There has been an epidemic of type 2 diabetes in the United States and worldwide. This graphic was made up a couple years ago; we have now 18 million Americans with diabetes, more than 90% of these have type 2 diabetes, and the problem is going to get worse.

We know that the prevalence of diabetes increases as a function of age. These data show that over the age of 65 years, almost 18% of people have type 2 diabetes. Even worse, 40.8% of people have impaired glucose tolerance (IGT).

Our population is aging. In 1986, 11% were over the age of 65 years. By the year 2020, that will double. So all things being equal, the problem of type 2 diabetes is going to double over the years.

The direct and indirect costs of diabetes are overwhelming. One in every 7 healthcare dollars is spent on someone with diabetes. The annual healthcare costs are 3-fold greater for people with diabetes than people without diabetes. Direct costs are about $85 billion. Total costs are over $100 billion.

Where does all this money go? Not for routine follow-up visits, not for acute metabolic problems, but for the most part, it goes for the treatment of the chronic complications of diabetes.

Microvascular complications include retinopathy, neuropathy, and nephropathy. Then macrovascular complications include cerebrovascular disease, peripheral vascular disease, and coronary heart disease. Mainly, the macrovascular complications are the big problem in people with type 2 diabetes.

These data show that ischemic heart disease, other heart disease, and stroke are the overwhelming causes of death in people with type 2 diabetes. For any given risk factor, the chances of getting cardiovascular disease are markedly increased in people with type 2 diabetes.

These are data from the Multiple Risk Factor Intervention Trial (MRFIT) study, where people with and without diabetes were classified as having: no risk factors at all, only 1 risk factor, 2 risk factors, or all 3 risk factors. Risk factors were hypertension, hyperlipidemia, and smoking. For any given number of risk factors, the chances of getting cardiovascular disease are markedly increased in people with type 2 diabetes. This increased risk is related to hyperglycemia.

There are several lines of evidence linking hyperglycemia to macrovascular disease. The first is epidemiologic data. Numerous studies have shown correlations between the frequency of cardiovascular disease and either plasma glucose levels or hemoglobin A1C (HbA1C).

These data from a Finnish study compare people with HbA1C levels below 6% with those who have HbA1C levels between 6% and 7.9%. They have roughly a doubling of the incidence of cardiovascular mortality.

Is it fasting or postprandial hyperglycemia that is important? Hemoglobin A1C measures total exposure to hyperglycemia over about a 3-month period of time. Both fasting and postprandial hyperglycemia contribute to this. We have no evidence that there's anything more toxic for postprandial hyperglycemia vs fasting hyperglycemia. The relative contributions depend on the relative degree of glycemic control. When your HbA1C is very high, when you have a fasting glucose level over 200 mg/dL, most of the HbA1C will be due to fasting hyperglycemia. However, earlier in the stage of diabetes, when HbA1C levels are lower, it's going to be the postprandial values that contribute most to HbA1C.

This shows the relationship of fasting and postprandial hyperglycemia to overall HbA1C over the range from less than 5% to 7.5%. Over this range, which covers the goals recommended by various groups, there is a very small increase in fasting plasma glucose levels, and a greater increase in postprandial levels. People who have HbA1C levels of 5.5% to 6% may have normal fasting plasma glucose levels, but may have a greater increase in postprandial hyperglycemia levels, and therefore these values contribute to high HbA1C.

Is postprandial hyperglycemia a risk factor by itself if you have normal fasting plasma glucose levels? The answer is yes, we know that from epidemiologic data, but we also have this from controlled clinical trials. Let me show some of the epidemiologic data.

These are data from the Paris Prospective Study. They looked at people with nondiabetic fasting plasma glucose levels of less than 140 mg/dL, and people with IGT, meaning normal fasting glucose levels and elevated 2-hour postprandial values. And the cardiovascular mortality is increased almost 2-fold.

This is the Funagata Diabetes Study. They looked at cardiovascular survival by whether someone had normal glucose tolerance, IGT, or type 2 diabetes. Compared with people with normal glucose tolerance, there was a marked reduction in cardiovascular survival in people with type 2 diabetes over 7 years. But for people with IGT, after 7 years, the survival is virtually the same in both groups.

