Endothelial dysfunction reflects an imbalance between release of vasodilator and vasoconstrictor endothelial-derived factors. A decrease in the bioavailability of NO, involves either a decrease in NO synthesis or inactivation of NO due to increased endothelial production of reactive oxygen species (ROS) [37]. With increased oxidative stress, tetrahydrobiopterin (BH₄), a cofactor that tightly regulates NO production, is oxidized resulting in the uncoupling of eNOS and reduced NO production [14]. Elevated levels of asymmetric dimethylarginine (ADMA), an endogenous inhibitor of eNOS through competition with l-arginine, may further reduce NO production [13]. This perpetuates a cycle of vascular oxidative stress through the transfer of electrons to molecular oxygen, forming oxidant species such as superoxide and peroxynitrite, which further consumes NO and increases oxidative stress (Fig. 13.1b) [14, 38].

Endothelial dysfunction may also involve altered levels of vasoconstrictors, such as endothelin-1 and angiotensin II, and other vasodilators such as endothelial-derived hyperpolarizing factor (EDHF) and prostacyclin [3, 37].

Several studies in diverse groups of subjects have shown that ED measured by the aforementioned techniques in different vascular beds is predictive of clinical events [39–54]. The principal studies are shown in Table 13.2. Some of these studies included type 2 diabetic patients. In type 2 diabetic patients with normal coronary arteries, coronary artery dysfunction, as assessed by cold-pressor test, was found to predict long-term cardiovascular outcomes, with a relative risk of 4.9 [49]. In a study of asymptomatic type 2 diabetic patients who underwent myocardial perfusion imaging, normal endothelial function had a 93 % negative predictive value in excluding CAD [50]. In patients with newly diagnosed CAD and impaired brachial artery FMD (40 % with diabetes), a persistently impaired FMD was an independent predictor of future cardiovascular events after six months of optimized lifestyle changes and pharmacotherapy [52]. A community based study in 1,016 older adults (72 % with diabetes) demonstrated that impaired forearm endothelial-dependent vasodilation was associated with a five year risk of major adverse cardiovascular events [54]. A recent meta-analysis of 14 observational studies and a recent review, both concluded that FMD is predictive of cardiovascular events and provides prognostic information that is at least equal to the information gained from conventional cardiovascular risk factors; however, future research is required to confirm FMD’s efficacy in the assessment of CVD risk [55, 56].

Endothelial dysfunction has been demonstrated in T2DM in both the resistance and conduit vessels of the peripheral circulation [57–61], as well as in the coronary circulation [62, 63]. Plasma levels of the soluble adhesion molecules E-selectin, vascular cellular adhesion molecule (VCAM)-1 and intercellular adhesion molecule (ICAM)-1 are elevated in subjects with T2DM [3, 29, 30, 64]. Similarly, increased plasma levels of von Willebrand factor (vWF), a measure of endothelial cell damage and activation, are found in diabetes [3, 30, 64]. Microalbuminuria is an independent predictor of ED and may indicate widespread vascular dysfunction in diabetes [3, 65].

The precise pathogenetic mechanisms underlying the development of ED in T2DM remain unclear, but at inception they probably involve uncoupling of both eNOS activity (Fig. 13.1b), and mitochondrial oxidative phosphorylation (Fig. 13.2), as well as the activation of vascular NAD(P)H oxidase. These three mechanisms essentially result in increased generation of superoxide (O₂ ^•−) radicals, eNOS uncoupling, and the overproduction of peroxynitrite. The main factors that combine to cause these biochemical disturbances are dyslipoproteinemia, oxidative stress [4], and inflammation [5–7]. Additional clinical factors that may contribute, either individually or synergistically, to ED in T2DM include hypertension [66], visceral obesity [67], insulin resistance [5, 68, 69], postprandial hyperlipidemia [70–72], fasting and postprandial hyperglycemia [73–75] and elevated levels of ADMA [37, 76].

The impact of insulin resistance in T2DM operates at an insulin signalling level in endothelial cells and in adipose tissue and skeletal muscle [5]. Impaired insulin receptor substrate-1 (IRS-1) and phosphatidylinositol (PI) 3-kinase insulin signalling results in decreased production of NO and ED on the one hand, and decreased glucose transporter (GLUT4) translocation and peripheral insulin resistance on the other. Insulin resistance also increases fatty acid availability which uncouples mitochondrial function in endothelial cells. This generates ROS by increasing advanced glycation end-products (AGES), protein kinase C (PKC) and N-acetylglucosamine (glcNAC), impairing eNOS activity and inducing ED. Inflammation, lipotoxicity and glucotoxicity are all increased in diabetes and collectively contribute to insulin resistance and ED [5]. Figure 13.3 suggests that the pathogenesis of ED in T2DM has oxidative stress as the central pathway for a wide spectrum of risk factors [3].

Strategies for treating ED in T2DM will necessarily target the pathophysiological factors that underlie endotheliopathy, such as hyperglycemia, insulin resistance, dyslipidemia, increased oxidative stress, inflammation and hypertension [77, 78]. Treatment options range from lifestyle interventions to nutritional supplements and specific pharmacological therapies. The results of selected intervention studies are summarized in Table 13.3.

