Endothelial Dysfunction in Obesity and Insulin Resistance: A Road to Diabetes and Heart Disease


Joslin Diabetes Center, Harvard Medical School, One Joslin Place, Boston, MA 02215. E-mail: enrique.caballero@joslin.harvard.edu


Obesity, insulin resistance, and endothelial dysfunction closely coexist throughout the natural history of type 2 diabetes. They all can be identified not only in people with type 2 diabetes, but also in various groups at risk for the disease, such as individuals with impaired glucose tolerance, family history of type 2 diabetes, hypertension, dyslipidemia, prior gestational diabetes, or polycystic ovary syndrome. Whereas their evident association cannot fully establish a cause-effect relationship, fascinating mechanisms that bring them closer together than ever before are rapidly emerging. Central or abdominal obesity leads to insulin resistance and endothelial dysfunction through fat-derived metabolic products, hormones, and cytokines. Insulin resistance leads to endothelial dysfunction through the frequent association with traditional cardiovascular risk factors and through some more direct novel mechanisms. Some specific and shared insulin signaling abnormalities in muscle, fat, and endothelial cells, as well as some new genetic and nontraditional factors, may contribute to this interesting association. Some recent clinical studies demonstrate that nonpharmacological and pharmacological strategies targeting obesity and/or insulin resistance ameliorate endothelial function and low-grade inflammation. All these findings have added a new dimension to the association of obesity, insulin resistance, and endothelial dysfunction that may become a key target in the prevention of type 2 diabetes and cardiovascular disease.


Obesity is a major worldwide health problem. At the present time, 30.5% of the adult population in the United States is considered to be obese (BMI > 30 kg/m2), and 60% falls into the overweight category (BMI > 25 kg/m2). Some minority groups, such as Hispanic and African Americans, have higher overweight and obesity rates than the white population (1, 2). Obesity is usually the result of the combination of genetic factors with an inappropriate lifestyle, characterized by inadequate nutrition and lack of regular physical activity. It is closely associated with the development of type 2 diabetes, hypertension, dyslipidemia, and cardiovascular disease, among other medical problems (3). An important mechanism by which obesity leads to the development of all of the above metabolic and vascular diseases is the development of insulin resistance. This frequent metabolic abnormality is typically defined as reduced insulin action in peripheral tissues, such as skeletal muscle, adipose tissue, and liver. We currently have a better understanding of how a state of adipose tissue excess, particularly visceral fat, is associated with a continuous production of mediators that impair insulin action in skeletal muscle, like free fatty acids (FFAs)1 and tumor necrosis factor alpha (TNF-α). Other fat cell-derived products that may also affect insulin action in humans, such as interleukin (IL)-6, resistin, and leptin (4), are currently under investigation. Furthermore, excessive adipose tissue is associated with a decreased production of adiponectin, a mediator known to improve insulin sensitivity (5). On the other hand, there is growing evidence that some of these mediators also have some direct and/or indirect effects on the vascular wall (6). Plasminogen activator inhibitor 1 (PAI-1) is typically increased in the obesity/insulin-resistance state and plays an important role in the genesis of vascular abnormalities (7). In addition, obesity and insulin resistance are frequently associated with other traditional cardiovascular risk factors such as type 2 diabetes, hypertension, dyslipidemia, and altered coagulation/fibrinolysis, all components of the insulin resistance or metabolic syndrome (8). In combination, all of the above abnormalities create a state of constant and progressive damage to the vascular wall, manifested by a low-grade progressive inflammatory process and endothelial dysfunction (9).

The study of human endothelial function in health and disease has allowed us to increase our understanding of how vascular disease develops, including atherosclerotic cardiovascular disease, the main cause of death in the adult population in the United States and many countries. Endothelial dysfunction is now considered a key element in the development of atherosclerosis and has also been closely associated with obesity/insulin resistance (9, 10). The purpose of this review is to present updated information on clinical endothelial dysfunction as a manifestation of vascular damage in individuals with obesity/insulin resistance, particularly in those groups that are considered at risk for type 2 diabetes and, hence, for cardiovascular disease. In addition, the effect of some pharmacological and nonpharmacological interventions that reduce obesity/insulin resistance on endothelial function are reviewed.

