Recent Advances in the Diagnosis, Pathogenesis, and Treatment of Type 2 Diabetes

Authors: Niyaz R. Gosmanov, MD, Steven D. Wittlin, MD, and John E. Gerich, MD, University of Rochester School of Medicine, Department of Medicine, Rochester, NY.

Editor’s Note—During the last decade, there has been a remarkable advance in our understanding of the pathogenesis of type 2 diabetes and the relationship between the control of glycemia and the development of the complications of diabetes. In addition, there has been the introduction of many new treatment modalities, which has enabled us to obtain better glycemic control in a larger percentage of our patients with less side effects.

In contrast to the studies of the 1980s, studies during the 1990s (clearly illustrated by studies of monozygotic twins) indicate that impaired beta cell function (inappropriate insulin release relative to insulin resistance) is the primary genetic factor involved in type 2 diabetes and that the insulin resistance associated with this disorder is largely explained by obesity, decreased physical activity, high fat diets, and secondary effects of high plasma glucose levels (glucose toxicity) and high plasma free fatty acid levels (lipotoxicity).

Nevertheless, although insulin resistance seems to be a largely acquired problem, it is important and its treatment, along with that of impaired insulin secretion, are vital parts of the therapeutic approach to the condition. Indeed, the argument can be made that both should be targeted from the start by use of combination therapy.

Several clinical trials (the Diabetes Control and Complications Trial [DCCT] and the Stockholm Intervention Trial in type 1 diabetes mellitus [T1DM], the Kumomoto Study, and the United Kingdom Prospective Diabetes Study [UKPDS] Trial in type 2 diabetes mellitus [T2DM]) have shown that maintenance of HbA1c levels below 7.0% can prevent development of microvascular complications. Epidemiological studies indicate that reduction of macrovascular complications may require maintenance of even lower HbA1c levels. For both types of complications the risk decreases with progressive lowering of HbA1c and shows no threshold.

The introduction of insulin sensitizers (metformin and thiazolidinediones) along with new insulin secretogogues (the sulfonylurea glimepiride and the meglitinides-repaglinide and nateglinide) along with more rapid, shorter-acting insulin analogs (lispro and aspart insulin) and the peakless long-acting analog (glargine insulin) now permit better glycemic control to be achieved with less risk of hypoglycemia. Various combinations of these agents have been shown capable of reducing HbA1c levels to less than 7.0%.

In summary, less than half of patients with diabetes in the United States have HbA1c levels less than 7.0%. We now have the tools to correct this and prevent development of diabetic complications in a safe, cost-effective manner.


During the past 10 years, there have been remarkable advances in our knowledge of the pathogenesis of T2DM as well as the relationship between glycemic control and the development of the complications of diabetes. In addition, several new treatment modalities have become available, which have improved our ability to obtain satisfactory glycemic control in a greater proportion of patients. This review will summarize these advances.

Glycemic Control and Complications

The relationship between glycemic control and the development of diabetic complications was controversial. This issue was put to rest by the results of several controlled clinical trials during the last decade. The first to be reported was the DCCT1—a study of intensive vs. conventional insulin therapy in patients with T1DM. Patients in the intensive group were managed by basal-bolus regimens of 3-4 insulin injections or pumps per day and used self-glucose monitoring to make dose adjustments; their HbA1c levels over the course of the study averaged about 7.0% compared to 9.0% in the conventional group, which took only 2 insulin injections a day and was not instructed how to adjust its doses based on self-glucose monitoring. This 2.0% difference in HbA1c reduced the development of retinopathy, nephropathy, and neuropathy by 63%, 39%, and 60%, respectively. Furthermore, the progression of microvascular complications already present was reduced by nearly 60%. Similar findings in patients with T1DM were reported in the Stockholm Diabetes Intervention Study.2

Most diabetologists considered that the results of these studies could be extrapolated to patients with T2DM since the adverse effects of hyperglycemia should be the same in both types of diabetes. However, there were still many skeptics. Their doubts were put to rest by publication of the results of the Kumomoto Study3 and the UKPDS4-6 in 1995 and 1998, respectively. The Kumomoto Study7 was virtually identical in design to the DCCT and the results were similar. Patients with T2DM randomized to intensive insulin treatment averaged a HbA1c of 7.0% during the 6-year study while those on conventional insulin therapy averaged about 9.5%. Intensive insulin therapy reduced the development of retinopathy and nephropathy by 76% and 62%, respectively, and the progression of these complications was reduced by about 60%.

The UKPDS4-6 was by far the most ambitious trial. It was started in 1978 as a response to the University Group Diabetes Program,8 which had failed to demonstrate a beneficial effect of glycemic control and had suggested that the sulfonylurea tolbutamide might be cardiotoxic. The UKPDS randomized more than 5000 patients with newly diagnosed T2DM to conventional or intensive therapy. The latter included initial monotherapy with sulfonylureas (glyburide or chlorpropamide) or insulin in nonobese patients and metformin or insulin in obese patients. Therapy was further intensified by either use of combinations of oral agents or conversion to progressively more intensive insulin regimens if goals were not being achieved. The goal of intensive therapy was a HbA1c below 7.0%. The study had an average follow-up of about 15 years. The purposes of the study were to determine whether good glycemic control affected not only microvascular but also macrovascular complications and whether any particular drug was advantageous. Many lessons were learned from this study.

The first lesson was that good glycemic control prevents both micro- and macrovascular complications. The differences in HbA1c between the intensively and conventionally treated groups were modest (ie, 0.9% in the nonobese group and 0.6% in the obese group—roughly 7.0% vs. 7.6-7.9%). Despite this modest difference in glycemic control, the results were dramatic. In the nonobese group, any diabetes-related end point was reduced 12%; microvascular end points were reduced by 25%; myocardial infarction was reduced by 16%; retinopathy was reduced by 21%; and nephropathy was reduced by 33%. In the obese group, any diabetes-related end point was reduced by 32%; myocardial infarction was reduced by 39%; and all-cause mortality was reduced by 36%.

The second lesson was that these regimens were safe: there was no evidence that sulfonylureas or insulin had adverse macrovascular effects (this had been a continuing concern); metformin, when used appropriately, did not cause lactic acidosis (no cases occurred); and severe hypoglycemia with sulfonylureas (~ 0.5%) and insulin treatment (~ 2%) was relatively uncommon.

