Diabetes and Atherosclerosis:  Epidemiology, Pathophysiology, and Management FREE

Diabetes mellitus magnifies the risk of cardiovascular morbidity and mortality.¹ Besides the well-recognized microvascular complications of diabetes, such as nephropathy and retinopathy, there is a growing epidemic of macrovascular complications, including diseases of coronary arteries, peripheral arteries, and carotid vessels, particularly in the burgeoning type 2 diabetic population. Despite this challenge, many primary care physicians have not yet adopted evidence-based management strategies. The traditional therapeutic approaches emphasize glycemic control, which limits microvascular disease but lacks an established benefit in macrovascular disease. Understanding atherosclerosis in diabetes and instituting therapy guided by emerging evidence should improve outcomes in patients. The evidence supports aggressive antiatherosclerotic management strategies upon diagnosis of type 2 diabetes to minimize the risk of cardiovascular morbidity and mortality.

This review of diabetes and atherosclerosis considers 3 main topics: epidemiology of atherosclerosis in diabetes, the pathophysiology of the diabetic blood vessel, and medical and invasive treatment for atherosclerotic complications of diabetes. We focus on type 2 diabetes, characterized by insulin resistance and inadequate beta cell insulin secretion, because these patients represent more than 90% of those with diabetes and atherosclerosis.

Clinical manifestations of atherosclerosis occur primarily in 3 vascular beds: coronary arteries, lower extremities, and extracranial carotid arteries. Diabetes increases the incidence and accelerates the clinical course of each. Fortunately, we now have therapies that, when implemented, can lessen their impact.

Coronary artery disease (CAD) causes much of the serious morbidity and mortality in patients with diabetes, who have a 2- to 4-fold increase in the risk of CAD.² In one population-based study,³ the 7-year incidence of first myocardial infarction (MI) or death for patients with diabetes was 20% but was only 3.5% for nondiabetic patients. History of MI increased the rate of recurrent MI or cardiovascular death events for both groups (18.8% in nondiabetic persons and 45% in those with diabetes). Thus, patients with diabetes but without previous MI carry the same level of risk for subsequent acute coronary events as nondiabetic patients with previous MI. Such results led the Adult Treatment Panel III of the National Cholesterol Education Program to establish diabetes as a CAD risk equivalent mandating aggressive antiatherosclerotic therapy.⁴

Diabetes also worsens early and late outcomes in acute coronary syndromes. In unstable angina pectoris or non–Q-wave MI compared with control, the presence of diabetes increases the risk of in-hospital MI, complications of MI, and mortality.⁵ In the OASIS registry, a 6-nation study of unstable angina and non–Q-wave MI, diabetes independently increased the risk of death by 57%.⁶ The age-adjusted relative risk (RR) of mortality for patients with diabetes in the GISSI-2 trial of fibrinolytic therapy in MI was 1.4 for men and 1.9 for women, regardless of intervention assignment.⁷ In the SHOCK trial of revascularization for MI complicated by cardiogenic shock, the RR of death for patients with diabetes was 1.36 compared with that of nondiabetic patients.⁸ Regardless of the severity of clinical presentation, patients who have diabetes and coronary events experience increased rates of MI and death.

Patients with diabetes also have an adverse long-term prognosis after MI, including increased rates of reinfarction, congestive heart failure, and death.⁶ A Finnish study⁹ on the trends of MI showed that diabetes increased 28-day mortality by 58% in men (hazard ratio [HR], 1.58; 95% confidence interval [CI], 1.15-2.18) and 160% for women (HR, 2.60; 95% CI, 1.71-3.95). In fact, the 5-year mortality rate following MI may be as high as 50% for diabetic patients—more than double that of nondiabetic patients.¹⁰

Epidemiological evidence confirms an association between diabetes and increased prevalence of peripheral arterial disease (PAD). Individuals with diabetes have a 2- to 4-fold increase in the rates of PAD,¹¹ more often have femoral bruits and absent pedal pulses,¹² and have rates of abnormal ankle-brachial indices ranging from 11.9% to 16%.^13- 14 The duration and severity of diabetes correlate with incidence and extent of PAD.¹⁵

Diabetes changes the nature of PAD. Diabetic patients more commonly have infrapopliteal arterial occlusive disease and vascular calcification than nondiabetic cohorts.¹⁵ The Hoorn study¹⁶ examined the rates of PAD among groups ranging from patients with normal glucose tolerance to those with diabetes requiring multiple medications. The 7% prevalence of abnormal ankle-brachial indices in individuals with normal glucose tolerance increased to 20.9% in those requiring multiple hypoglycemic medications.