Obesity/Insulin Resistance in the Natural History of Type 2 Diabetes

Individuals who advance toward the development of type 2 diabetes experience progressive deterioration of glucose tolerance over time. They generally progress from normoglycemia to impaired glucose tolerance (IGT) and/or impaired fasting glucose (IFG) and finally to overt diabetes. Insulin resistance, defined as the decreased ability of insulin to promote glucose uptake in skeletal muscle and adipose tissue and to suppress hepatic glucose output, may be present for many years before the development of any abnormality in plasma glucose levels (11, 12). Insulin resistance has a well-known but not completely defined genetic influence, frequently transmitted along generations in any given family. In addition, obesity, which also has an important genetic component, invariably exacerbates any degree of insulin resistance (11). Thus, obesity and insulin resistance are usually present for many years before the appearance of other abnormalities such as hypertension, dyslipidemia, type 2 diabetes, and cardiovascular disease. In certain individuals, obesity and insulin resistance may be present during childhood and adolescence (13). β-cell dysfunction (decreased insulin secretory capacity) is a necessary pathophysiological defect for obese/insulin-resistant individuals to progress from normoglycemia to IGT and/or IFG and ultimately, type 2 diabetes (14). People that are considered at increased risk of developing type 2 diabetes include those with IGT and/or IFG (true prediabetic state), first-degree relatives of patients with type 2 diabetes, those with hypertension or dyslipidemia (high triglycerides, low high-density lipoprotein-cholesterol), women with a history of gestational diabetes, and those with polycystic ovary syndrome. Some of these groups are also at risk for cardiovascular disease, particularly those in the prediabetic state (12). People from minority groups in the United States, such as Hispanic, African, Asian, and Native Americans, have a particularly high risk of developing type 2 diabetes (15).

Individuals in the aforementioned categories at risk for type 2 diabetes have several characteristics in common: they are usually overweight or obese, have more insulin resistance than the healthy general population, and often exhibit multiple features of the insulin resistance or metabolic syndrome. Obesity and insulin resistance are abnormalities that can be identified throughout the natural history of type 2 diabetes, starting many years before the appearance of any derangement in glucose homeostasis and continuing during the development of the prediabetic state (IGT and/or IFG) and during the full course of type 2 diabetes (Figure 1).

Figure 1.

Obesity/insulin resistance and endothelial dysfunction closely coexist throughout the natural history of type 2 diabetes. They both are present in people with type 2 diabetes and in those in the prediabetic state, which includes people with IGT and/or IFG. Several groups at risk for type 2 diabetes, even at a stage of normal glucose homeostasis, have also been described with these abnormalities.

Endothelial Dysfunction

In recent years, endothelial function has gained increasing attention in the study of vascular health and disease. The endothelium plays a vital role in vascular homeostasis, vascular tone regulation, vascular smooth muscle cell proliferation, trans-endothelial leukocyte migration, and thrombosis and thrombolysis balance. In response to various mechanical and chemical stimuli, endothelial cells synthesize and release a large number of vasoactive substances, growth modulators, and other factors that mediate these functions (16).

Endothelial dysfunction can be defined as the partial or complete loss of balance between vasoconstrictors and vasodilators, growth promoting and inhibiting factors, pro-atherogenic and anti-atherogenic factors, and pro-coagulant and anti-coagulant factors (16). Endothelial dysfunction is now regarded as an early pivotal event in atherogenesis (9) and has been shown to precede the development of clinically detectable atherosclerotic plaques in the coronary arteries (17). It has also been considered an important event in the development of microvascular complications in diabetes (18).

There is no single and standard test to fully assess endothelial function in vivo. Because the endothelium regulates several vascular functions, the assessment of each of them is a potential way to evaluate its integrity. The chosen test should also consider the vascular bed and tissue or organ of interest. A common approach to the evaluation of endothelial function is the assessment of blood flow and vascular reactivity. Invasive and noninvasive techniques, such as catheterization, ultrasound, positron emission tomography, laser Doppler flowmetry, and plethysmography, comprise the list of the various methods to evaluate these functions (19). The assessment of vasodilatory responses in the brachial artery through the use of high-resolution ultrasound images is a widely used method due to its noninvasive and practical nature (20, 21,22). Endothelial dysfunction in the brachial artery highly correlates with endothelial dysfunction in the coronary circulation, which is emerging as an independent risk factor for cardiovascular disease (23). The reliability of the brachial artery ultrasound method depends on controlling several important technical and biological factors (22). The assessment of vascular reactivity is largely based on our current understanding of the nitric oxide pathway, highly important in vascular health and various disease states. Nitric oxide is produced in endothelial cells through the conversion of l-arginine to l-citrulline by the enzyme nitric oxide synthase (type III). Nitric oxide then diffuses to the underlying vascular smooth muscle cells to stimulate the production of cyclic guanosine monophosphate (cGMP) through the activation of the enzyme guanylate cyclase. The increase in cGMP induces vasodilation (Figure 2). Nitric oxide not only produces vasodilation, but it also participates in various processes that are beneficial to the vasculature, such as the reduction of vascular smooth muscle cell migration and growth, platelet aggregation and thrombosis, monocyte and macrophage adhesion, and inflammation. There are various pharmacological and nonpharmacological stimuli that can be used to assess the production of nitric oxide in endothelial cells in vivo as well as its effect on the underlying vascular smooth muscle cells (endothelium-dependent and -independent vasodilation, respectively; Figure 2). The intra-arterial infusion of acetylcholine or methacholine has widely been used to stimulate endothelial cell production of nitric oxide, whereas sodium nitroprusside has been used to directly stimulate vascular smooth muscle cells. The evaluation of vascular reactivity through high-resolution ultrasound includes the measurement of changes in diameter and flow in the brachial artery in response to “shear stress” (flow-mediated or endothelium-dependent vasodilation) and to sublingual nitroglycerin (endothelium-independent vasodilation). Shear stress is the mechanical stimulus that increased blood flow produces on the endothelium to stimulate nitric oxide production. Nitroglycerin works as a direct donor of nitric oxide to vascular smooth muscle cells.