Even scarier are the data from the Norfolk, United Kingdom cohort of the European Prospective Investigation into Cancer (EPIC-Norfolk) study. This was a study of about 5000 people in various centers in Europe. They simply assessed cardiovascular risk factors: hypertension, hyperlipidemia, smoking, preexisting cardiovascular disease, and HbA1C. Then they looked 3 or 4 years later to see who was alive, who was dead, and what they died from. And they used, as their control group, people with HbA1C levels of 5% or below; they were given a relative risk of 1. In people who had HbA1C levels from 5% to 5.4%, there was roughly a 2.5-fold increased risk of dying from cardiovascular disease.

As strange as it may seem, in this study, where they measured all these cardiovascular risk factors and HbA1C, they did not measure plasma glucose levels. So who are these people between 5% and 5.4% HbA1C? We accessed our database that had almost 500 people in it who had oral glucose tolerance tests and HbA1C levels. In the group with HbA1C of 5% or below, about 85% had normal glucose tolerance.

In the group between 5% and 5.5%, who had 2.5-fold increased mortality in the EPIC-Norfolk study, about 37% either had IGT or type 2 diabetes. This could possibly explain the increased cardiovascular risk.

This would be consistent with what was found in the United Kingdom Prospective Diabetes Study (UKPDS); at the time of diagnosis, people had macrovascular complications. Seventeen percent had cardiovascular complications by the time type 2 diabetes was diagnosed. You could argue they might have had diabetes for several years, but some of this cardiovascular morbidity may be related to having had IGT for several years.

How can glucose contribute to cardiovascular disease?

We have evidence from in vitro experiments to show that there are many biochemical mechanisms to explain how hyperglycemia may lead to atherogenesis. With hyperglycemia you get glycosylation of proteins. You have more glucose moving through the polyol-hexosamine pathways. You have the generation of free radicals. All of these things can lead to an increase in protein kinase C and diacylglycerol.

These free radicals, these compounds, and glycation can: lead to increases in plasminogen activator inhibitor-1 (PAI-1), therefore decreasing fibrinolysis; increase the production of matrix proteins in the atheroma; reduce nitric oxide (NO) production, a vasodilator that keeps the endothelium healthy; increase local generation of endothelin-1 (ET-1), which can predispose to hypertension; and increase the production of growth factors within the endothelium and subendothelium that can lead to atheroma.

All of these things together, as the product of hyperglycemia, can lead to cell proliferation, increased coagulability, decreased fibrinolysis, decreased vasodilation, increased vascular permeability, and increased matrix formation, all known to be involved in the process of atherogenesis.

Can we prevent these micro- and macrovascular complications?

The answer is definitely yes. There have been controlled clinical trials showing that if you optimally control blood pressure, you can reduce both micro- and macrovascular complications. You can reduce macrovascular complications by optimizing control of lipids. And we've had controlled clinical trials showing that strict glycemic control can prevent these complications.

The first big study was the Diabetes Control and Complications Trial (DCCT), done in people with type 1 diabetes. They were randomized to an intensive insulin regimen or, what was at the time, conventional insulin therapy. And there was about a difference of 1.9% in HbA1C levels, 7% vs 8.9%.

The other big study was the UKPDS. This was a study of over 4000 newly diagnosed patients with type 2 diabetes who were randomized to conventional or strict metabolic control using various oral hypoglycemic agents. The difference between the intensively treated group and the conventionally treated group was only .9% HbA1C. A relatively small difference, but what a big difference it made in complications.

In the DCCT, tight glycemic control decreased retinopathy by 63%, nephropathy by 39%, and neuropathy by 60%.

This is an extrapolation of the data from the DCCT to show the risk of developing these complications as a function of HbA1C. When you get HbA1C down to about 7%, the relative risk compared with the normal population is about 1. This epidemiologic relationship between HbA1C and microvascular complications is the basis of the American Diabetes Association (ADA) recommending a target HbA1C of 7% or less.