Patients with diabetes more commonly develop the symptomatic forms of PAD, intermittent claudication and amputation.¹⁷ In the Framingham cohort,¹⁸ the presence of diabetes increased the risk of claudication by 3.5-fold in men and 8.6-fold in women. Worse, diabetes causes most nontraumatic lower extremity amputations in the United States.¹⁹ The RR for lower extremity amputation in patients with diabetes was 12.7 (95% CI, 10.9-14.9) compared with that of nondiabetic patients in the Medicare population and as high as 23.5 (95% CI, 19.3-29.1) for diabetic persons aged 65 to 74 years.¹⁹

Diabetes adversely affects cerebrovascular arterial circulation, akin to its effects in the coronary and lower extremity vasculature. Patients with diabetes have more extracranial atherosclerosis.²⁰ In patients undergoing dental panoramic radiographs, diabetic patients had 5-fold excess prevalence of calcified carotid atheroma.²¹

The frequency of diabetes among patients presenting with stroke is 3 times more than that of matched controls.²² The risk of stroke is increased 150% to 400% for patients with diabetes,^23- 25 and worsening glycemic control relates directly to stroke risk. In the Multiple Risk Factor Intervention Trial (MRFIT)²⁶ of 347 978 men, subjects taking medications for diabetes were 3 times as likely to develop a stroke (P<.01).

Diabetes particularly affects the risk of stroke among younger patients.²⁷ In the stroke population younger than 55 years, diabetes increases the risk of stroke more than 10-fold (odds ratio, 11.6; 95% CI, 1.2-115.2).²⁸ The Baltimore-Washington Cooperative Young Stroke Study²⁹ examined 296 cases of incident ischemic stroke among black and white subjects aged 18 to 44 years. The presence of diabetes markedly increased the odds ratio for stroke, ranging from 3.3 for black women to as high as 23.1 for white men.

Diabetes affects stroke outcome as well. It increases the risk of stroke-related dementia more than 3-fold,³⁰ doubles the risk of recurrence,³¹ and increases total and stroke-related mortality.³²

Although women experience relative protection from cardiovascular disease compared with men in the general population, diabetes blunts the benefit of female sex. Diabetes increases the incidence of MI, claudication, and stroke in women more than in men, equalizing the age-adjusted rates.^{18,25,33- 34} Indeed, outcomes in women with diabetes have lagged compared with that of their nondiabetic cohorts. In the First National Health and Nutrition Examination Survey (NHANES) and the NHANES Epidemiologic Follow-up Survey conducted 10 years later, age-adjusted heart disease mortality decreased in nondiabetic men and women, less so in diabetic men, but increased by 23% in diabetic women.³⁵

The systemic nature of atherosclerosis and its complications implies a commonality of effects on blood vessels. Indeed, diabetes alters functions of arteries that predispose these patients to the development and progression of atherosclerosis.

The abnormal metabolic state that accompanies diabetes causes arterial dysfunction. Relevant abnormalities include chronic hyperglycemia, dyslipidemia, and insulin resistance. These factors render arteries susceptible to atherosclerosis. Diabetes alters function of multiple cell types, including endothelium, smooth muscle cells, and platelets, indicating the extent of vascular disarray in this disease.

A single layer of endothelial cells lines the inner surface of all blood vessels, providing a metabolically active interface between blood and tissue that modulates blood flow, nutrient delivery, coagulation and thrombosis, and leukocyte diapedesis.³⁶ It synthesizes important bioactive substances, including nitric oxide and other reactive oxygen species, prostaglandins, endothelin, and angiotensin II, that regulate blood vessel function and structure. Nitric oxide potently dilates vessels and mediates much of the endothelium's control of vascular relaxation.³⁷ Further, it inhibits platelet activation, limits inflammation by reducing leukocyte adhesion to endothelium and migration into the vessel wall, and diminishes vascular smooth muscle cell proliferation and migration.^37- 39 Taken together, these properties inhibit atherogenesis and protect the blood vessel.