Figure 2.

The nitric oxide pathway is central in vascular health and disease. It also serves as the basis for the evaluation of vascular reactivity in various tissues. The stimulation of the endothelium to produce nitric oxide is used to assess endothelium-dependent vasodilation, whereas the stimulation of VSMCs is used to assess endothelium-independent vasodilation. EC, endothelial cells; NOS, nitric oxide synthase; NO, nitric oxide; GTP, guanosine triphosphate.

Another approach to the assessment of endothelial function is to measure plasma levels of markers of endothelial activation, such as soluble vascular cell adhesion molecule (sVCAM), soluble intercellular adhesion molecule (sICAM), endothelin 1 (ET-1), E-selectin, and others; markers of coagulation/fibrinolysis such as PAI-1, tissue plasminogen activator, or von Willebrand factor (vWF); and markers of low-grade inflammation such as C-reactive protein (CRP), IL-1, IL-6, or TNF-α.

Endothelial Dysfunction in Obesity/Insulin Resistance throughout the Natural History of Type 2 Diabetes

People with type 2 diabetes almost invariably have abnormal endothelial function as determined by the assessment of vascular reactivity and/or by the measurement of plasma markers of endothelial activation, coagulation/fibrinolysis, or inflammation. They have been consistently found to have abnormal small and large vessel reactivity for both endothelium-dependent and -independent vasodilatory pathways, demonstrating that there is not only a reduction in nitric oxide production in diabetes but also a decreased response to its effect in vascular smooth muscle cells (24, 25). These abnormalities are often the result of hyperglycemia, hypertension, dyslipidemia, altered fibrinolysis, and obesity/insulin resistance (26). By the time someone is diagnosed with type 2 diabetes, severe impairment to endothelial function has very likely been established. For instance, our group found the same degree of impairment in vascular reactivity in the skin microcirculation and in the plasma levels of some markers of endothelial activation such as sVCAM and sICAM in obese individuals with type 2 diabetes with and without microalbuminuria (25). This is of clinical importance because microalbuminuria has traditionally been recognized as a marker of endothelial dysfunction; however, other abnormalities that reflect endothelial damage may occur earlier. Both men and women with type 2 diabetes frequently have endothelial dysfunction. Of particular interest is the finding that whereas estrogen enhances vascular reactivity in healthy women, those with type 2 diabetes have severely impaired vascular reactivity and an attenuated response to estrogen stimulation (27, 28, 29).

Several studies have reported the presence of endothelial dysfunction in individuals in different categories of risk for type 2 diabetes, all of whom have obesity/insulin resistance as a predominant feature. Steinberg et al. (10) showed that severely obese (mean BMI, 34 kg/m2) insulin-resistant individuals with normal glucose tolerance have the same degree of impairment in blood flow and vascular reactivity as those people with established type 2 diabetes. These findings suggest that obesity/insulin resistance has a stronger detrimental influence on endothelial function than hyperglycemia, the most important metabolic difference in these two study groups. Not surprisingly, severely obese children have also been reported to have endothelial dysfunction (30).