From an epidemiological point of view, the results of lowering HbA1c in the UKPDS (T2DM) closely coincided with those of the DCCT (T2DM). In fact, data from all of the clinical trials indicate that if the HbA1c could be maintained below 7.0% in patients either T1 or T2DM, then risk of developing microvascular complications would approach that of the general population. The situation, however, appears to be different for prevention of macrovascular complications, which is the major cause of morbidity and mortality in T2DM.9

Numerous studies9-19 have shown that people with impaired glucose tolerance, whose HbA1c would be expected to be below 7.0%, have an increased risk for cardiovascular disease. The results of the recent EPIC-NORFOLK study20 illustrate this. In this prospective study, subjects were categorized by their initial HbA1c level and those with levels 5.0% or below were contrasted with those with higher levels. It was found that people with HbA1c levels between 5.0 and 5.5 had a 40% increase in risk for cardiovascular death in follow-up. These results are consistent with the results of the Honolulu Study10 and several others.16 Therefore, it appears that if we wish to prevent both macro- and microvascular complications of diabetes, we must try to achieve HbA1c levels lower than those currently recommended by the American Diabetes Association (ADA).

The current ADA recommendations consider a HbA1c < 7.0% as a goal and urge action be taken to improve control if the HbA1c is above 8.0%. These recommendations imply that HbA1c levels between 7.0% and 8.0% constitute adequate control and are not consistent with scientific data regarding the relation between glycemic control and macrovascular complications discussed above. These data suggest, at a minimum, the goal for therapy should be a HbA1c less than 6.0% and that action to improve control must be initiated if the HbA1c is above 7.0%. For several years, The European Diabetes Group consensus panel21 has recommended as a goal a HbA1c < 6.5% and intensification of glycemic control if the HbA1c exceeds 7.5%. Similar guidelines have recently been proposed by the American Academy of Clinical Endocrinology (

It is appreciated that these goals may not be appropriate for all patients (eg, elderly patients or those already having severe complications) and may be difficult to achieve by those involved in primary care and general internal medicine practices who lack time and ancillary resources (ie, dietitians, nurse educators, etc). Therefore, when patients with T2DM cannot achieve an HbA1c of 7.0% or below on maximal doses of 2 different classes of oral agents alone or in combination with a simple insulin regimen, referral to a specialist should be seriously considered.

Pathogenesis of Type 2 Diabetes

The pathogenesis of T2DM is much more complicated than T1DM, which is simply due to autoimmune destruction of pancreatic islet beta cells making insulin. T2DM is a heterogeneous disorder. About 5-10% of patients appearing to have T2DM actually have a late onset variety of T1DM; these patients are generally lean and progress to insulin dependency within a few years.22 Another 5% have Maturity Onset Diabetes of Youth (MODY),22 an autosomal dominant disorder of which at least 5 subtypes exist. Patients with this disorder generally only have impaired insulin release and develop their diabetes younger than the age of 25. The remaining 85-90% of patients have classic T2DM in which both insulin resistance and impaired insulin secretion are involved.22 Each of these factors is affected by genetic and acquired/environmental influences.22 Furthermore, it is now apparent that T2DM is a polygenic disorder, meaning that it takes the interaction of several diabetogenic genes with acquired/environmental factors to develop/acquire T2DM.22

Until a few years ago, it was generally thought that insulin resistance was the primary genetic component. Recent studies, however, have challenged this concept and it now seems most likely that the main genetic component involves pancreatic beta cell dysfunction.

The evidence for this view can be summarized as follows:

1. Most of the insulin resistance found in T2DM can be attributed to nondiabetes-specific factors such as obesity (especially visceral obesity),22 physical inactivity,23 high-fat diets,24 glucose,25 and lipotoxicity.26 To some extent, obesity and fat distribution are genetically determined,22 but such genes are not diabetes specific.

2. Although most (> 90%) patients with classic T2DM are obese (and therefore insulin resistant), most insulin resistant obese individuals are not diabetic. What distinguishes obese individuals with and without diabetes is the ability to compensate for insulin resistance with increased insulin secretion.27

A clear example of this is gestational diabetes.28 All women become insulin resistant during the third trimester of pregnancy. Some women develop diabetes during this period. What distinguishes women who do and do not develop diabetes are their plasma insulin levels, which are higher in women who do not develop gestational diabetes. After pregnancy, women who have gestational diabetes return to normal glucose tolerance. Nevertheless, gestational diabetes is a strong risk factor for subsequent development of T2DM, which approaches 50%.28

When women who had developed gestational diabetes have been studied after return to normal glucose tolerance, an impairment in beta cell function has been detected while insulin sensitivity has been found to be normal.29 This not only illustrates that, as discussed below, impaired beta cell function precedes insulin resistance, but also the importance of insulin resistance, even if not genetic, in the pathogenesis of T2DM since with resolution of pregnancy-associated insulin resistance, glucose tolerance returns to normal.

3. A third line of evidence is that impaired beta cell function can be detected earlier in people with normal glucose tolerance who are genetically predisposed to develop T2DM (eg, first-degree relatives of individuals with T2DM).30 The strongest evidence for this comes from studies of monozygotic twins, in which one twin has T2DM while the other still has normal glucose tolerance.31 The twin with normal glucose tolerance has about an 80% chance of developing T2DM and thus can be considered to be a true prediabetic individual. All 4 studies of such twin pairs have found that the twin with normal glucose tolerance had impaired beta cell function22 and the only study that simultaneously assessed insulin sensitivity found it to be normal.31

4. Another line of evidence comes from examining the reversibility of insulin resistance and impaired beta cell function in obese individuals with T2DM. Several studies22 have demonstrated complete normalization of insulin sensitivity after weight loss but persistence of impaired beta cell function. The marked improvement in glycemic control after weight loss and normalization of insulin sensitivity illustrates again the importance of insulin resistance, but the fact that patients remained diabetic also points out the essentiality of impaired beta function.

Mechanisms for Impaired Beta Cell Function and Insulin Resistance

The mechanisms responsible for impaired beta cell function are not known. Some studies have suggested that a reduced beta cell mass may be involved. The UKPDS32 demonstrated that beta cell function was already reduced by 50% in patients with newly diagnosed T2DM and that there was subsequent progressive deterioration. Other factors may involve amyloid deposition, alterations in growth factors necessary for establishment maintenance of normal beta cell function/mass, and toxic effects of increased exposure to glucose and free fatty acids (FFA).27

Obesity and Type 2 Diabetes

The insulin resistance of obesity has been shown to be associated with reduced numbers of insulin receptors, reduced insulin receptor kinase activity, reduced activation of insulin signaling proteins, and glucose transport.33 Quantitatively similar defects have been found in obese individuals with T2DM.33 Several factors may be involved in the insulin resistance associated with obesity: higher circulating levels of 1) free fatty acids; 2) tumor necrosis factor alpha;34 and 3) the newly discovered hormone resistin,35 all of which are released from adipose tissue. Reduced physical activity, high-fat diets, increased plasma glucose levels, and certain drugs (eg, nicotinic acid, glucocorticoids, diuretics) may also contribute to the insulin resistance found in patients with T2DM.