Diabetes impairs endothelium-dependent (nitric oxide–mediated) vasodilation before the formation of atheroma (Figure 1).^40- 41 A number of fundamental mechanisms contribute to the decreased bioavailability of endothelium-derived nitric oxide in diabetes. Hyperglycemia inhibits production of nitric oxide by blocking eNOS synthase activation and increasing the production of reactive oxygen species, especially superoxide anion (O²⁻), in endothelial and vascular smooth muscle cells.⁴² Superoxide anion directly quenches nitric oxide by forming the toxic peroxynitrite ion,⁴³ which uncouples eNOS by oxidizing its cofactor, tetrahydrobiopterin, and causes eNOS to produce O².⁴³

Other common abnormalities in type 2 diabetes also decrease endothelium-derived nitric oxide. Insulin resistance leads to excess liberation of free fatty acids from adipose tissue,⁴⁴ which activate the signaling enzyme protein kinase C, inhibit phosphatidylinositol-3 (PI-3) kinase (an eNOS agonist pathway), and increase the production of reactive oxygen species—mechanisms that directly impair nitric oxide production or decrease its bioavailability once produced.⁴⁵ Production of peroxynitrite decreases synthesis of the vasodilatory and antiplatelet prostanoid prostacyclin.⁴⁶ Thus, as nitric oxide bioavailability progressively decreases, concomitant increases in peroxynitrite further impair production of subsidiary vasodilators.

In addition to reducing ambient concentrations of nitric oxide, diabetes increases the production of vasoconstrictors, most important, endothelin-1, which activates endothelin-A receptors on vascular smooth muscle cell to induce vasoconstriction (Figure 1). In addition to its modulation of vascular tone, endothelin-1 increases renal salt and water retention, stimulates the renin-angiotensin system, and induces vascular smooth muscle hypertrophy.⁴⁷ Endothelin-1 activity may rise in response to insulin-mediated increases in gene expression and receptor formation, stimulation of the receptor for advanced glycation end products, and increased gene transcription induced by oxidized low-density lipoprotein (LDL) cholesterol.^48- 50 Diabetes increases other endothelium-derived vasoactive substances such as vasoconstrictor prostanoids and angiotensin II, and investigation into their pathophysiological relevance in diabetes continues.^51- 52

Migration of T-cell lymphocytes and monocytes into the intima participates integrally in atherogenesis. T cells secrete cytokines that modulate lesion formation.⁵³ Monocytes, upon reaching the subendothelial space, ingest oxidized LDL via scavenger receptors and become foam cells. Localized accumulation of foam cells leads to formation of fatty streaks, the hallmark of early atherosclerotic lesions.⁵⁴ Diabetes augments these pathologic processes. Hyperglycemia via decreased nitric oxide, increased oxidative stress, and receptor for advanced glycation end products activation increases the activation of the transcription factors nuclear factor κB and activator protein 1 (Figure 1). These factors regulate the expression of the genes encoding a number of mediators of atherogenesis; for example, leukocyte-cell adhesion molecules on the endothelial surface, leukocyte-attracting chemokines, such as monocyte chemoattractant proteins that recruit lymphocytes and monocytes into the vascular wall, and proinflammatory mediators found in atheroma, including interleukin 1 and tumor necrosis factor.^55- 57 Lipid abnormalities commonly found in diabetes, such as increased very low-density lipoprotein (VLDL) and excess free fatty acid liberation, also increase endothelial nuclear factor κB and subsequent cell adhesion molecule and cytokine expression.⁵⁸

In addition to enhancing the initiation of atherogenesis, diabetes promotes plaque instability and clinical sequelae. Diabetic endothelial cells elaborate cytokines that decrease the de novo synthesis of collagen by vascular smooth muscle cells.⁵⁹ Diabetes also enhances the production of matrix metalloproteinases that lead to breakdown of collagen.⁶⁰ Collagen confers mechanical stability to the plaque's fibrous cap. When collagen breakdown increases and synthesis decreases, plaques may rupture more readily, a trigger to thrombus formation. Finally, an important modulator of the severity of plaque rupture is the extent of vascular occlusion by thrombus formation. In diabetes, endothelial cells increase production of tissue factor, the major procoagulant found in atherosclerotic plaques along with alterations in soluble coagulation and fibrinolytic factors (Figure 1).⁶¹

Arteries affected by diabetes and atherosclerosis have altered vasomotor function. In particular, patients with type 2 diabetes have impaired nitric oxide–mediated vasodilation, reflecting an abnormality of vascular smooth muscle cell function or signal transduction.⁴⁰ These patients also have decreased vasoconstriction to infusion of endothelin-1 and angiotensin compared with that of controls.^51,62- 63 Most patients with diabetes have peripheral autonomic impairment on diagnosis,⁶⁴ a condition that decreases arterial resistance.⁶⁵ Despite evidence of increased endothelin-1, angiotensin II, and abnormal sympathetic nervous system activity, the mechanism of vascular smooth muscle cell dysfunction and hypertension in diabetes remains unknown.