We reported the results of a study assessing endothelial function and vascular reactivity in two other groups at risk for developing type 2 diabetes, subjects with impaired glucose tolerance and subjects with normal glucose tolerance but with a parental history of type 2 diabetes (relatives) (31). Micro- and macrovascular reactivities were reduced in these two groups compared with healthy controls but were at a better level than in those with type 2 diabetes. Figure 3 shows the results of endothelium-dependent vasodilation in the macrocirculation in these groups. The results are expressed as percent increases in the diameter of the brachial artery in response to reactive hyperemia (shear stress), assessed through high-resolution ultrasound images. People with IGT also had higher plasma levels of sICAM and ET-1, whereas those with parental history of diabetes exhibited higher plasma concentrations of sVCAM and ET-1 compared with controls (Table 1). The BMI and waist-to-hip ratio were significantly higher in those with IGT and in the diabetic group. These two measures were not statistically different in the relatives compared with controls. It was then interesting to observe that endothelial dysfunction was present not only in individuals with IGT but also in those with parental history of type 2 diabetes who were mildly overweight (mean BMI, 28 kg/m2) and who do not yet exhibit typical clinical abnormalities of metabolic syndrome such as impaired glucose tolerance, hypertension, or dyslipidemia. These findings suggest that minimal, not clinically evident, metabolic abnormalities associated with very incipient stages of obesity/insulin resistance are associated with endothelial dysfunction. It is also possible that some genetic abnormalities leading to endothelial dysfunction could be present in people with family history of type 2 diabetes, although another study found that only those first degree relatives with demonstrable insulin resistance had endothelial dysfunction (32). In support of this association, decreased insulin sensitivity has been also associated with abnormal plasma levels of some markers of endothelial activation in healthy volunteers (33).

Figure 3.

Impaired endothelium-dependent vasodilation in people at risk for type 2 diabetes. The brachial artery diameter change in response to reactive hyperemia, also known as flow-mediated dilation (endothelium-dependent vasodilation) is reduced in relatives of type 2 diabetic patients (Relatives), subjects with IGT, and type 2 diabetic patients compared with healthy controls. Results are presented as mean percent increase in diameter over baseline. *p < 0.05 vs. control; **one or both parents with diabetes.

Table 1.  Clinical characteristics and markers of endothelial activation in studied groups
  • Mean ± SD unless stipulated otherwise.

  • *

    p = <0.05 vs. controls.

  • OGTT, oral glucose-tolerance test.

Age (years)48 ± 949 ± 1050 ± 1053 ± 9
Sex (M/F)14/1619/2015/1721/21
Systolic blood pressure (mm Hg)114 ± 11122 ± 8*123 ± 11*124 ± 9*
Diastolic blood pressure (mm Hg)73 ± 879 ± 7*79 ± 8*80 ± 6*
BMI26.4 ± 3.728.0 ± 4.632.6 ± 7.6*32.2 ± 6.3*
Fasting glucose (mg/dL)92 ± 795 ± 7105 ± 10170 ± 58*
2-hours post-OGTT glucose104 ± 23108 ± 20168 ± 20* 
Triglycerides (mg/dL, median range)79 (37 to 209)79.5 (33 to 288)125 (31 to 599)*171 (54 to 507)*
Total cholesterol (mg/dL)194 ± 34198 ± 35219 ± 37*210 ± 38
High-density lipoproteins (mg/dL)54 ± 1549 ± 1748 ± 1342 ± 10*
Low-density lipoproteins (mg/dL)120 ± 31128 ± 36143 ± 33134 ± 36
vWF (%)110 ± 49103 ± 41121 ± 45135 ± 51*
ET-1 (pg/mL)3.8 (1.4 to 12.3)5.3 (1.2 to 33.4)*6.8 (1.3 to 43.1)*5.4 (1.5 to 39.1)*
sICAM (ng/mL)222 ± 57251 ± 89264 ± 56*301 ± 106*
sVCAM (ng/mL)661 ± 176747 ± 171*759 ± 254831 ± 257*

Women with a history of gestational diabetes represent another group at risk for type 2 diabetes. Whereas blood glucose levels may return to normal in a good number of them, some metabolic abnormalities associated with obesity/insulin resistance continue for a long time after the end of pregnancy. Anastasiou et al. reported reduced flow-mediated dilatation in the brachial artery in a group of women with normal glucose tolerance and history of recent gestational diabetes (34). This impairment in vascular reactivity was present in both obese (BMI > 27 kg/m2) and nonobese women (BMI < 27 kg/m2) compared with controls. However, the “nonobese women” had some biochemical abnormalities that could be associated with insulin resistance, such as elevated uric acid levels, representing subtle but important metabolic changes associated with a slight increase in weight and decreased insulin sensitivity.

Hypertensive individuals are also at increased risk for type 2 diabetes. They are often overweight, insulin resistant, and have endothelial dysfunction. Interestingly, Ferri et al. reported that even nonobese (BMI < 26 kg/m2) hypertensive individuals have abnormalities in endothelial function, findings that suggest that hypertension impairs endothelial function independently from the effects of weight (35).

Hypertriglyceridemia, an excellent clinical marker of insulin resistance, is frequently present in obese individuals, and it is considered a good predictor of the development of type 2 diabetes. High plasma levels of triglycerides have been directly associated with endothelial dysfunction (36).