Metabolic Consequences of Insulin Resistance and Impaired Beta Cell Function

As a consequence of insulin resistance and impaired beta cell function, there is overproduction of glucose by liver and kidney in the fasting state and impaired suppression of hepatic glucose release after meal ingestion.36,37 Although glucose uptake by peripheral tissues is normal in an absolute sense, it is not appropriate for the prevailing hyperglycemia and plasma insulin concentrations37 (ie, glucose uptake is less efficient). Circulating FFA levels are generally increased in both the fasting and postprandial states; these promote greater release of glucose26 and impede tissue glucose uptake.26 The failure to "turn off" release of glucose after meal ingestion has been specifically linked to impaired early insulin release.38 This exacerbates postprandial hyperglycemia, which can result in late hyperinsulinemia. These insulin levels, however, are still not appropriate for the prevailing hyperglycemia.

Revised Diagnostic Criteria for Diabetes

In 1997, a consensus panel of the ADA recommended changes in the Diagnostic Criteria for Diabetes.39 The old criteria were a fasting plasma glucose above 140 mg/dL or a value above 200 mg/dL at 2 hours of a 75 g oral glucose tolerance test (OGTT). Two-hour values above 140 mg/dL but below 200 mg/dL categorized a person as having impaired glucose tolerance. The new criteria were established mainly because review of epidemiologic data indicated that a fasting criterion value of 140 mg/dL was too high. Many people with a 2-hour value greater than 200 mg/dL during the OGTT had a fasting plasma glucose level considerably below 140 mg/dL. Examination of the data indicated that a fasting value above 126 mg/dL was more appropriate because most people with values above 126 mg/dL would also have 2-hour values greater than 200 mg/dL. A fasting value of 110 mg/dL or below was deemed normal since most of these individuals had normal values during OGTTs, whereas values between 110 and 126 mg/dL categorized a person as having impaired fasting glucose tolerance, presumably carrying the same implications of the category of impaired glucose tolerance (ie, increased risks to develop diabetes and coronary heart disease).

This change in diagnostic criteria also came with the recommendation to rely more on fasting plasma glucose levels while discouraging use of the OGTT. It was argued that fasting plasma glucose levels were more reproducible than 2-hour values of the OGTT and that because fasting values would be more convenient, this would encourage more testing.

The consensus panel concluded that these changes would not increase the actual number of people with diabetes, but rather would diminish the number of people with undiagnosed diabetes. In other words, more people will be diagnosed earlier and receive appropriate treatment earlier, which should reduce the incidence of diabetic complications. In the UKPDS, more than 25% of patients with newly diagnosed diabetes already had diabetes-related complications.40

Another implication of the new criteria is that a greater proportion of the HbA1c level of the patients diagnosed earlier would be determined by postprandial hyperglycemia. Treatment of patients with fasting plasma glucose levels between 126 mg/dL and 140 mg/dL with potent long-acting sulfonylureas may increase the risk for hypoglycemia. In such patients, use of the short-acting meglitinide insulin secretogogues, which primarily target postprandial hyperglycemia, may have advantages.

New Treatment Modalities

Until 1992, the only pharmacologic treatments available in the United States for T2DM were sulfonylureas and insulin. Since that time, a new so-called third-generation sulfonylurea (glimepiride); nonsulfonylurea secretogogues (the so-called meglitinides, repaglinide, and nateglinide); metformin—a biguanide on the market in Europe for 40 years; alpha glucosidase inhibitors (acarbose and miglitol); and insulin analogs either with a more rapid onset and shorter duration of action (lispro and aspart insulin) or with a longer duration than neutral protamine Hagedorn (NPH) and lack of a peak (insulin glargine) have become available. Some of the basic clinical pharmacology of the new oral agents are summarized in Table 1.

Table 1. Efficacy and Side Effects of Commonly Used Oral Antidiabetic Drugs
Class Dose Efficacy
in HbA1c)
Common Side Effects of Class

1. Sulfonylureas Hypoglycemia
a. Glyburide (Diabeta®, Glynase®)
b. Glipizide (Glucotrol®)
c. Glimepiride (Amaryl®)
1.25-10 mg
5-20 mg
2-8 mg/d
Weight gain
2. Meglitinides Hypoglycemia
a. Repaglinide (Prandin®)
b. Nateglinide (Starlix®)
1-4 mg premeal
120 mg premeal
Weight gain
1. Metformin (Glucophage®) 500-2500 mg/d 1.5-2.0% Anorexia; diarrhea; dyspepsia; lactic acidosis
2. Thiazolidinediones Fluid retention; 
weight gain; possible precipitant of CHF
a. Rosiglitazone (Avandia®)
b. Pioglitazone (Actos®)
2-8 mg/d
15-45 mg/d
Alpha Glucosidase Inhibitors Flatulence; diarrhea
1. Acarbose (Precose®)
2. Miglitol (Glyset
25-100 mg t.i.d.
25-100 mg t.i.d.

Insulin Secretogogues

Sulfonylureas and meglitinides (nateglinide and repaglinide) act by promoting insulin secretion.41 With improvement in glycemic control, there will also be a modest improvement in insulin sensitivity due to alleviation of glucose toxicity.25,42 Both groups (sulfonylureas and meglitinides) bind to receptors on pancreatic beta cells. This leads to closure of ATP-sensitive potassium channels, which causes depolarization of the cell membrane and an influx of calcium through voltage-sensitive channels. The resultant increase in intracellular calcium stimulates the release of insulin. Meglitinides and sulfonylureas differ in their binding sites and binding kinetics. This leads to differences in their onset and duration of action as well as the different characteristics of the response in plasma insulin levels. Sulfonylureas are slower in onset, have a longer duration of action (up to 24 hours), and primarily affect second-phase insulin release whereas meglitinides have a rapid onset (< 15 minutes), relatively short duration (~ 3-4 hours), and primarily affect first phase insulin release.41,43

Glimepiride has a similar efficacy to the other commonly used sulfonylureas (glyburide and glipizide), lowering HbA1c on average 1.5-2.0%.44 It has a duration of action comparable to glipizide XL (ie, ~ 24 hours and thus is administered once a day in doses ranging from 2-8 mg/d).44 Compared to glyburide, it has been reported to cause less hypoglycemia and less weight gain and not to impair cardiac preconditioning; compared to glyburide, it is more selective in its binding to islet receptors and those in cardiac muscle.45-47 The clinical significance of the latter finding is unclear.