Diabetes stimulates atherogenic activity of vascular smooth muscle cells. Hyperglycemia activates protein kinase C, receptor for advanced glycation end products, and nuclear factor κB in vascular smooth muscle cells, as it does in endothelial cells. Activation of these systems augments production of O²⁻, contributing to the oxidant-rich milieu.⁴⁵ Vascular smooth muscle cells are integral in the development of atherosclerosis. Once the macrophage-rich fatty streak forms, vascular smooth muscle cells in the medial layer of the arteries migrate into the nascent intimal lesion, replicate, and lay down a complex extracellular matrix, important steps in the progression to advanced atherosclerotic plaque. Arterial vascular smooth muscle cells cultured from patients with type 2 diabetes demonstrate enhanced migration.⁶⁶ As the source of collagen, vascular smooth muscle cells strengthen the atheroma, making it less likely to rupture and cause thrombosis. Indeed, lesions that have disrupted and caused fatal thrombosis tend to have few vascular smooth muscle cells.⁵⁴ Advanced atherosclerotic lesions in diabetic patients have fewer vascular smooth muscle cells compared with those of controls.⁶⁷ Hyperglycemic lipid modifications of LDL may in part regulate the increased migration and then apoptosis of vascular smooth muscle cells in atherosclerotic lesions. Low-density lipoprotein that has undergone nonenzymatic glycation induces vascular smooth muscle cell migration in vitro, while oxidized glycated LDL can induce apoptosis of vascular smooth muscle cells.⁶⁸ Thus, diabetes alters vascular smooth muscle function in ways that promote atherosclerotic lesion formation, plaque instability, and clinical events.

Platelets can modulate vascular function and participate significantly in thrombus formation. Abnormalities in platelet function may exacerbate the progression of atherosclerosis and the consequences of plaque rupture. Intraplatelet glucose concentration mirrors the extracellular concentration, since glucose entry into the platelet does not depend on insulin.⁶⁹ In the platelet, as in endothelial cells, elevated glucose levels lead to activation of protein kinase C, decreased production of platelet-derived nitric oxide, and increased formation of O²⁻.⁷⁰ In diabetes, platelets also show disordered calcium homeostasis.⁷¹ Disordered calcium regulation may contribute significantly to abnormal activity, since intraplatelet calcium regulates platelet shape change, secretion, and aggregation and thromboxane formation. Moreover, patients with diabetes have increased platelet-surface expression of glycoprotein Ib (GpIb), which mediates binding to von Willebrand factor, and GpIIb/IIIa, which mediates platelet-fibrin interaction.⁶⁹ These abnormalities may result from decreased endothelial production of the antiaggregants nitric oxide and prostacyclin, increased production of fibrinogen, and increased production of platelet activators, such as thrombin and von Willebrand factor.⁶⁹ Taken together, diabetic abnormalities increase intrinsic platelet activation and decrease endogenous inhibitors of platelet activity. These mechanisms may explain the enhanced thrombotic potential characteristic of diabetes.

In addition to potentiating platelet function, diabetes augments blood coagulability, making it more likely that atherosclerotic plaque rupture or erosion will result in thrombotic occlusion of the artery. Patients with type 2 diabetes have impaired fibrinolytic capacity because of elevated levels of plasminogen activator inhibitor type 1 in atherosclerotic lesions and in nonatheromatous arteries.⁷² Diabetes increases the expression of tissue factor, a potent procoagulant, and plasma coagulation factors such as factor VII and decreases levels of endogenous anticoagulants such as antithrombin III and protein C.^73- 75 Many of these abnormalities correlate with the presence of hyperglycemia and proinsulin split products.⁷⁶ Thus, in diabetes an increased tendency toward coagulation, coupled with impaired fibrinolysis, favors formation and persistence of thrombi.

Throughout the last decade, the perception of type 2 diabetes has evolved from a focus on dysregulated glucose and insulin to encompass a global metabolic disorder characterized by dyslipidemia, hypertension, and hypercoagulability in addition to hyperglycemia and hyperinsulinemia (Figure 2). Each of these abnormalities plays an important role in cardiovascular disease development and progression and provides targets for therapy.