Polycystic ovary syndrome (PCOS) has been recognized as a feature of insulin resistance or the metabolic syndrome (8). Many, but not all, women with PCOS are overweight or obese and insulin resistant. Paradisi et al. found that obese women with PCOS have endothelial dysfunction and resistance to the vasodilating action of insulin (37). Endothelial dysfunction had a close association with elevated androgen levels and insulin resistance in the studied individuals.

Thus, many groups traditionally considered at risk for type 2 diabetes exhibit endothelial dysfunction as an almost invariable characteristic. Because all the individuals in these studies did not have type 2 diabetes, hyperglycemia cannot be considered as a causal factor for impaired vascular function. Some of the individuals in these groups did not have hypertension or dyslipidemia. The most remarkable common features among all these individuals were being overweight or obese and having insulin resistance. Of particular importance is the fact that even those individuals who are mildly overweight and do not yet have significant typical clinical or biochemical abnormalities such as hyperglycemia, hypertension, or dyslipidemia may have impaired vascular function.

Endothelial dysfunction and obesity/insulin resistance frequently and closely coexist throughout the natural history of type 2 diabetes. Figure 1 illustrates this association as well as the various groups of individuals at risk for type 2 diabetes who have been found with these abnormalities.

Potential Mechanisms for the Association of Obesity/Insulin Resistance and Endothelial Dysfunction

Figure 4 integrates the potential mechanisms through which obesity, insulin resistance, and endothelial dysfunction are closely associated. Obesity, particularly “central or abdominal” obesity, usually involving increased visceral fat, leads to an imbalanced production of several metabolic products, hormones, and cytokines (adipocytokines), which favor decreased insulin sensitivity in liver and skeletal muscle and impair endothelial function through direct and/or indirect mechanisms.

Figure 4.

Mechanisms through which obesity, insulin resistance, and endothelial dysfunction are closely associated. Obesity leads to insulin resistance and endothelial dysfunction, mainly through fat-derived metabolic products, hormones, and cytokines (adipocytokines). Insulin resistance leads to endothelial dysfunction and may contribute to obesity. Insulin resistance is frequently associated with other abnormalities that can affect endothelial function, such as hyperglycemia, hypertension, dyslipidemia, and altered coagulation/fibrinolysis. Endothelial dysfunction may also favor insulin resistance.

An increased efflux of FFAs from the more lipolytically active intra-abdominal adipocytes leads to decreased insulin action in liver and skeletal muscle through mechanisms that may affect the intracellular insulin signaling cascade (3, 38). TNF-α is a cytokine overexpressed in some animal models of obesity that seems to have some effects on adipose cells in a paracrine fashion and that reduces insulin action in skeletal muscle (11, 39). The mechanisms by which other adipose cell—derived molecules increase insulin resistance in skeletal muscle are not clear. Leptin is a hormone that affects energy expenditure, satiety, and neuroendocrine function. In theory, it may improve insulin sensitivity by an indirect mechanism related to decreased satiety and body weight regulation and in a more direct way by affecting insulin signaling in muscle (40, 41). Resistin, another fat-derived hormone, has been found to be linked to obesity and insulin resistance in animal models; however, its role in human obesity is not clear (42, 43). Adiponectin is a more-recently described hormone secreted by adipose cells that has antagonistic effects to all of the above molecules (5). It improves insulin sensitivity by enhancing intracellular insulin signaling, and it is reduced in states of obesity/insulin resistance (5, 44).

Some of these fat-derived products can also affect vascular function in addition to inducing insulin resistance (Figure 4). FFAs are closely associated with impaired vascular reactivity, a measure of endothelial dysfunction (45). TNF-α may also have some direct effect in the vasculature, although the mechanisms are not very clear yet (46). Interestingly, leptin may have some direct vasodilatory effects (47). The adipocytokines IL-1 and IL-6 have been widely studied and have been closely linked to endothelial dysfunction and subclinical inflammation (48). IL-6 is a potent stimulus for the production of CRP in the liver, which, in turn, may have some direct deleterious effects in the vascular wall (49). CRP is considered an excellent marker of low-grade inflammation in the vascular wall, a well recognized mechanism in the development of atherosclerosis (50).

Insulin resistance also contributes to this complex cascade of events. In individuals with insulin resistance, an increase in insulin production by pancreatic β-cells usually occurs as an attempt to maintain normal plasma glucose levels (compensatory hyperinsulinemia). Hyperinsulinemia may promote lipogenesis and weight gain as fat cells are usually more sensitive to insulin action than skeletal muscle. The real implication of this theoretical mechanism in the clinical setting is not known.