Dyslipidemia is a common risk factor and a strong predictor of CVD in T2DM patients [180]. Elevated plasma concentrations of triglycerides and reduced high-density lipoprotein (HDL)-cholesterol, in both the fasting and postprandial states, are the major lipoprotein abnormalities in diabetic dyslipidemia. The accumulation of small dense low-density lipoprotein (sdLDL) particles and triglyceride-rich lipoproteins (TRLs), including chylomicron remnants and very-low density lipoprotein (VLDL) remnants, are also characteristic of the atherogenic lipid profile [181–184]. These abnormalities are reflected by increased plasma concentrations of non-HDL cholesterol and apolipoprotein B-100 (apoB) [182]. In the postprandial state there is an increase in plasma TRLs and their remnants and qualitative changes in low-density lipoprotein (LDL) and HDL particles [181]. Therefore, hypertriglyceridemia is a marker of a range of lipoprotein abnormalities not routinely measured in clinical practice [182].

The aetiology of diabetic dyslipidemia is complex [185, 186]. It relates collectively to hyperglycemia [187], insulin resistance [187], hyperinsulinemia [187, 188], abdominal visceral adipose disposition and increased liver fat content [188], and dysregulated fatty acid metabolism [188]. Insulin resistance increases fatty acid flux from visceral adipose tissue to the liver, inducing increased liver fat content, over production of VLDL₁ particles [181, 188] and a reduction in the inhibitory effect of insulin on hepatic apoB secretion [189–191]. Hyperglycemia further aggravates the overproduction of VLDL₁, in particular increased VLDL₁ triglyceride production rate [187]. Collectively, plasma glucose, insulin and free fatty acids explain approximately half of the variation in VLDL₁ production rate [187].

Impaired chylomicron clearance in T2DM results from the reduced activity of lipoprotein lipase (LPL), an endothelial bound enzyme, and decreased receptor-mediated endocytosis in the liver [181, 192, 193]. VLDL₁ particles compete with chylomicrons and its remnants for clearance by saturating the lipolytic capacity of LPL and the activity of hepatic receptors [181, 192]. Hepatic secretion of apolipoprotein CIII (apoCIII) is also increased in insulin resistance. This small protein, which is attached to VLDL, contributes to the delayed clearance of TRLs by inhibiting LPL and the binding of remnant TRLs to hepatic clearance receptors [192]. These mechanisms collectively account for postprandial lipaemia [181, 192] and may be an important causal mechanism of ED in T2DM and treatment targets for reducing residual cardiovascular risk.

Important compositional and atherogenic changes in lipoproteins are seen in T2DM [181]. An increased VLDL triglyceride pool leads to cholesterol depletion and triglyceride enrichment of LDL and HDL, mediated via the action of cholesteryl ester transfer protein (CETP) [181, 182]. Increased phospholipid transfer protein (PLTP) activity may contribute to hypertriglyceridemia and compositional changes in HDL. Further, the over activity of hepatic lipase, commonly increased in T2DM, increases the lipolysis of triglyceride enriched LDL and HDL particles [3, 4]. Compositional changes in HDL are also mediated by the actions of lipoprotein lipase (LPL) [181]. Collectively, these compositional changes produce smaller and denser lipoprotein particles that are potentially more atherogenic. Small dense LDL particles more easily penetrate the arterial wall and have a higher binding affinity to intimal proteoglycans than more buoyant larger LDL particles [3, 181, 194, 195]. In the intima, retained LDL particles are modified when exposed to oxidative stress, with sdLDL having an increased sensitivity to oxidation; glycation of LDL further increases this susceptibility to oxidation [181]. A predominant feature of diabetic dyslipidemia is low HDL-cholesterol concentrations with greater reductions in HDL₂ than HDL₃ [181]. In parallel to these reductions in HDL particles are reductions in plasma levels of apolipoprotein A-I (apoA-I) and apoA-II and HDL lipoproteins containing both Apo A-I and Apo A-II (LpA-I:A-II) [3, 181, 196]. These compositional changes in HDL particles are important in respect to ED and atherogenicity, as they are associated with a reduction in rates of reverse cholesterol transport and a decrease in the direct anti-atherogenic effects of HDL, including its anti-oxidant, anti-inflammatory and anti-thrombotic effects [3, 197–199]. These lipoprotein abnormalities and the associated risk of ED and cardiovascular disease in T2DM may be determined by various genes that regulate lipid and lipoprotein functionality, for example, LPL [200], apolipoprotein E [201], apoC-III [202, 203] and CETP [204]. This suggests a connection for the genetic control of ED in T2DM [3].

Current clinical guidelines emphasize lifestyle modifications and pharmacotherapy for the reduction of dyslipidemia and CVD risk in T2DM patients [205, 206].

The rationale for lipid regulation of diabetic dyslipidemia is well supported by studies showing correction of endothelial dysfunction. The results of selected intervention studies utilizing lipid regulating therapy are reviewed below and summarized in Table 13.4.