Hyperglycemia, a cardinal manifestation of diabetes, adversely affects vascular function, lipids, and coagulation. Intensive treatment of hyperglycemia reduces the risk of microvascular complications such as nephropathy and retinopathy, as shown by the United Kingdom Prospective Diabetes Study (UKPDS),⁷⁷ but does not confer the same benefit in the conduit, muscular arteries. In the UKPDS, treatment with either oral hypoglycemic agents or insulin did not significantly reduce macrovascular end points.⁷⁷ Although these data support a distinct pathogenesis of microvascular and macrovascular sequelae in diabetes, they do not exclude glycemic control as an important part of the treatment of the dysmetabolic syndrome. Epidemiological studies support the concept that increasing levels of glycemia commensurately increase cardiovascular events. In the UKPDS, this increase in risk began above a hemoglobin A_1c (HbA_1c) level of 6.2%.⁷⁸ In a meta-analysis of more than 95 000 diabetic patients, increases in cardiovascular risk depended directly on plasma glucose concentrations and began with concentrations below the diabetic threshold.⁷⁹ Continued improvements in therapy will permit even more effective glycemic control and enable retesting of the hypothesis that intense control of glycemia reduces atherosclerotic macrovascular complications.

Insulin resistance, another cardinal feature of type 2 diabetes, may promote atherosclerosis. The degree of insulin resistance relates directly to increasing rates of MI,⁸⁰ stroke,^24,81 and PAD.⁸² In the UKPDS, improving insulin resistance with metformin decreased macrovascular events. This result has engendered some controversy because the addition of metformin to sulfonylurea therapy seemed to increase cardiovascular risk.^77,83 Ongoing clinical trials should determine whether improvements in insulin resistance will decrease cardiovascular events.

The recent availability of the thiazolidinediones (TZDs) provides a novel approach to glycemic control. Thiazolidinediones improve glycemia by decreasing insulin resistance.⁸⁴ The TZDs bind and activate peroxisome proliferator-activated receptor γ, a nuclear receptor that participates in adipose and vascular cell differentiation.⁸⁵ The first agent in this class, troglitazone, was removed from the market because of idiosyncratic hepatotoxicity. Two other TZDs, rosiglitazone and pioglitazone, now available in the United States, appear not to share this liability. Because peroxisome proliferator-activated receptor γ activity may have anti-inflammatory activity, TZDs may directly benefit the atherosclerotic lesion.⁸⁶ Studies in progress will evaluate this possibility.

Characteristic abnormalities in the lipid profile in type 2 diabetes include elevated triglyceride levels, decreased atheroprotective high-density lipoprotein (HDL) levels, and increased levels of small dense LDL. Increased efflux of free fatty acids from adipose tissue and impaired insulin-mediated skeletal muscle uptake of free fatty acids increase hepatic free fatty acid concentrations.⁸⁷ In response, the liver increases VLDL production and cholesteryl ester synthesis.⁸⁸ Free fatty acids combine with a cholesterol molecule to form a cholesteryl ester. Cholesteryl ester concentrations may regulate VLDL production, with increased concentrations resulting in elevated VLDL synthesis.⁸⁹ Overproduction of triglyceride-rich lipoproteins and impaired clearance by lipoprotein lipase lead to the hypertriglyceridemia common in diabetes.⁸⁹

Low levels of HDL represent the second common abnormality in type 2 diabetes. Elevated levels of triglyceride-rich lipoproteins lower HDL levels by promoting exchanges of cholesterol from HDL to VLDL via cholesteryl ester transfer protein.⁸⁹ Diabetic patients with CAD more commonly have the combination of elevated triglycerides and low HDL than elevated total and LDL cholesterol levels.⁹⁰ Functional defects in HDL may also contribute. High-density lipoprotein in diabetic patients does not prevent LDL oxidation as well as HDL in nondiabetic patients does.⁹¹

Diabetic patients commonly have elevated concentrations of small dense LDL even in the face of average plasma LDL levels.^92- 93 This change in LDL particles results from increased VLDL secretion and abnormal cholesterol and triglyceride transfer between VLDL and LDL. The modification depends on increased concentrations of VLDL and usually occurs when triglyceride concentrations exceed 130 mg/dL (1.47 mmol/L).⁹⁴ Small dense LDL particles are proatherogenic, bind more readily to intimal proteoglycans, enhancing their metabolism, and readily undergo oxidative modification, which drives their uptake by monocytes and vascular smooth muscle cells in the vessel wall.^95- 96