The mechanisms by which insulin resistance leads to endothelial dysfunction are certainly multiple and complex. All major abnormalities that are part of the insulin resistance syndrome, such as hyperglycemia, hypertension, dyslipidemia, and altered coagulation/fibrinolysis, are directly and independently linked to endothelial dysfunction (Figure 4). It is still unknown whether insulin resistance per se causes endothelial dysfunction and ultimately atherosclerosis. There is some interesting information that favors this concept. Hyperinsulinemia, a surrogate marker for the presence of insulin resistance, was found to be an independent risk factor for cardiovascular disease in nondiabetic individuals in the Quebec Cardiovascular Study (51). In addition, The Insulin Resistance and Atherosclerosis Study found an independent correlation of insulin resistance assessed through intravenous glucose tolerance testing and the intimal medial thickness of the carotid artery, a measure of the degree of atherosclerosis (52).

New insights into a more direct link between insulin resistance and the vasculature come from the observation that insulin has direct vascular effects. Insulin is known to have a direct vasodilatory effect mediated through stimulation of nitric oxide production in endothelial cells (53). In the insulin-resistant state, the ability of insulin to stimulate nitric oxide production in the endothelium is diminished. Interestingly, the stimulation of nitric oxide production in endothelial cells and the stimulation of glucose uptake in muscle and fat tissue by insulin occur through the phosphatidylinositol 3-kinase (PI3-K) and Akt pathway (54, 55). In contrast, other effects of insulin action on the vasculature, including the stimulation of migration and growth of smooth muscle cells and the production of PAI-1, are mediated through the mitogen-activated protein kinase (MAP-K) pathway (56).

Jiang et al. found that Zucker rats, a well-established model of insulin resistance, have decreased insulin-mediated glucose uptake in muscle and decreased insulin-mediated nitric oxide production in endothelial cells (57). These two insulin effects are at least partially mediated through the PI3-K pathway. In the same model, insulin-mediated stimulation of VSCM migration and growth and PAI-1 production, mediated through the MAP-K pathway, is preserved. Extrapolating these findings to a theoretical framework in humans, it could be speculated that individuals who exhibit the insulin resistance syndrome may have an abnormality in nitric oxide production by endothelial cells and, at the same time, a constant stimulation of proatherogenic changes in the vasculature in response to the hyperinsulinemia that frequently accompanies the syndrome. It has recently been reported that pharmacologically induced hyperinsulinemia causes endothelial dysfunction in healthy volunteers (58). However, whether this interesting mechanism of abnormal selective signaling pathways described in animals operates in humans and whether true endogenous hyperinsulinemia is in fact detrimental to the vasculature deserve further investigation.

Many other abnormalities may certainly contribute to the link between insulin resistance and endothelial dysfunction, such as the production of reactive oxygen species, i.e., superoxide (59), inflammation (49), production of cytokines such as TNF-α (46), activation of the renin-angiotensin system (60), and elevation of ET-1 (61). It is also possible that some genetic factors, which are important for vascular function, may be involved in the association of obesity/insulin resistance and endothelial dysfunction. For instance, the polymorphism of ecNOS 4a/b has been demonstrated to substantially influence basal nitric oxide production in health and disease states (62).

The possibility of endothelial dysfunction leading to insulin resistance is under current investigation. It has been suggested that insulin “opens its own way” by producing vasodilation and recruiting capillaries, particularly in skeletal muscle. Therefore, a state of endothelial dysfunction would limit insulin's ability to reach its target organ to produce its metabolic effects (63). However, some studies have not found a close link between glucose uptake and endothelial function (64, 65).

Therefore, a large number of interesting and complex mechanisms closely link obesity, insulin resistance, and endothelial dysfunction in a multidirectional and dynamic framework (Figure 4).

Effect of Weight Loss and the Reduction of Insulin Resistance on Endothelial Function

Further support for the close association of obesity, insulin resistance, and endothelial dysfunction comes from studies that have assessed the effects of weight loss and/or an improvement in insulin sensitivity on human endothelial function. As described in prior sections, increased adipose tissue, particularly in the abdominal region, is associated with the production of several hormones, mediators, and cytokines that impact insulin sensitivity in liver and skeletal muscle and vascular function. Thus, weight loss and/or an improvement in insulin sensitivity through pharmacological agents would lead to a more favorable profile of these fat-derived substances, which would translate into an improvement of endothelial function and a reduction of the low-grade inflammatory state. Several studies now support this concept. Weight loss leads to a reduction in the plasma levels of various adipocytokines, to an attenuation of the pro-inflammatory state, and to improvement in endothelial function (66, 67, 68, 69). Hamdy et al. recently reported the effect of a 6-month weight loss program on endothelial function in severely obese, insulin-resistant subjects (70). This intervention resulted in a significant reduction in body weight (6.6 ± 1%) and marked improvement in insulin sensitivity. Flow-mediated dilation in the brachial artery significantly improved (12.9 ± 1.2% vs. 7.9 ± 1.0%, final vs. baseline, respectively; p < 0.001) with a linear relationship with percentage weight reduction. Similar observations were seen when the subjects were subclassified according to their glucose tolerance into groups of individuals with normal glucose tolerance, impaired glucose tolerance, and type 2 diabetes. A significant reduction in sICAM and PAI-1 levels was also seen.

