Endothelial dysfunction in pre-diabetes

Cardiovascular complications are the principal cause of morbidity and mortality among patients with NIDDM, mainly as a result of an acceleration of atherosclerosis and increased thrombosis¹. One of the early signs in the development of atherosclerosis is endothelial dysfunction. Vascular endothelium is considered as the organ of first-line defense against atherosclerosis: by secreting a variety of dilating and constricting substances it modulates vascular tone. Furthermore it can affect platelet adhesion and aggregation influencing thrombogenicity of the blood, and over the long term participate in cell proliferation and, consequently, in the development and progression of atherosclerosis^2,3.

The predominant relaxing factor secreted by the endothelium is nitric oxide (NO), which is synthesized from the amino acid L-arginine by constitutive activity of nitic oxide synthase, an enzyme present in endothelial cells. Nitric oxide release activates the smooth muscle cell guanylyl cyclase, leading to increased cGMP production and subsequent relaxation⁴.

Endothelial function may be clinically assessed through measurements of endothelium dependent vasodilation. Several tests have been developed using either pharmacological or mechanical stimuli; a preserved vasodilatory response is an indication of endothelial integrity. Numerous endothelium dependent agonists have been identified, including acetylcholine, serotonin, bradykinin, thrombin and substance P⁵. In addition, mechanical stimuli, such as an increase in blood flow, result in increased shear stress at the endothelial cell surface leading to enhanced synthesis and release of endothelium-derived NO⁶. The above mentioned pharmacological and mechanical stimuly activate receptors that mediate calcium influx into the endothelial cells. The increased intracelular free calcium levels activate the nitric oxide synthase resulting in NO production⁷. Endothelium dependent vasodilation has been studied in both the coronary and peripheral vessels.

Three clinical measures are usually employed in order to assess endogenous NO activity, namely quantitative coronary arteriographic diameter changes in response to varying concentrations of acetylcholine, venous plethysmographic changes in forearm blood flow after acetylcholine infusion and high-resolution ultrasonography. The latter technique is a non-invasive approach suitable for young symptom-free subjects. It assesses changes in the diameter of a conduit artery (such as the brachial artery) caused by a reactive hyperemia induced by a shear stress after vessel occlusion and release⁸. Moreover, it has been shown that abnormalities of flow-mediated dilation (FMD) of the brachial artery correlate significantly with acetylcholine-induced coronary artery vasodilation^9,10. The three above-mentioned techniques include the concomitant measurement of endothelium independent vasodilation induced through the administration of an exogenous NO donor, such as sodium nitroprusside or nitroglycerine to assess the integrity of smooth muscle cell response.

Recently, endothelial dysfunction in both the coronary and brachial arteries has been found to be associated with the presence of the traditional coronary risk factors, before any evidence of atherosclerosis is detected even by intravascular ultrasound. Specifically, it has been associated with hypertension¹¹, hypercholesterolemia¹², smoking¹³, obesi ty¹⁴. Futhermore, improvement in indothelial function has been demonstrated after interventions that are know to reduce cardiovascular risk factors (e.g. exercise¹⁵ reduction of cholesterol in cholesterol lowering trials¹⁶).

Based on the above, an attractive current hypothesis is that the assessment of endothelial function may serve as an integrating index of overall coronary risk factors.

Diabetes mellitus, both type 1^17,18 and type 2^19,20, has been associated with endothelial dysfunction in several studies. The mechanisms responsible for the impaired vasodilation are incompletely understood, as impaired NO activity by either impaired synthesis or increased breakdown or both may be involved²¹. A factor which plays an important role in endothelial dysfunction is hyperglycemia: acute hyperglycemia impairs endothelium-dependent vasodilation in healthy humans in vivo²², while chronic hyperglycemia may lead to formation of free radicals directly by glucose auto-oxidation²³ or as a byproduct of increased prostaglandin synthesis²⁴. Endothelium-derived NO is thus inactivated. Another proposed mechanisms is hyperglycemia-mediated activation of protein kinase C thorugh increases synthesis of diacylglycerol. Activated protein kinase C results in increased generation of vasoconstrictor prostanoids, which contribute to the formation of reactive oxygen species and perhaps decreased NO synthesis and activity^25,26. A third possible mechanisms is through formation of advanced glycosilation end-products which form during hyperglycemia via non enzymatic protein glycosilation, and result in NO inactivation and endothelium dysfunction²⁷. Moreover, patients with type 2 diabetes mellitus usually also have additional cardiovascular risk factors, such as dyslipemia, which in themselves contribute to varying degress of endothelial dysfunction.