The lipid abnormalities that develop in diabetes and their role in atherogenesis have important therapeutic implications. Poor glycemic control exacerbates diabetic dyslipidemia. Thus, strict glycemic regulation may lessen free fatty acid flux and hepatic VLDL production.⁹⁷ Metabolic and lipid abnormalities improve with weight loss, exercise, smoking cessation, and dietary modification⁹⁸; hence, lifestyle modification is the first mode of therapy. Pharmacological interventions also decrease cardiovascular events in diabetes, as borne out by large-scale clinical trials (Table 1). Indeed, diabetic patients at increased risk of recurrent cardiovascular events may experience greater risk reduction by lipid lowering than nondiabetic patients.^99- 100 The 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors (statins) increase LDL clearance and decrease VLDL secretion.¹⁰¹ In the Scandinavian Simvastatin Survival Study,¹⁰⁰ simvastatin reduced the risk of total mortality by 43% in diabetic patients vs 29% in nondiabetic patients and reduced the risk of MI by 55% in diabetic patients vs 32% in nondiabetic patients. In the Heart Protection Study (HPS),¹⁰² which included more than 4000 subjects with diabetes, simvastatin decreased the risk of acute coronary syndrome, stroke, or revascularization by 25% in the diabetic subgroup.

Fibric acid derivatives represent other medications potentially of benefit in diabetic dyslipidemia because they raise HDL levels and lower triglyceride levels. In the Veterans Affairs High-Density Lipoprotein Cholesterol Intervention Trial (VA-HIT),¹⁰³ patients with diabetes represented 25% of the 2531 male participants. Treatment with gemfibrozil reduced risk of MI by 24%, comparable to that observed in nondiabetic patients. As with TZDs, fibrates may have antiatherogenic effects independent of lipid lowering. Fibrates bind to peroxisome proliferator-activated receptor α (as opposed to TZDs, which bind to peroxisome proliferator-activated receptor γ). Peroxisome proliferator-activated receptor α ligation by fibrates can exert anti-inflammatory effects. For example, fibrates decrease endothelial cell activation by proinflammatory cytokines and reduce tissue factor production by human macrophages.^104- 105 Fibric acid derivatives can prove useful in diabetic patients with persistently elevated triglyceride levels and low HDL levels, despite tight glycemic control. Combination therapy of statins and fibrates warrants careful monitoring for muscle injury. Nicotinic acid may also increase HDL levels in diabetic patients, but glycemic control requires supervision in this population.¹⁰⁶

In addition to improving lipid abnormalities via improvements in glycemic control, TZDs may decrease the concentration of small dense LDL¹⁰⁷ and increase the resistance of LDL to oxidation,¹⁰⁸ although total, LDL, and HDL cholesterol concentrations increase.¹⁰⁹ The clinical effects of TZDs on vascular outcomes await clinical trials.

Hypertension, a common comorbid condition in the dysmetabolic syndrome, occurs more often in patients with type 2 diabetes than in matched controls.¹¹⁰ Vigorous control of blood pressure decreases the rate of cardiovascular events more in patients with diabetes than in those without diabetes.^111- 112 Indeed, control of high blood pressure represents the most important intervention, limiting cardiovascular events far more effectively than tight glycemic control (Table 2). In the UKPDS, aggressive blood pressure reduction significantly reduced stroke and deaths.¹¹³ Captopril and atenolol as initial therapy had similar efficacy.¹¹⁴ Achieving American Diabetes Association target blood pressure (130/80 mm Hg) usually requires more than one agent.^{110,112- 113} Diuretics, β-blockers, angiotensin-converting enzyme (ACE) inhibitors, angiotensin receptor blockers, and calcium channel antagonists all effectively decrease blood pressure in diabetic patients. Modulation of the renin-angiotensin system seems particularly important in diabetes. In the recent Heart Outcomes and Prevention Evaluation (HOPE) study,¹¹⁵ ramipril significantly decreased the rates of MI, stroke, and death in patients with diabetes and one additional cardiovascular risk factor. Further, in the Losartan Intervention for Endpoint Reduction in Hypertension (LIFE) study,¹¹⁶ losartan reduced total and cardiovascular mortality more than atenolol did in diabetic patients with hypertension and left ventricular hypertrophy. ACE inhibitors seem to reduce cardiovascular end points more than dihydropyridine calcium channel blockers as well.¹¹⁷ We recommend the use of drugs that inhibit the renin-angiotensin system as an integral component in the treatment of diabetes.