Nateglinide (120 mg) and repaglinide (1-4 mg) are given just prior to meal ingestion (ie, 3-4 times/d).41,48,49 With nateglinide, in contrast to repaglinide, no dose titration is needed; the starting dose of 120 mg per dose is also the maintenance dose. These agents have only a modest effect on fasting plasma glucose levels but are particularly effective in reducing postprandial hyperglycemia43,50 and have the same side effects as sulfonylureas. However, their short duration of action has been reported to lead to less hypoglycemia and weight gain than sulfonylureas.51,52 There have been no head-to-head comparisons of nateglinide and repaglinide in clinical trials in T2DM. Repaglinide has been reported to lower HbA1c to a comparable extent to glyburide50,53 and glipizide.54 In short-term studies (8 weeks), nateglinide (120 mg/d) was reported to reduce overall glycemia less than glyburide (5 mg b.i.d.).43 Nateglinide has also been reported to lower HbA1c and fasting plasma glucose less than metformin, but to reduce postprandial plasma glucose excursions more.48,55 No long-term comparison of monotherapy with repaglinide and metformin is available; however, repaglinide has been reported to be more effective in lowering HbA1c than the thiazolidinedione troglitazone.56 Meglitinides are ineffective in patients already on maximal doses of sulfonylureas and in patients who have failed sulfonylurea therapy.50,53 As monotherapy, meglitinides may be advantageous in patients whose hyperglycemia is mainly postprandial and in elderly patients prone to develop hypoglycemia.51,55 No adjustments in doses are needed in patients with renal insufficiency.49,53,57

Insulin Sensitizers

Metformin58 and thiazolidinediones59 (rosiglitazone and pioglitazone) fall into this category. These agents improve all of the actions of insulin in insulin-resistant patients with T2DM and have been shown to reduce insulin resistance in nondiabetic individuals with polycystic ovary syndrome.58,60

Metformin reduces the excessive glucose release found in people with T2DM, primarily by reducing gluconeogenesis.61-63 To a lesser extent, it improves the efficiency of glucose disposal. Although in animals it reduces intestinal glucose absorption, this does not apparently occur at doses used in humans.63 Metformin is available as 500-mg and 850-mg tablets taken 1-3 times daily and a sustained release preparation taken once a day (Glucophage XRâ); its maximally effective dose is 2000-2550 mg/d. As monotherapy, it is generally as effective as sulfonylureas, lowering HbA1c levels 1.5-2.0%4,58 in obese as well as nonobese patients.58 About a third of patients transiently develop gastrointestinal side effects (loose stools/epigastric discomfort), which can be minimized by starting with a low dose (500 mg) once or twice daily taken with meals.58 When used as monotherapy, hypoglycemia is not observed. Patients generally do not gain weight and actually may lose weight because it reduces appetite.4 Whereas sulfonylureas and meglitinides are generally neutral with respect to lipids, metformin may lower plasma triglycerides, especially if patients are hypertriglyceridemic,58 and has been reported to decrease plasma plasminogen activator inhibitor-1 (PAI-1) and fibrinogen levels.58 These and other actions may explain why in the UKPDS Trial, metformin had more beneficial effect on macrovascular end points than sulfonylureas.4

The most serious adverse event associated with metformin therapy is lactic acidosis, which has been reported to occur with an incidence of 0.03/1000 patient years—about the same as death from prolonged sulfonylurea hypoglycemia—and generally in patients who should not have been on metformin in the first place.58 In the UKPDS, there was not a single case.4 The contraindications for metformin use are basically those conditions associated with overproduction or underuse of lactate. These conditions include: pregnancy, heart failure, and hypoxic disorders because of overproduction due to increased anaerobic glycolysis; hepatic disease, renal insufficiency, and alcoholism because of decreased lactate use. In addition, because metformin is not metabolized and its sole route of excretion is via the kidney, it will accumulate in renal insufficiency. Diabetic patients undergoing contrast media studies can develop acute renal failure, and it is therefore recommended that metformin be withheld 24 hours before the procedure and not reinstituted until renal function is known to be intact. It is also recommended that metformin not be used in patients with heart disease requiring medication because an acute deterioration in cardiac output could precipitate lactic acidosis.

Thiazolidinediones are useful alternatives to metformin when the latter is contraindicated or not tolerated. These agents lower HbA1c levels on average 1.0-1.5% as monotherapy.59 They are thought to act primarily by binding to intracellular PPAR-g receptors and this complex then binds to DNA and alters regulation of synthesis of various factors involved in lipid and glucose metabolism.59 Some of their actions may be indirect, mediated through reductions in FFA, tumor necrosis factor alpha and resistin and through reductions in intracellular lipid in liver, muscle, and pancreatic islets.35,64 As insulin sensitizers, like metformin, they improve all actions of insulin in insulin-resistant patients. It is presently controversial whether they exert a selective effect on glucose disposal in contrast to metformin’s apparent selective effect on glucose production.61,62 Rosiglitazone (2-8 mg) and pioglitazone (15-45 mg) are both effective at once-daily dosing. Their onset of action is slower than those of secretogogues and metformin and it may take up to 4 months to observe their maximum effect. Like secretogogues and insulin, their use is associated with weight gain, part of which is due to fluid retention (which can cause a dilutional reduction in hemoglobin values) and a redistribution of fat from visceral to subcutaneous sites.59 Because of the fluid retention that has been reported to exacerbate heart failure, use of these agents is contraindicated in patients with heart failure Class III and IV and other disorders in which fluid retention would be a problem.59

As a class, these agents increase HDL and LDL cholesterol levels and are either neutral or lower plasma triglycerides.59 The increase in plasma LDL is in the buoyant fraction rather than the small dense particles so that the net effect on plasma lipids is probably beneficial.65 Like metformin, hypoglycemia is not observed when these agents are used as monotherapy and several nonglucose-related effects have been reported, eg, decreases plasma PAI-1,66 improvement of endothelial function,66 and, in animal experiments, preservation of beta cell function.67 The latter have been confirmed in short-term (2-year) human studies.68 If confirmed in longer-term studies, use of these agents to prevent T2DM and as early first-line therapy alone or in combination with other agents may be warranted. Currently, since these agents are relatively expensive compared to sulfonylureas and metformin, slow to reach maximum effect, and are not more effective than sulfonylureas or metformin, they are generally not used as initial monotherapy. In Europe, they are approved only as combination therapy (see below).