Whether weight loss directly translates into an actual reduction in cardiovascular events in people with obesity and insulin resistance is still not known. The lifestyle modification arm of the Diabetes Prevention Program, which included weight loss, demonstrated an impressive reduction in the progression to type 2 diabetes in subjects with IGT after a mean follow-up of only 2.9 years (71). However, revealing the impact of this strategy on cardiovascular outcomes may take many more years. The LookAhead Program funded by the NIH in the United States is currently ongoing and aims at assessing the effects of sustained weight loss on cardiovascular outcomes in obese individuals with type 2 diabetes.

There are several drugs that can affect weight and/or insulin resistance. Among these drugs, there are two classes that are widely used in people with obesity and type 2 diabetes. These are the thiazolidenediones (TZDs) and the biguanides (metformin). TZDs directly reduce insulin resistance by enhancing insulin action in peripheral tissues and, therefore, have been attractive candidates for studying the effects on vascular function of reducing insulin resistance. The main mechanism of action of these drugs is through activation of the nuclear peroxisome proliferator-activated receptor-γ (PPAR-γ) and regulation of gene expression in adipose tissue and, to a lesser degree, in skeletal muscle (72). They may promote weight gain by increasing subcutaneous fat and through fluid retention. However, TZDs actually decrease the visceral fat content, which may be a beneficial effect. PPAR-γ is also expressed in all major cells of the vasculature, including endothelial cells, vascular smooth muscle cells (VSMCs), monocytes, and macrophages (73, 74). In vitro studies have shown that the activation of PPAR-γ in the vasculature inhibits VSMC proliferation and migration and decreases foam cell formation and inflammation (75). A good amount of work on vascular cells and animals has demonstrated the great potential of TZDs in preventing or delaying atherosclerosis. From the clinical perspective, long-term clinical trials with TZDs and cardiovascular endpoints are currently in progress. Meanwhile, increasing information in regard to the beneficial effects of TZDs on endothelial function in humans is rapidly becoming available.

In 1997, troglitazone was the first TZD to become available in the United States. After a few years, it was withdrawn from the market due to its liver toxicity. Nevertheless, interesting information was obtained through various studies with this medication. Several studies showed the significant impact of troglitazone on glycemia, insulin resistance, and traditional cardiovascular risk factors (76). Some data suggest that troglitazone had some anti-inflammatory effects (77). We conducted a randomized, double-blind, placebo-controlled trial to investigate the effect of troglitazone on the vascular reactivity in the micro- and macrocirculation and on the endothelial function of people with early and late type 2 diabetes with and without macrovascular complications (78). We studied 87 type 2 diabetic patients who were divided into three groups. Group A consisted of patients with recently diagnosed diabetes (within the prior 3 years) and no clinical manifestations of macrovascular disease, group B included those with long-term diabetes and no clinically evident macrovascular disease, and group C was comprised of those with long-term diabetes with documented macrovascular disease (cardiovascular, cerebrovascular, or peripheral vascular disease). The flow-mediated dilation (FMD) improved only in the troglitazone-treated patients in group A [7.72 ± 3.4 vs. 5.27 ± 2.0, p < 0.05, (final vs. baseline, percent increase in brachial artery diameter, mean ± SD)]. The fasting insulin level also significantly improved in this group (19.7 ± 10 vs. 15.6 ± 10, final vs. baseline p < 0.05) and was strongly correlated to changes in FMD (r = −0.73, p < 0.01). No changes were found in the FMD in the troglitazone-treated patients in groups B or C. The nitroglycerin-induced dilation was not changed by troglitazone treatment in any of the three groups. No differences were found in the skin microcirculation reactivity measurements or in the biochemical markers of endothelial dysfunction in the three study groups. We concluded that troglitazone treatment for 12 weeks improved endothelial function in the macrocirculation only in those patients with recently diagnosed type 2 diabetes and no clinical evidence of macrovascular disease (78). These findings suggest that an improvement in endothelial function with TZDs may not be evenly seen throughout the many years of exposure to diabetes. These findings also suggest that it is during the early stages of the disease that it seems more prudent and effective to reduce insulin resistance.