In diabetic patients with CAD, platelet inhibition and anticoagulation assume particular importance (Table 3). Platelet antagonists, as reviewed by the Antiplatelet Trialists' Collaboration Group,¹¹⁸ lowered the combined risk of vascular death, MI, and stroke by 19% in diabetic patients (P<.01). In unstable angina, low-molecular-weight heparin confers event reduction in patients with diabetes similar to that in those without diabetes.¹¹⁹ Platelet GpIIb/IIIa inhibitors also have benefits in unstable coronary syndromes. In the PRISM-PLUS study,¹²⁰ the addition of tirofiban to heparin decreased the rates of death and MI in patients with diabetes more than in nondiabetic patients. A meta-analysis of 10 clinical trials of GpIIb/IIIa inhibitors showed that diabetic patients experienced twice the absolute event rate reduction of patients without diabetes.¹²¹ Several studies have established that diabetic patients benefit from the use of thrombolytic therapy in acute MI.^122- 123 In a meta-analysis of more than 43 000 patients, including 4529 patients with diabetes, the Fibrinolytic Therapy Trialists' Collaborative Group¹²⁴ demonstrated that the reduction in absolute mortality in those with diabetes exceeded that of patients without diabetes, 3.7% vs 2.1% (P<.01).¹²⁴

Practitioners have had reservations regarding the use of β-blockers in diabetic patients because of the perceived risk of masking hypoglycemia and reduced insulin production. However, β-blockers decrease the risk of reinfarction and cardiac mortality of patients with diabetes more than of matched controls.^125- 126 In a retrospective study of more than 45 000 patients,¹²⁶ β-blockers reduced the risk of MI by 23% (95% CI, 0.67-0.88) in patients with type 2 diabetes without increasing diabetes-related complications. We advocate the use of β-blockers in diabetes after MI and urge patient education to decrease the frequency of diabetes-related events.

Diabetic and nondiabetic patients have similar immediate success rates with percutaneous coronary revascularization.¹²⁷ However, soon after diabetic patients leave the catheterization laboratory, they experience clinical events at higher rates than nondiabetic patients do. Patients with diabetes have a clear trend toward increased rates of in-stent thrombosis.¹²⁸ In the National Heart, Lung, and Blood Institute PTCA Registry,¹²⁹ diabetic patients more commonly reached a composite end point that included mortality, nonfatal MI, and emergency surgery (11.0% vs 6.7%; P<.01) and had higher rates of death (3.2% vs 0.5%; P<.01) than nondiabetic patients. Several studies have demonstrated a greater long-term risk of restenosis after balloon angioplasty.^130- 131 Moreover, the severity of diabetes affects outcome after stent implantation. In a study of 954 patients who underwent coronary artery stent implantation,¹³² insulin-requiring patients with diabetes faced a 28% risk of late revascularization and a 40% risk of adverse cardiac events compared with 16.3% and 24% in nondiabetic control patients, respectively. In multivariate analysis, insulin requirement entailed a 2-fold increased risk of adverse cardiac events and target vessel revascularization at 1 year (P<.01).¹³² Increased rates of restenosis may in part result from an increased intimal proliferative response, but the mechanisms remain unclear.^130,133

Diabetic patients also have a worse prognosis following surgical revascularization than nondiabetic patients. In the Bypass Angioplasty Revascularization Investigation (BARI),¹³⁴ diabetic patients had significantly lower 5-year survival rates (73.3% vs 91.3%) than nondiabetic patients. Yet insulin-requiring diabetic patients do significantly better with surgery than with percutaneous intervention.^135- 136 In the BARI trial, diabetic patients who underwent multivessel bypass surgery had significantly higher rates of survival than those who underwent percutaneous interventions, 80.6% vs 65.5%.¹³⁵ The benefit may result from use of the internal mammary artery in the surgical arm.

Despite the marked increase in risk of lower extremity atherosclerosis, we have inadequate information regarding the role of medical therapies in diabetic patients with PAD. No evidence shows that tight glycemic control, aggressive blood pressure management, or the use of antiplatelet agents decreases the incidence of intermittent claudication or critical limb ischemia.¹¹³ Although simvastatin decreased the rates of reported claudication in patients with CAD in the Scandinavian Simvastatin Survival Study, the trial did not report data specific to patients with diabetes.¹³⁷ Even in the absence of definitive data, we recommend that patients with diabetes receive therapies of proven benefit in broader patient populations because cardiovascular events remain the principal cause of death in patients with PAD.¹⁷

Diabetic persons have a particular propensity to develop foot ulcers. Risk factors for diabetic ulcers include male sex, hyperglycemia, and diabetes duration.¹³⁸ Foot ulcers often result from severe macrovascular disease,^139- 140 and diabetic neuropathy exacerbates the risk.^140- 141 Patients with diabetes need intensive self-examination and physician examination of the foot, specifically checking the nails and skin for cracking and small ulceration. Use of therapeutic footwear can decrease the risk of ulceration.¹⁴² Once ulceration occurs, patients with diabetes have a much higher risk of amputation, highlighting the paramount importance of prevention.