A major concern with this class has been hepatoxicity. The first thiazolidinedione on the market in the United States, troglitazone, was withdrawn because about 2-4% of patients developed an idiosyncratic reaction similar to that observed with isoniazid. This resulted in numerous deaths or liver transplantation when the drug was not stopped in time. So far, rosiglitazone and pioglitazone have not shown this hepatoxicity and it is generally not considered to be a class effect, but rather one that was unique for troglitazone. Differences in the metabolism of troglitazone offer an explanation of why this may not be a class effect.59 Nevertheless, it is currently recommended that liver function tests be monitored before therapy and every other month thereafter for a year, in patients treated with these agents, and that they not be used in patients with plasma hepatic enzyme levels more than 2 times the upper limit of normal.

a Glucosidase Inhibitors

These agents (acarbose and miglitol) are competitive inhibitors of intestinal a glucosidase and consequently slow the breakdown of dietary starches and thus the intestinal absorption of glucose derived from ingested complex carbohydrates.69,70 The primary effect of these agents, therefore, is on postprandial hypoglycemia with little or no effect on fasting hyperglycemia. In general, they lower HbA1c levels by 0.5-1.0% and, like the meglitinides, must be given just prior to each meal. To reduce side effects, initial doses are generally 25 mg with subsequent increments of 25 mg per dose every 2 or 3 weeks up to a maximal dose of 100 mg before each meal. Although extremely safe, these agents have a common side effect (flatus, which often makes them unacceptable to patients) due to undigested carbohydrate reaching the large intestine, being metabolized by gas-producing bacteria. This side effect can be reduced by starting at low doses, which allow induction of the enzyme in the distal part of the small intestine.

New Insulins

Insulin Glargine. As indicated earlier, T2DM is a disorder of progressive deterioration in beta cell function and therefore, eventually, many patients will require some form of insulin therapy. It has been common practice that when patients on 2 or more oral agents fail to achieve adequate glycemic control (HbA1c < 7.0%), to add bedtime NPH to the regimen while keeping the patient on one oral agent.71 Such a regimen has been reported to be capable of reducing HbA1c levels from 9.8% to 7.0%.71 Recently, an alternative to NPH, insulin glargine, has become available.72 This insulin analog is the result of an exchange of glycine for asparagine at the end of the A chain of insulin and the addition of 2 arginine residues on the B chain. These modifications cause an altered isoelectric point72,73 that results in a clear soluble insulin that precipitates in subcutaneous tissue. As a consequence, its absorption into the circulation is peakless and prolonged.74 Its duration of action, for example, averages about 24 hours compared to about 14 hours with NPH.74 Clinical trials comparing glargine with NPH in patients with T1 or T2DM have demonstrated that glargine is at least as efficacious as NPH while significantly reducing the frequency of nocturnal hypoglycemia.73,75,76 From a practical point of view, an additional advantage of the long duration of this insulin is that at steady state, missing a bedtime injection of glargine would be expected to have a minimal effect on the next day’s fasting plasma glucose level and the patient could inject the glargine in the morning with minimal effect on daytime glycemia. In contrast, with NPH in this scenario, fasting plasma glucose levels would be increased and taking NPH in the morning would significantly alter daytime values. When a patient is already taking a single dose of NPH, glargine can be substituted unit for unit. When a patient is taking 2 injections of NPH, it is recommended that glargine be started at 80% of the sum of the NPH dosage.72 Glargine is also used as a part of the basal component of a basal-bolus insulin regimen (see below).

Insulin Lispro (Humalog R) and Aspart (Novolog R). With the progressive deterioration in beta cell function as part of the natural history of T2DM, many patients will ultimately require more sophisticated insulin regimens. Indeed, the Veteran’s Administration Study showed that about a third of obese T2DM patients ultimately need a basal-bolus regimen (ie, preprandial short-acting insulin plus a long-acting insulin) to maintain HbA1c levels below 7.0%.77

Insulin lispro results from inverting the 28th and 29th B-chain aminoacids from proline-lysine to lysine-proline. Insulin aspart is the result of a substitution of aspartic acid for proline at position 28 on the B-chain.78 As a consequence, both these insulin analogs are less prone to self-aggregation and exist as monomers after subcutaneous injection, leading to more rapid absorption into the circulation than regular insulin, which exists as hexomers after injection. This rapid absorption also results in a shorter duration of action of these analogs than regular insulin (2-4 h vs 4-6 h). To achieve maximal effectiveness, regular insulin must be injected 30-45 minutes prior to meals.79 Most patients do not do this. These new insulin analogs can be injected ~ 5 minutes prior to meal ingestion with maximal effectiveness. Thus, use of these analogs has been shown to improve glycemic control compared to the way people customarily use regular insulin.80 Furthermore, because of the short duration of these analogs, there is less hypoglycemia,80 but in patients with T1DM this often requires 2 injections of NPH and occasionally of glargine. Overall, these insulins represent a substantial improvement in terms of convenience and safety for patients. There have been no head-to-head comparisons of lispro and insulin aspart in terms of clinical efficacy so that at this time no recommendation between them can be made. There are subtle differences in their pharmacokinetics81 but the clinical importance of these differences is unclear. Lispro (75/25) and aspart (70/30) insulins are also available as mixtures of free and protamine conjugated analog at least comparable to 70/30 mixtures of regular and NPH insulin.72,82

Combination Therapy. The UKPDS showed that less than 50% of patients who initially achieved a HbA1c < 7.0% on sulfonylurea or metformin monotherapy still had an HbA1c < 7.0% at 3 years.83 The results of the Belfast Study84 indicate that this is due to progressive deterioration in beta cell function rather than changes in insulin sensitivity. It can therefore be inferred that even if patients achieve initial adequate glycemic control with monotherapy, most will ultimately require additional medication. It is of no benefit to change medication class in patients who have failed to achieve adequate glycemic control on maximal dose of either a secretogogue or a sensitizer.85 Rather, addition of a second agent, which works via a different mechanism, is used in combination when a single agent has failed to provide adequate glycemic control. The results of several clinical trials indicating the reductions to be expected from different combinations are presented in Table 2.55,68,77,85-89

Table 2. Effect of Different Combination Therapies
Combination of Drugs HbA1c

Initial Drug Add-on Drug Reference Duration Initial Decrement
Glyburide Metformin 85 29 wks 8.7 1.6
Metformin Rosiglitazone 68 26 wks 8.8 1.2
Metformin Pioglitazone 88 72 wks 9.86 1.36
Nateglinide Metformin 55 24 wks 8.4 0.9
Metformin Repaglinide 87 3 mos 8.3 1.1
Metformin Acarbose 89 12 mos 7.8 0.8
Glyburide Acarbose 89 12 mos 8.0 0.9

A combination pill of glyburide and metformin (Glucovance®) is already on the market (1.25/250, 2.5/500, 5/500 mg glyburide/metformin). This combination has been shown to lower HbA1c more than monotherapy with larger doses of the individual drugs, with less overall side effects (less hypoglycemia than glyburide; less gastrointestinal side effects than metformin).90 Since T2DM is accompanied by both beta cell dysfunction and insulin resistance, an argument can be made for using combination therapy (a secretogogue and a sensitizer) as initial management to treat both of these defects.