In the UKPDS there was a reduction in all diabetes-related endpoints and microvascular complications. The reduction in myocardial infarction (MI), 16%, was P = .052 when the data were originally reported. Many people like myself said, "P = .052, is as good as P < .05. That's significant." You had purists who said, "It was above .05, not statistically significant." There was an evaluation of the data with the final HbA1C levels, and that was statistically significant.

These are data from that analysis showing the relationship between the updated HbA1C levels and fatal and nonfatal MI. For every 1% reduction in HbA1C, there was a 14% decrease in MI. These studies showed that you needed better glycemic control to prevent macrovascular disease than microvascular complications.

I mentioned earlier about IGT as a risk factor for cardiovascular disease. There was a study called the Study to Prevent Non-Insulin-Dependent Diabetes Mellitus (STOP-NIDDM). It was designed to see if the use of an alpha-glucosidase inhibitor, which only works on postprandial hyperglycemia, could prevent people with IGT from progressing to overt type 2 diabetes.

They also looked at the development of cardiovascular disease during the 3-year study. The progression to type 2 diabetes was reduced, but also the development of cardiovascular events was reduced by 50%. This was just by treating postprandial hyperglycemia. We know that once people have cardiovascular complications, good control matters.

The Diabetes Mellitus Insulin Glucose Infusion in Acute Myocardial Infarction (DIGAMI) study took patients in Sweden who were in coronary care units with MIs, and randomized them to an intensive basal-bolus insulin regimen or their conventional treatment. The people who were placed on the basal-bolus intensive regimen had in-hospital reduction of cardiovascular mortality, and it was reduced more and more over time. At 5 years, the relative reduction in cardiovascular mortality was increased. So even in established macrovascular disease, good glycemic control matters.

Perhaps one of the mechanisms is the improvement in cardiovascular risk factors. These are data from 4 different studies that looked at the effect of improving glycemic control in people with type 2 diabetes by use of insulin regimens on various cardiovascular risk factors. There's a reduction in triglycerides, low-density lipoprotein (LDL) cholesterol, and very low density lipoprotein (VLDL), an increase in high-density lipoprotein (HDL), a reduction in apolipoprotein B, a reduction in blood pressure, a reduction in insulin resistance, a reduction in albuminuria, an improvement in endothelial function, a decrease in C-reactive protein (CRP), a decrease in PAI-1, and a decrease in thromboxane generation.

The goals of glycemic control. The ADA, primarily based on evidence regarding microvascular complications, recommends HbA1C below 7%. The International Diabetes Federation (IDF) and the American Association of Clinical Endocrinologists (AACE), taking into consideration that you need stricter glycemic control to prevent macrovascular complications, recommend a target of less than 6.5%.

What do we have to do to get to an HbA1C of 6.5%? Going back to that database that I mentioned earlier, we accessed people who had HbA1C levels between 6% and 6.5% and compared them with people with levels between 6.5% and 7%. All of these people had diabetes, but were under no treatment other than diet at the time.

What's the difference between these 2 groups? Fasting plasma glucose levels are near normal. They're not different between the 2 groups. Postprandial hyperglycemia, there's the difference. So to take a patient with HbA1C between 7% and 7.5% and move them below 7%, to 6.5%, it's going to take concentrating on treating postprandial hyperglycemia because people in those HbA1C ranges have normal fasting glucose levels.

How good are we at getting good glycemic control? These are data from National Health and Nutrition Examination Survey III (NHANES III), and only 38% of people on oral agents achieved HbA1C below 7%. It's even worse with insulin, 27%. This shows that we're not using enough of the pills, or people who should be on insulin are still on pills. Looking at insulin, you have to draw the conclusion that either insulin doesn't work or we're not using it aggressively. We know from studies such as the DCCT that insulin does work.

NHANES III ended in 1994. At that time we were basically limited to sulfonylurea treatment and conventional insulins: regular insulin, neutral protamine Hagedorn (NPH), and mixtures. Since that time we have had the introduction of numerous new agents, many of which work via different mechanisms, many of which are better than our old preparations. And with these new tools we should be able to get better glycemic control than we have in the past.