Two noninvasive therapies have demonstrated benefit in improving walking distance in patients with PAD: exercise therapy and cilostazol. Supervised exercise therapy produces impressive increases in walking distance. In a meta-analysis of exercise programs, supervised exercise programs increased walking distance 122%.^143- 144 The mechanism underlying the benefit is unclear but may derive from improved cardiovascular fitness, increased production of nitric oxide, or modification of cardiovascular risk factors.^145- 146 Even though we lack data on the benefits of exercise specifically in diabetes, we advocate a prescription for supervised exercise for its benefits in walking distance, risk factor modification, and insulin sensitivity.

Cilostazol, a type III phosphodiesterase inhibitor, antagonizes platelet activity and exerts favorable lipid effects, yet the mechanism by which it improves walking distance remains unclear.^147- 149 Cilostazol increases walking distance by 35% to 50% above placebo in patients with claudication; however, we lack specific information for diabetic patients.¹⁵⁰ Pentoxifylline, a xanthine derivative, may affect blood rheology and decrease blood viscosity.¹⁵¹ Pentoxifylline improves claudication in some studies but not in others, and outcomes specific to patients with diabetes remain unknown.¹⁵²

As in coronary circulation, outcomes of percutaneous revascularization depend on many variables, including lesion location, lesion length, stenosis or occlusion, and the nature of the distal runoff.¹⁴⁴ Diabetes alters the distribution of lower extremity atherosclerosis so that these patients tend to have severe arterial occlusive disease below the knee in the runoff vessels. As distal runoff declines, the results of percutaneous interventions worsen.

The success of iliac artery stenting in diabetic patients varies among studies, but several groups have shown better than 90% patency at 1 year.¹⁴⁴ With femoral artery interventions, 1-year patency rates range from 29% to 80%, and diabetes adversely affects these rates of success.^153- 154 This finding may result in part from poor runoff in patients with diabetes because in patients with good runoff, patency rates were comparable to that of nondiabetic patients.¹⁵⁵ For infrainguinal ischemia, the outcomes of surgical revascularization in diabetes resemble those in patients without diabetes in terms of limb salvage,¹⁵⁶ albeit more distal than that of nondiabetic patients.¹⁵⁷ In general, in patients with severe claudication or critical limb ischemia, surgery is probably superior to percutaneous transluminal angioplasty for revascularization in the femoral, popliteal, and infrapopliteal vessels, but this comes at a price of increased periprocedural cardiovascular morbidity and mortality.¹⁴⁴

Diabetic patients with cerebrovascular atherosclerosis should also receive platelet antagonists, statins, and ACE inhibitors. A strategy of surgical revascularization with medical therapy for asymptomatic and symptomatic patients with hemodynamically significant internal carotid artery atherosclerosis resulted in fewer strokes then medical therapy alone.^158- 159 Diabetic subjects represented 23% of the total population in the Asymptomatic Carotid Atherosclerosis Study (ACAS) and 19% in the North American Symptomatic Carotid Endarterectomy Trial (NASCET) and therefore should receive treatment similar to that of nondiabetic patients with carotid artery disease. Several studies have demonstrated increased cardiovascular mortality following carotid endarterectomy in patients with diabetes, both at 30 days and 1 year.¹⁶⁰ The mortality derives from an increased rate of coronary heart disease events.¹⁶¹ The rates of perioperative major and minor stroke do not differ significantly between diabetic and nondiabetic patients¹⁶²; however, hospital length of stay tends to increase.¹⁶³ Despite the increased use of stenting for the treatment of carotid artery atherosclerosis, we lack direct outcome data in diabetic patients, but evidence suggests that outcomes may be similar to those of patients without diabetes.¹⁶⁴

Atherosclerosis causes most of the death and much of the disability in patients with diabetes. Dysresgulation of metabolism in diabetes adversely affects every cellular element within the vascular wall. Intensive treatment of the gamut of metabolic abnormalities associated with diabetes beyond hyperglycemia reduces the rates of MI and death. As this high-risk population increases, strategies to encourage the aggressive use of targeted medical and revascularization therapies will reduce the magnified rates of death and disability.