The question arises as to how to proceed when a patient fails to achieve adequate glycemic control on a combination of 2 oral agents. Should a third oral agent be added, or should bedtime insulin be used? Aside from cost factors (3 pills are more expensive than 1 pill + insulin), it has been shown that addition of a third pill can further lower HbA1c.86,91 Thus, if a patient on 2 pills has an HbA1c > 8.5%, they will usually not achieve adequate glycemic control (HbA1c < 7.0%) on 3 pills and instead, insulin should be started. For patients with lower HbA1c levels, a third oral agent may be tried on the basis of convenience. All classes of oral agents may be used in combination with insulin.71,91 It has been demonstrated that metformin plus bedtime NPH was superior to glyburide plus bedtime NPH in terms of greater HbA1c lowering and less hypoglycemia.71 There are no comparisons of metformin with thiazolidinediones or with insulin secretogogues other than glyburide used with bedtime insulin. Conceivably, different results may be achieved when using glargine rather than NPH as bedtime insulin.

When a patient fails to achieve adequate glycemic control using a combination of an oral agent plus bedtime insulin, several progressively more complex insulin regimens may be used: 1) twice daily NPH; 2) morning 70/30 or 75/25 insulin mixtures plus bedtime NPH or glargine; 3) prebreakfast and predinner 70/30 or 75/25 insulin mixtures and finally; 4) a basal bolus regimen consisting of preprandial short-acting insulin (regular, lispro or aspart) plus NPH or glargine. Researchers tend to reserve premixed insulins for those who have difficulty mixing insulins or the elderly. A drawback is that one may not change the dose of one insulin species without changing the other. In general, when a patient is on 2 or more injections of insulin, concomitant use of an oral agent provides little benefit vs. using more insulin and is less cost-effective. The Veterans Administration Study77 showed that a third of obese T2DM subjects required a basal-bolus regimen. In this study, overall daily insulin doses averaging about 1.3 U/kg (130 U/d) and achieved an average HbA1c of 6.9% with little or no weight gain or serious hypoglycemia.

Summary and Conclusions

During the last decade, great strides have been made in our knowledge of the pathogenesis of T2DM, the relationship between glycemic control and complications, and in our ability to achieve adequate glycemic control due to the introduction of new therapeutic modalities. Both acquired insulin resistance and genetically determined deterioration in insulin secretion are key elements in the pathogenesis of this condition and both, therefore, are logical targets for therapeutic intervention. Clinical trials have clearly demonstrated that to prevent the development and slow the progression of micro- and macrovascular complications, it is desirable to maintain HbA1c below 6.5%. Values above 7.5% are no longer considered acceptable. Use of new oral agents that work by different mechanisms alone or in combination with each other or insulin, and use of various insulin regimens as needed can reduce HbA1c with greater safety and efficacy than was previously possible.


We thank Mary Little for her excellent editorial assistance. The present work was supported in part by Division of Research Resources-GCRC Grant 5MO1 RR-00044 and the National Institute of Diabetes and Digestive and Kidney Diseases Grant DK-20411.


1. DCCT Research Group. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin dependent diabetes mellitus. N Engl J Med. 1993;329:977-986.

2. Reichard P, et al. Complications in IDDM are caused by elevated blood glucose level: The Stockholm Diabetes Intervention Study (SDIS) at 10-year follow up. Diabetologia. 1996;39:1483-1488.

3. Ohkubo Y, et al. Intensive insulin therapy prevents the progression of diabetic microvascular complications in Japanese patients with non-insulin-dependent diabetes mellitus: A randomized prospective 6-year study. Diabetes Res Clin Pract. 1995;28: 103-117.

4. UK Prospective Diabetes Study (UKPDS) Group. Effect of intensive blood-glucose control with metformin on complications in overweight patients with type 2 diabetes (UKPDS 34). Lancet. 1998;352:854-865.

5. Fuller J, et al. Mortality from coronary heart disease and stroke in relation to degree of glycemia: The Whitehall study. BMJ. 1983;287:867-870.

6. UK Prospective Diabetes Study Group. Tight blood pressure control and risk of macrovascular and microvascular complications in type 2 diabetes: UKPDS 38. BMJ. 1998;317:703-713.

7. Fukagawa N, et al. Insulin-mediated reduction of whole body protein breakdown. J Clin Invest. 1985;76:2306-2311.

8. University Group Diabetes Program. A study on the effects of hypoglycemic agents on vascular complications in patients with adult-onset diabetes. II. Mortality results. Diabetes. 1970;19 (Suppl 9):747-830.

9. Stratton I, et al. Association of glycaemia with macrovascular and microvascular complications of type 2 diabetes (UKPDS 35): Prospective observational study. BMJ. 2000;321:405-412.

10. Rodriguez B, et al. Glucose intolerance and 23-year risk of coronary heart disease and total mortality: The Honolulu Heart Program. Diabetes Care. 1999;22:1262-1265.

11. Shaw J, et al. Isolated post-challenge hyperglycaemia confirmed as a risk factor for mortality. Diabetologia. 1999;42:1050-1054.

12. Tominaga M, et al. Impaired glucose tolerance is a risk factor for cardiovascular disease, but not impaired fasting glucose. The Funagata Diabetes Study. Diabetes Care. 1999;22:920-924.

13. Donahue R, et al. Postchallenge glucose concentration and coronary heart disease in men of Japanese ancestry. Honolulu Heart Program. Diabetes. 1987;36:689-692.

14. Hanefeld M, Temelkova-Kurktschiev T. The postprandial state and the risk of atherosclerosis. Diabet Med. 1997;14(Suppl 3): S6-S11.

15. The DECODE (Diabetes Epidemiology: Collaborative Analysis of Diagnostic Criteria in Europe) Study Group. Is fasting glucose sufficient to define diabetes? Epidemiological data from 20 European studies. Diabetologia. 1999;42:647-654.

16. Gerstein H, et al. Relationship of glucose and insulin levels to the risk of myocardial infarction: a case-control study. J Am Coll Cardiol. 1999;33:612-619.

17. Ceriello A. The post-prandial state and cardiovascular disease: relevance to diabetes mellitus. Diabetes Metab Res Rev. 2000;16:125-132.

