-1), or heat shock protein 32 (HSP32), is a stress protein induced by a variety of factors, and oxidative stress can activate its expression. HO-1 induction is considered to be a common marker of oxidative stress (96). Shi et al (97) showed that activation and intra-nuclear expression of Nrf2 protein can promote the expression of anti-oxidative protein HO-1, which can significantly reduce renal oxidative stress injury in type 1 diabetic mice. The overexpression of HO-1 can protect the myocardia from ischemia/reperfusion injury, and bilirubin produced by HO-1 can protect cardiac function under an ischemic condition (98–102). Therefore, HO-1 is an important antioxidant protease of the Nrf2-ARE pathway. The Nrf2-ARE pathway is a key approach for the body to resist oxidative stress injury. Oxidative stress is an important mechanism of the pathogenesis of DC. Resisting oxidative stress and scavenging ROS are potent strategies for the treatment of DC and have been demonstrated to be an effective anti-oxidative stress enzyme system in vivo, and also an important cardioprotective factor for the prevention and treatment of DC. Activation of Nrf2 is a key step in cellular antioxidative stress. It is important to elucidate whether HO-1 exerts cardioprotective effects against DC by activating the Nrf2-ARE pathway (98).

TGF-β1 and its downstream Smad proteins are involved in the pathological process of myocardial interstitial fibrosis and cardiac hypertrophy (103). In normal hearts, the levels of TGF-β1, type I and type III collagen are extremely low, while diabetic hearts are stimulated by multiple factors such as hyperglycemia, local renin-angiotensin system (RAS), and oxidative stress inflammation. Therefore, the expression of TGF-β1 is upregulated and eventually leads to the differentiation of cardiac fibroblasts into myofibroblasts, which also induces the over-synthesis of collagen, fibronectin and proteoglycans (104). The process of myocardial fibrosis is accompanied by upregulation of TGF-β1 and its downstream Smad2 and Smad3 proteins (105–111). In vascular smooth muscle cells, TGF-β1 promotes the phosphorylation of Smad2 and Smad3 to form a Smad4 trimer (112). This complex translocates into the nucleus and binds to Smad-associated DNA sequences to promote the transcription of fibrogenic factors such as fibronectin and type I collagen; TGF-β1-mediated gene expression in human fibroblasts depends on the presence of Smad3 (105–112). In Smad3 gene-deficient mice, myocardial fibrosis can be reduced by 60%, indicating that Smad3 is essential for the development of myocardial fibrosis (113). At present, myocardial fibrosis still lacks effective treatment methods. In-depth study of the role of TGF-β1/Smad signal transduction pathways in diabetic myocardial fibrosis is expected to provide new and effective therapies for the treatment of DC.

The NF-κB signaling pathway mainly mediates inflammatory responses. It has been shown that NF-κB can activate the expression of inflammatory factors such as interleukin (IL)-1, tumor necrosis factor (TNF) and interferon (IFN), which in turn can promote activation of NF-κB bypass (114). The activation of the NF-κB signaling pathway is closely related to the occurrence of cardiovascular disease in diabetes, and is an important intermediate link in various signaling pathways in the pathogenesis of diabetic vascular complications (115). Excessive glucose in diabetic patients causes non-enzymatic glycation, producing advanced glycation end products (AGEs) that bind to receptors of AGEs and release large amounts of ROS to activate NF-κB translocation into the nucleus, which in turn initiates transcription of inflammatory factors such as TNF-α, ultimately leading to damage of cardiovascular endothelial cells and proliferation of smooth muscle cells, subsequently promoting cardiovascular disease in diabetes (116–127).

At present, it is believed that the pathogenesis of DC is related to inflammation. Feng et al (128) studied myocardial ischemia-reperfusion injury in rats and found that knocking out the MyD88 gene reduced the myocardial infarction size and improved cardiac function compared to wild-type mice. Furthermore, less inflammatory cell infiltration and lower expression of monocyte chemoattractant protein-1 (MCP-1) and intercellular adhesion molecule-1 (ICAM-1) were found in MyD88-deficient mice, suggesting that the knockdown of MyD88 attenuates the NF-κB activation during ischemia-reperfusion.

Protein kinase C (PKC) is a family of serine/threonine kinases that are widely present in cells. Most of the neurotransmitters, hormones and growth factors can activate PKC through activating phospholipase C, which produces a lipid-derived secondary messenger (such as diacylglycerol) that phosphorylates the serine/threonine residue of the target protein (129). Previous studies have found that excessive activation of PKC can lead to cardiac hypertrophy and fibrosis, suggesting that PKC inhibitors may prevent and delay cardiac hypertrophy and myocardial fibrosis (130,131). Hyperglycemia increases the de novo synthesis of diacylglycerols in cardiomyocytes, vascular smooth muscle cells and endothelial cells, which in turn activates PKC. In addition, Ca²⁺, angiotensin, endothelin, vascular endothelial growth factor (VEGF) and osmotic factors can activate PKC. The possible mechanism is that the above factors bind to the cell membrane receptors, and diacylglycerol is activated by membrane phospholipids to activate PKC (130,131). Similarly, the PKC inhibitory therapy should be an essential part of DC treatment. PKC inhibitors have been initially used in the clinical treatment of diabetic vascular complications, but their long-term efficacy and side effects still need to be further observed (132–135).