Gestational diabetes mellitus (GDM) refers to glucose intolerance that has its onset or first recognition during pregnancy. Pregnancy itself is characterized by the development of significant insulin resistance which, when superimposed on pre-existing defects in insulin action, unmasks even the slightest defects in insulin secretion, resulting in GDM. The vast majority of GDM women will return to normal glucose tolerance after delivery, but are at markedly increased risk to develop type 2 DM in later life. In this way GDM is considered a pre-diabetic state and offers an opportunity to study abnormalities that may appear very early in the natural history of DM type 2^28,29.

In view of the above, we recently assessed endothelial function using FMD of the brachial artery as an index in both non-obese normotensive, normoglycemic women with a history of GDM, and in non obese controls who had normal glucose tolerance during pregnancy³⁰. Both groups of women with previous GDM had markedly decreased FMD in the brachial artery compared with control subjects. Furthermore, repeated studies performed after a one-year period in a subgroup of these previous GDM women showed similar results suggesting that endothelial dysfunction is a consistent findings in women with a history of GDM. Flow-mediated dilation was significantly inversely correlated with serum uric acid levels, basal insulin resistance as calculated by homeostasis model assessment, BMI, and serum total cholesterol levels.

This effect could either be secondary to the known risk factors related to the insulin resistance syndrome or alternatively it could be due to a more basic pre-existing abnormality. An interesting possibility is that early endothelial dysfunction could be related to the chronic insulin resistance that is known to characterize this group³¹. It has been shown that there is a positive correlation between basal vascular endothelial nitric oxide production and insulin sensitivity in healthy individuals³². Furthermore it has been postulated that a defect in the phosphatidylinositol 3-kinase (PI3-K), a signaling pathway of insulin action, may be common in both insulin mediated glucose uptake in muscle cells as well as in insulin stimulated NO production in endothelial cells³³. The highly significant inverse correlation of FMD with uric levels is in agreement with this explanation, as the latter have been associated with insulin resistance^34,35. Similar evidence that endothelial dysfunction is present very early and is associated with insulin resistance is provided by a study of normal glucose tolerant first-degree relatives of patients with DM type 2³⁶.

Apart from the above mentioned hypothesis that this dysfunction may be intrinsic to the insulin resistance state, Took and Goh³⁷ suggest other interesting mechanisms such as a generalized change in cell membrane biophysical propertial that might alter the presentation of insulin and shear receptors³⁸. Another attractive hypothesis, widely discussed currently, is the association of reduced fetal growth with insulin resistance and the future development of type 2 diabetes^39,40. Interestingly Leeson et al⁴¹ showed that low birth weight was associated with impaired endothelial function in 9 to 11 year old children independently of conventional cardiovascular risk factors.

Nevertheless, we cannot exclude the possibility that other abnormalities which accompany the insulin resistance syndrome such as disturbances in lipid metabolism or oxidative stress have a synergistic detrimental effect of endothelial impairment. Dyslipidemia, a feature of the insulin resistance syndrome, is a well-known factor causing endothelial dysfunction^12,42 although according to our findings, non-obese previous GDM women have no significant differences in lipid levels compared to controls. Another possible mechanism that may link the insulin resistance syndrome and endothelial dysfunction is oxidative stress⁴³. We recently showed that oral administration of ascorbic acid, a know anti-oxidant, acutely improves endothelium dependent vasodilation of brachial artery⁴⁴.

In conclusion, it seems that endothelial dysfunction and possible increased risk of atherogenesis is already present in the prediabetic state. Further investigation is needed to explore the full mechanism that links endothelial dysfunction with the insulin resistance state.