18. Lefebvre P, Scheen A. The postprandial state and risk of cardiovascular disease. Diabet Med. 1998;15(Suppl 4):S63-S68.

19. Haller H. Postprandial glucose and vascular disease. Diabet Med. 1997;14(Suppl 3):S50-S56.

20. Khaw KT, et al. Glycated haemoglobin, diabetes, and mortality in men in Norfolk cohort of European Prospective Investigation of Cancer and Nutrition (EPIC-Norfolk). BMJ. 2001;322:15-18.

21. Sinclair AJ, et al. A desktop guide for the management of non-insulin-dependent diabetes mellitus (NIDDM): An update. European NIDDM Policy Group. Diabet Med. 1995;12:724.

22. Gerich J. The genetic basis of type 2 diabetes mellitus: Impaired insulin secretion versus impaired insulin sensitivity. Endocr Rev. 1998;19:491-503.

23. Yki-Järvinen H, Koivisto V. Effects of body composition on insulin sensitivity. Diabetes. 1983;32:965-969.

24. Oakes N, et al. Mechanisms of liver and muscle insulin resistance induced by chronic high-fat feeding. Diabetes. 1997;46: 1768-1774.

25. Yki-Järvinen H. Glucose toxicity. Endocr Rev. 1992;13:415-431.

26. Boden G. Role of fatty acids in the pathogenesis of insulin resistance and NIDDM. Diabetes. 1997;46:3-10.

27. Porte D Jr., Kahn S. Beta-cell dysfunction and failure in type 2 diabetes: Potential mechanisms. Diabetes. 2001;50(Suppl 1):S160-S163.

28. Buchanan T. Pancreatic B-cell defects in gestational diabetes: implications for the pathogenesis and prevention of type 2 diabetes. J Clin Endocrinol Metab. 2001;86:989-993.

29. Dornhorst A, et al. A defect in insulin release in women at risk of future non-insulin-dependent diabetes. Clin Sci. 1991;81: 195-199.

30. Pimenta W, et al. Pancreatic beta-cell dysfunction as the primary genetic lesion in NIDDM. JAMA. 1995;273:1855-1861.

31. Vaag A, et al. Intracellular glucose and fat metabolism in identical twins discordant for non-insulin-dependent diabetes mellitus (NIDDM): Acquired versus genetic metabolic defects. Diabetic Med. 1996;13:806-815.

32. UK Prospective Diabetes Study Group. UK Prospective Diabetes Study 16: Overview of 6 year’s therapy of type II diabetes: a progressive disease. Diabetes. 1995;44:1249-1258.

33. Dohm GL, et al. An in vitro human muscle preparation suitable for metabolic studies. Decreased insulin stimulation of glucose transport in muscle from morbidly obese and diabetic subjects. J Clin Invest. 1988;82:486-494.

34. Hotamisligil G, Spiegelman B. Tumor necrosis factor a: a key component of the obesity-diabetes link. Diabetes. 1994;43: 1271-1278.

35. Steppan C, et al. The hormone resistin links obesity to diabetes. Nature. 2001;409:307-312.

36. Mitrakou A, et al. Contribution of abnormal muscle and liver glucose metabolism in postprandial hyperglycemia in noninsulin-dependent diabetes mellitus. Diabetes. 1990;39:1381-1390.

37. Dinneen S, et al. Carbohydrate metabolism in noninsulin-dependent diabetes mellitus. N Engl J Med. 1992;327:707-713.

38. Mitrakou A, et al. Role of reduced suppression of glucose production and diminished early insulin release in impaired glucose tolerance. N Engl J Med. 1992;326:22-29.

39. Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Report of the expert committee on the diagnosis and classification of diabetes mellitus. Diabetes Care. 1998;21:S5-S19.

40. UK Prospective Diabetes Study Group. UK Prospective Diabetes Study (UKPDS) VIII. Study design, progress and performance. Diabetologia. 1991;34:877-890.

41. Lebovitz H. Insulin secretogogues: Old and new. Diab Rev. 1999;7:139-153.

42. Gerich J. Oral hypoglycemic agents. N Engl J Med. 1989;321:1231-1245.

43. Hollander P, et al. Importance of early insulin secretion. Comparison of nateglinide and glyburide in previously diet-treated patients with type 2 diabetes. Diabetes Care. 2001;24:983-988.

44. Langtry H, Balfour J. Glimepiride. A review of its use in the management of type 2 diabetes mellitus. Drugs. 1998;55: 563-584.

45. Klepzig H, et al. Sulfonylureas and ischaemic preconditioning. A double-blind, placebo-controlled evaluation of glimepiride and glibenclamide. Eur Heart J. 1999;20:439-446.

46. Harrower A. Comparative tolerability of sulphonylureas in diabetes mellitus. Drug Saf. 2000;22:313-320.

47. Schade D, et al. A placebo-controlled, randomized study of glimepiride in patients with type 2 diabetes mellitus for whom diet therapy is unsuccessful. J Clin Pharmacol. 1998;38:636-641.

48. Dunn C, Faulds D. Nateglinide. Drugs. 2000;60:607-615.

49. Jovanovic L, et al. Repaglinide in type 2 diabetes: A 24-week fixed-dose efficacy and safety study. J Clin Pharmacol. 2000;40:49-57.

50. Landgraf R, et al. A comparison of repaglinide and glibenclamide in the treatment of type 2 diabetic patients previously treated with sulphonylureas. Eur J Clin Pharmacol. 1999;55:165-171.

51. Hatorp V, et al. Pharmacokinetic profiles of repaglinide in elderly subjects with type 2 diabetes. J Clin Endocrinol Metab. 1999;84:1475-1478.

52. Damsbo P, et al. A double-blind randomized comparison of meal-related glycemic control by repaglinide and glyburide in well-controlled type 2 diabetic patients. Diabetes Care. 1999;22: 789-794.

53. Marbury T, et al. Repaglinide versus glyburide: a one-year comparison trial. Diabetes Res Clin Pract. 1999;43:155-166.

54. Madsbad S, et al. Comparison between repaglinide and glipizide in type 2 diabetes mellitus: A 1-year multicenter study. Diabet Med. 2001;18:395-401.

55. Horton E, et al. Nateglinide alone and in combination with metformin improves glycemic control by reducing mealtime glucose levels in type 2 diabetes. Diabetes Care. 2000;23: 1660-1665.

56. Raskin P, et al. Repaglinide/troglitazone combination therapy. Improved glycemic control in type 2 diabetes. Diabetes Care. 2000;23:979-983.