Several PPAR isoforms, including PPAR-α, PPAR-β/δ, and PPAR-γ, have been shown to be expressed in cardiomyocytes and act as key regulators for glucose and lipid metabolism, and energetic homeostasis. Interestingly, they are also involved in other cellular events such as inflammation and oxidative stress (136). PPAR-α is relatively highly expressed in the heart, and its activation influences cellular free fatty acid (FFA) uptake and mitochondrial FFA oxidation (137). PPAR-α regulates the assembly and transport of lipoproteins, and regulates the defenses of both oxidant and anti-oxidant. Previous research has shown that overexpression of PPAR-α results in decreased sarcoplasmic reticulum Ca²⁺ uptake, contractile dysfunction, left ventricular hypertrophy and increased B-type natriuretic peptide (138). Conversely, depletion of PPAR-α prevents fasting-induced FFA metabolic gene expression and enhances glucose metabolism (139). With the progression of DC, long-term exposure to increased FFAs was found to result in reduced PPAR-α expression, which was shown to further injure the cardiac function by inhibiting FFA oxidation and increasing intracellular lipid accumulation in rodent cardiomyocytes (140,141). However, clinical studies have suggested that PPAR-α is not significantly overexpressed in the cardiomyocytes of patients with type 2 diabetes mellitus (142,143). The activation of PPAR-α and the following FFA oxidation in cardiomyocytes under DC situation may act as a compensatory mechanism for substrate supply. Furthermore, a reduction in PPAR-α in advanced disease may have maladaptive consequences in terms of cardiac metabolism, including glucotoxicity and functional cardiac abnormalities. The role of PPAR-α in the progression of cardiac function decrease in DC has not been fully understood and warrants further investigation. Similarly, PPAR-β/δ isoforms are also abundantly expressed in cardiomyocytes and can regulate FFA metabolism (144). Enhanced PPAR-β/δ signaling promotes FFA metabolism and vice versa. Additionally, PPAR-γ provides anti-hypertrophic and anti-inflammatory effects in cardiocytes (145). PPAR-γ agonists promote insulin sensitivity and enhance cardiomyocyte glucose intake (136). Therefore, PPAR-γ may help maintain glucose and FFA metabolism, and can protect cardiac function.

MAPK overactivation has been reported to contribute to the pathogenesis of DC and cardiac dysfunction. Erk1/2, p38 MAPK, and JNKs are 3 essential MAPK members that regulate cardiomyocyte growth, hypertrophy and remodeling (146–150). Increased phosphorylation of Erk1/2 and p38 MAPK was observed in streptozotocin-induced DC models (151). Previous research has demonstrated that obesity- or insulin resistance-induced heart failure is associated with the overactivation of S6 kinase 1 and Erk1/2 signaling (152,153). JNK can be activated under the conditions of oxidative stress and inflammation (154). In turn, enhanced JNK signaling in the diabetic heart contributes to oxidative stress, endoplasmic reticulum stress and interstitial fibrosis (4,154). In contrast, inhibition of JNK phosphorylation by a curcumin analog was found to prevent high glucose-induced inflammation and apoptosis in diabetic hearts (154,155). In addition to these observations, JNK may play an important role in cardiomyocyte apoptosis (155). As such, JNK activation has been showed to cause increased cardiomyocyte apoptosis as early as at day 3 and 7 in a type 1 diabetic rodent model (155). Collectively, activation of both MAPK and JNK signaling seems to significantly contribute to the development of DC.

AMPK is a serine/threonine kinase that can detect cellular energy status and regulate energy homeostasis (156). Activation of AMPK is involved in multiple cellular processes such as autophagy (157). Activation of AMPK has been shown to inhibit mTOR, which is a protein kinase that regulates autophagy (158). A previous study showed that high glucose inhibited autophagic activity and the AMPK pathway in a DC cell model, thereby stimulating the interaction of beclin 1 (BECN1) and the anti-apoptotic protein BCL2 (159). Activation of AMPK results in phosphorylation of MAPK8, which in turn drives BCL2 phosphorylation and dissociates from the BECN1-BCL2 complex (15). Thus, AMPK restores cardiomyocyte autophagy through BCL2, preventing cardiomyocyte apoptosis. Dissociation of BCL2 from BECN1 via activation of MAPK8-BCL2 signaling may restore autophagy by driving AMPK activation and is important for the prevention of DC processes (160).

DC is associated with the dysregulated expression of miRNAs, short single-stranded noncoding RNAs of ~20 nucleotides in length (161). Importantly, miRNAs bind to the 3′-untranslated region (3′UTR) of the mRNAs of target genes. Therefore miRNAs can participate in the regulation of multiple events in the pathogenesis and progression of DC, such as mitochondrial function, Ca²⁺ handling, ROS production, apoptosis, autophagy and fibrosis (162,163). miR-15a, miR-21, miR-24, miR-29, miR-30d, miR-103, miR-126, miR-146a, miR-150, miR-191, miR-223, miR-320, miR-375 and miR-486 have all been reported to be overexpressed in type 2 diabetic patients (163–168). miR-103, miR-107, miR-143 and miR-181 play essential roles in the regulation of cellular insulin sensitivity and glucose metabolism (163,169). High expression levels of miR-454, miR-500, miR-142-3p/5p and miR-1246 have been found in the circulating blood of patients with cardiac diastolic dysfunction (170). Other microRNAs, such as miR-113a, miR-133a and miR-150, have been reported to be related to cardiomyocyte hypertrophy and interstitial fibrosis (171).