There are two currently available TZDs, pioglitazone and rosiglitazone. They both significantly reduce insulin resistance and improve glycemic control in people with type 2 diabetes. They also seem to have an overall positive effect on blood pressure, lipids, and fat distribution (76). Despite not having hard cardiovascular outcome data with these drugs yet, some initial clinical experience with pioglitazone suggests that they may have a beneficial effect in the atherosclerotic process (79,80). They have also been reported to have a beneficial impact on a number of nontraditional cardiovascular risk factors and markers of endothelial dysfunction. Rosiglitazone has been found to reduce serum levels of matrix metalloproteinase-9 (implicated in the pathogenesis of plaque rupture) as well as CRP, a marker of inflammation in patients with type 2 diabetes (81). TZDs also reduce PAI-1 levels (82). Both pioglitazone and rosiglitazone have also been reported to reduce urinary albumin excretion in type 2 diabetes, an effect that reflects improved vascular function (83,84). Therefore, through studies mostly conducted in people with type 2 diabetes, TZDs have demonstrated very favorable vascular effects. There are now some prospective, long-term, randomized clinical trials aiming at assessing the effect of pioglitazone and rosiglitazone in people with IGT on the prevention of type 2 diabetes and cardiovascular disease. These studies will be crucial for ultimately establishing the usefulness of these drugs for the prevention of cardiovascular disease.

Metformin is a biguanide that primarily reduces hepatic glucose output with a more subtle effect on decreasing insulin resistance in skeletal muscle and fat. It has been shown to reduce cardiovascular morbidity and mortality in people with type 2 diabetes (85,86). The protective cardiovascular effect seems to be the result of a positive effect on lipids, glucose, blood pressure, and weight (87). Consistent with the beneficial effects of metformin on cardiovascular risk factors, a study recently conducted in people with type 2 diabetes showed an improvement in endothelium-dependent vasodilation assessed through forearm plethysmography. This improvement was associated with a reduction of insulin resistance (88). In the BIGPRO1 study, metformin use led to a reduction in PAI-1 levels in obese nondiabetic subjects, primarily in those that lost weight (89). In the same study, metformin had a beneficial effect on tissue plasminogen activator and vWF, independent of the effect on weight (89). In a group of subjects with IGT, we recently found that metformin improved plasma levels of some markers of endothelial activation, such as sVCAM, sICAM, and vWF, whereas it had no significant effect on some markers of inflammation, such as CRP and TNF-α (90). Therefore, whereas metformin has a very weak anti-inflammatory effect, it seems to have some favorable effects on endothelial function, partially through its effect on weight and other mechanisms that have not yet been identified.

Thus, some nonpharmacological and pharmacological strategies targeting obesity and insulin resistance improve human endothelial function and reduce low-grade inflammation, strengthening the associations among weight, insulin sensitivity, and vascular health. Long-term clinical trials may demonstrate that all of these beneficial effects on endothelial function translate into a significant reduction in clinical cardiovascular outcomes in people with obesity and insulin resistance.


Obesity and insulin resistance are major metabolic abnormalities in the natural history of type 2 diabetes. These abnormalities are often identified many years before the appearance of any impairment in glucose homeostasis and usually continue during the prediabetic state and during the full course of type 2 diabetes. Endothelial dysfunction is a pivotal event in atherogenesis and cardiovascular disease that can be identified in the clinical research setting by evaluating blood flow and vascular reactivity and by measuring various plasma markers of endothelial activation, coagulation/fibrinolysis, and inflammation. Obesity, insulin resistance, and endothelial dysfunction closely coexist throughout the natural history of type 2 diabetes. The mechanisms by which they are interrelated are numerous and complex. Visceral fat-derived metabolic products, hormones, and cytokines play a major role in affecting insulin action in skeletal muscle and in creating a state of low-grade inflammation and endothelial dysfunction. Some specific and shared insulin signaling abnormalities in muscle, fat, and endothelial cells may also contribute to this interesting association. Nonpharmacological and pharmacological interventions targeting obesity and/or insulin resistance demonstrate an amelioration of endothelial dysfunction and low-grade inflammation. These strategies, particularly in individuals at risk for type 2 diabetes, may improve clinical cardiovascular outcomes.


No outside funding/support was provided for this review.


  • 1

    Nonstandard abbreviations: FFA, free fatty acid; TNF-α, tumor necrosis factor α IL, interleukin; PAI-1, plasminogen activator inhibitor 1; IGT, impaired glucose tolerance; IFG, impaired fasting glucose; sVCAM, soluble vascular cell adhesion molecule; sICAM, soluble intercellular adhesion molecule; ET-1, endothelin-1; vWF, von Willebrand factor; CRP, C-reactive protein; PCOS, polycystic ovary syndrome; TZD, thiazolidenediones; PPAR-γ, peroxisome proliferator-activated receptor-γ VSMC, vascular smooth muscle cell; cGMP, cyclic guanosine monophosphate; FMD, flow-mediated dilation; PI3-K, phosphatidylinositol 3-kinase; MAP-K, mitogen-activated protein kinase.