57. Schumacher S, et al. Single- and multiple-dose pharmacokinetics of repaglinide in patients with type 2 diabetes and renal impairment. Eur J Clin Pharmacol. 2001;57:147-152.

58. Cusi K, DeFronzo R. Metformin: A review of its metabolic effects. Diab Rev. 1998;6:89-131.

59. Mudaliar S, Henry R. New oral therapies for type 2 diabetes mellitus: The glitazones or insulin sensitizers. Annu Rev Med. 2001;52:239-257.

60. Iuorno M, Nestler J. Insulin-lowering drugs in polycystic ovary syndrome. Obstet Gynecol Clin North Am. 2001;28:153-164.

61. Stumvoll M, et al. Metabolic effects of metformin in non-insulin-dependent diabetes mellitus. N Engl J Med. 1995;333:550-554.

62. Inzucchi S, et al. Efficacy and metabolic effects of metformin and troglitazone in type II diabetes mellitus. N Engl J Med. 1998;338:867-872.

63. Jackson R, et al. Mechanism of metformin action in non-insulin-dependent diabetes. Diabetes. 1987;36:632-640.

64. Hotamisligil G. The role of TNFa and TNF receptors in obesity and insulin resistance. J Intern Med. 1999;245:621-625.

65. Thomas J, Taylor K. Effects of thiazolidinediones on lipoprotein subclasses in patients who are insulin resistant. Diabetes. 2001;50(Suppl 2):A455(Abstract).

66. Parulkar A, et al. Nonhypoglycemic effects of thiazolidinediones. Ann Intern Med. 2001;134:61-71.

67. Finegood D, et al. B-cell mass dynamics in Zucker diabetic fatty rats. Rosiglitazone prevents the rise in net cell death. Diabetes. 2001;50:1021-1029.

68. Fonseca V, et al. Effect of metformin and rosiglitazone combination therapy in patients with type 2 diabetes mellitus. JAMA. 2000;283:1695-1702.

69. Campbell L, et al. Miglitol: Assessment of its role in the treatment of patients with diabetes mellitus. Ann Pharmacother. 2000;34:1291-1301.

70. Göke B, Herrmann-Rinke C. The evolving role of a-glucosidase inhibitors. Diabetes Metab Rev. 1998;14(Suppl 1):S31-S38.

71. Yki-Järvinen H, et al. Comparison of bedtime insulin regimens in patients with type 2 diabetes mellitus. A randomized, controlled trial. Ann Intern Med. 1999;130:389-396.

72. Bolli G, et al. Insulin analogues and their potential in the management of diabetes mellitus. Diabetologia. 1999;42:1151-1167.

73. Raskin P, et al. A 16-week comparison of the novel insulin analog insulin glargine (HOE 901) and NPH human insulin used with insulin lispro in patients with type I diabetes. Diabetes Care. 2000;23:1666-1671.

74. Lepore M, et al. Pharmacokinetics and pharmacodynamics of subcutaneous injection of long-acting human insulin analog glargine, NPH insulin, and ultralente human insulin and continuous subcutaneous infusion of insulin lispro. Diabetes. 2000;49:2142-2148.

75. Ratner R, et al. Less hypoglycemia with insulin glargine in intensive insulin therapy for type I diabetes. Diabetes Care. 2000;23:639-643.

76. Rosenstock J, et al. Basal insulin therapy in type 2 diabetes. Diabetes Care. 2001;24:631-636.

77. Abraira C, et al. Response to intensive therapy steps and glipizide dose in combination with insulin in type 2 diabetes. Diabetes Care. 1998;21:574-579.

78. Mudaliar S, et al. Insulin aspart (B28 Asp-Insulin): a fast-acting analog of human insulin. Absorption kinetics and action profile compared with regular human insulin in healthy nondiabetic subjects. Diabetes Care. 1999;22:1501-1506.

79. Dimitriadis G, Gerich J. Importance of timing preprandial subcutaneous insulin administration in the management of diabetes mellitus. Diabetes Care. 1983;6:374-377.

80. Home P, et al. Improved glycemic control with insulin aspart: a multicenter randomized double-blind crossover trial in type 1 diabetic patients. Diabetes Care. 1998;21:1904-1909.

81. Hedman C, et al. Direct comparison of insulin lispro and aspart shows small differences in plasma insulin profiles after subcutaneous injection in type 1 diabetes. Diabetes Care. 2001;24: 1120-1121.

82. Heise T, et al. Time-action profiles of novel premixed preparations of insulin lispro and NPL insulin. Diabetes Care. 1998;21:800-803.

83. Turner R, et al. Glycemic control with diet, sulfonylurea, metformin, or insulin in patients with type 2 diabetes mellitus. Progressive requirement for multiple therapies (UKPDS 49). JAMA. 1999;281:2005-2012.

84. Levy J, et al. Beta-cell deterioration determines the onset and rate of progression of secondary dietary failure in type 2 diabetes mellitus: the 10-year follow-up of the Belfast Diet Study. Diabetic Med. 1998;15:290-296.

85. DeFronzo R, Goodman A. Efficacy of metformin in patients with non-insulin-dependent diabetes mellitus. N Engl J Med. 1995;333:541-549.

86. Gavin L, et al. Troglitazone add-on therapy to a combination of sulfonylureas plus metformin achieved and sustained effective diabetes control. Endocr Pract. 2000;6:305-310.

87. Moses R, et al. Effect of repaglinide addition to metformin monotherapy on glycemic control in patients with type 2 diabetes. Diabetes Care. 1999;22:119-124.

88. Einhorn D, et al. Pioglitazone hydrochloride in combination with metformin in the treatment of type 2 diabetes mellitus: a randomized, placebo-controlled study. Clinical Therapeutics. 2000;22:1395-1409.

89. Chiasson J, et al. The efficacy of acarbose in the treatment of patients with non-insulin-dependent diabetes mellitus. A multicenter controlled clinical trial. Ann Intern Med. 1994;121: 928-935.

90. Garber A, et al. Durable antidiabetic effect of glyburide/metformin tablets as initial therapy for type 2 diabetes. Diabetes. 2001;50(Suppl 2):A113(Abstract).

91. Bell D, Ovalle F. How long can insulin therapy be avoided in the patient with type 2 diabetes mellitus by use of a combination of metformin and a sulfonylurea? Endocr Pract. 2000;6:293-295.

Readers are Invited. . .

Readers are invited to submit questions or comments on material seen in or relevant to Primary Care Reports. Send your questions to: Robin Mason, Primary Care Reports, c/o American Health Consultants, P.O. Box 740059, Atlanta, GA 30374. For subscription information, you can reach the editors and customer service personnel for Primary Care Reports via the internet by sending e-mail to