Is treatment of insulin resistance beneficial independent of glycemia?

The clinical definition of the Insulin Resistance or Metabolic Syndrome is shown in Table 1. Subjects who fulfill these criteria are at increased risk for cardiovascular disease (CVD). The criteria of the National Cholesterol Education Program are more clinically relevant because those measurements are readily available. Although those with impaired fasting glucose and/or impaired glucose tolerance (IGT) are at increased risk for the subsequent development of type 2 diabetes (3), abnormal glycemia is not an independent predictor of cardiovascular disease once the other risk factors are taken into account (4,5). Three recent randomized, controlled studies have shown that lifestyle intervention with diet and exercise can decrease the development of type 2 diabetes in subjects with IGT (6–8). However, although exercise is associated with less CVD (9), the hypothesis that exercise decreases CVD has never been directly tested in a randomized, controlled clinical trial. Insulin resistance, highly associated with visceral fat accumulation (10), characterizes the Insulin Resistance or Metabolic Syndrome. Furthermore, measured insulin resistance predicts CVD (11). In this issue of Diabetes Care, Shadid and Jenson (12) compare the effects of diet and exercise versus pioglitazone in insulin-resistant nondiabetic men and premenopausal women with upper body obesity (i.e. visceral fat accumulation). Insulin sensitivity increased in both groups. Those given diet and exercise showed weight loss and decreased visceral, subcutaneous, total body, and leg fat, whereas the subjects prescribed pioglitazone had no change in subcutaneous and visceral fat and increased their weight and total body and leg fat. Although the differences in fat and weight changes suggest different mechanisms, both approaches decreased insulin resistance. While no one would argue with using diet and exercise to reduce insulin resistance, it is well recognized that the long-term effectiveness of this lifestyle intervention is disappointing. This raises the question of whether pharmacological means to increase insulin sensitivity would be beneficial and whether they should be considered. Similar to the situation with lifestyle intervention, there are no studies currently available demonstrating a beneficial effect on clinical cardiovascular end points after reducing insulin resistance by pharmacological means (although such trials are in process). However, there are a number of reports of beneficial effects on risk factors for and surrogate end points of CVD. All three glitazones raise HDL cholesterol levels (13–15) and change small, dense, more atherogenic LDL particles to larger, less atherogenic ones (16– 18). Their effect on triglycerides is variable, with pioglitazone the most effective at lowering them and rosiglitazone the least effective (13–15). The blood pressure–lowering effect of rosiglitazone is proportional to the drug’s effect on decreasing insulin resistance (19). In addition to the classical risk factors of dyslipidemia and hypertension for CVD, accumulating evidence has also implicated prothrombotic and proinflammatory states, increased levels of certain endothelial cell molecules, and endothelial dysfunction in the pathogenesis of CVD. Increased plasminogen activator inhibitor-1 levels, leading to impaired fibrinolysis and therefore a prothrombotic state, are part of the Insulin Resistance Syndrome and were lowered by troglitazone (20–22) and rosiglitazone (19). Creactive protein, a marker of the proinflammatory state, was also reduced by troglitazone (21,22), rosiglitazone (19,23), and pioglitazone (24), the latter having this effect even in patients who did not respond glycemically to the drug. Early events in the pathogenesis of atherosclerosis involve the effect of oxidized and glycated LDL initiating the formation of fat-filled macrophages or foam cells. This process involves triggering inflammatory responses by increased expressions of cytokines and adhesion molecules, e.g., intracellular adhesion molecule (ICAM) and E-selectin, and the stimulation of monocyte migration to the vessel wall by chemotactic factors, e.g., MCP (25). Troglitazone lowered Eselectin (26), ICAM (20), and MCP-1 (20) and rosiglitazone decreased E-selectin (27). One of the matrix-degrading metalloproteinases (MMP-9), which has been implicated in atherosclerotic plaque rupture, the precipitating event in the acute coronary syndrome, was decreased in type 2 diabetic patients with CVD by troglitazone (22) and rosiglitazone (28). Reducing insulin resistance also improved the morphologic, physiological, and clinical outcomes of CVD in type 2 diabetic patients. Morphologically, intima-media thickness (IMT) of the carotid arteries correlates with the extent of coronary artery atherosclerosis (29). Both troglitazone (30) and pioglitazone (31) decreased IMT within 3 months. Physiologically, endothelial dysfunction, which also characterizes atherosclerotic disease in type 2 diabetic patients, was improved by troglitazone (32) and rosiglitazone but not by metformin (33). Furthermore, in type 2 diabetic patients, pulse wave velocity, a direct measure of arterial distensibility, which correlates well with IMT (34), is a good marker of vascular damage (35), and predicts mortality (36), was decreased by pioglitazone (24), even in patients who did not respond glycemically to the drug. From a clinical perspective, rosiglitazone improved myocardial blood flow measured by positron emission tomography scanning in type 2 diabetic patients with disease duration 8 years (37). Both the effects of troglitazone on endothelial dysfunction (32) and rosiglitazone on myocardial blood flow (37) were not significant in type 2 diabetic patients with longer durations of disease, suggesting that lowering insulin resistance is more likely to prevent or delay CVD if it occurs early on. Finally, in type 2 diabetic patients in whom stents were placed in their coronary arteries, the restenosis rate was markedly reduced by rosiglitazone compared with the control group in the absence of a glycemic difference (38). Similarly, the minimal lumen diameter was significantly greater 6 E D I T O R I A L ( S E E S H A D I D A N D J E N S E N , P . 3 1 4 8 )