Diabetes Management: Pharmacologic Therapy

PHARMACOLOGIC THERAPY

As stated earlier, insulin is secreted by the beta cells of the islets of Langerhans and works to lower the blood glucose level after meals by facilitating the uptake and utilization of glucose by muscle, fat, and liver cells. In the absence of adequate insulin, pharmacologic therapy is essential.

Insulin Therapy and Insulin Preparations

Because the body loses the ability to produce insulin in type 1 di-abetes, exogenous insulin must be administered for life. In type 2 diabetes, insulin may be necessary on a long-term basis to control glucose levels if diet and oral agents fail. In addition, some pa-tients in whom type 2 diabetes is usually controlled by diet alone or by diet and an oral agent may require insulin temporarily during illness, infection, pregnancy, surgery, or some other stress-ful event. In many cases, insulin injections are administered two or more times daily to control the blood glucose level. Because the insulin dose required by the individual patient is determined by the level of glucose in the blood, accurate monitoring of blood glucose levels is essential; thus, SMBG has become a cornerstone of insulin therapy. A number of insulin preparations are available. They vary according to three main characteristics: time course of action, species (source), and manufacturer.

TIME COURSE OF ACTION

Insulins may be grouped into several categories based on the onset, peak, and duration of action (Table 41-3). Human insulin preparations have a shorter duration of action than insulin from animal sources because the presence of animal proteins triggers an immune response that results in the binding of animal insulin, which slows its availability.

Rapid-acting insulins such as insulin lispro (Humalog) and in-sulin aspart (Novolog) are blood glucose-lowering agents that produce a more rapid effect that is of shorter duration than reg-ular insulin. These insulins have an onset of 5 to 15 minutes, a peak action of 1 hour after injection, and a duration of 2 to 4 hours. Because of their rapid onset, patients should be instructed to eat no more than 5 to 15 minutes after injection. Because of the short duration of action of these insulin analogs, patients with type 1 diabetes and some patients with type 2 or gestational diabetes also require a long-acting insulin to maintain glucose control. Basal insulin is necessary to maintain blood glucose levels irrespective of meals. A constant level of insulin is required at all times. Intermediate-acting insulins function as basal insulins but may have to be split into two injections to achieve 24-hour coverage.

Short-acting insulins, called regular insulin (marked R on the bottle), have an onset of 30 minutes to 1 hour; peak, 2 to 3 hours; and duration, 4 to 6 hours. Regular insulin is a clear solution and is usually administered 20 to 30 minutes before a meal, either alone or in combination with a longer-acting insulin. Humulin R, Iletin Regular, and Novolin R are examples of regular insulin.

Intermediate-acting insulins, called NPH insulin (neutral pro-tamine Hagedorn) or Lente insulin, have an onset of 3 to 4 hours; peak, 4 to 12 hours; and duration, 16 to 20 hours. Intermediate-acting insulins, which are similar in their time course of action, ap-pear white and cloudy. If NPH or Lente insulin is taken alone, it is not crucial that it be taken 30 minutes before the meal. It is im-portant, however, for the patient to eat some food around the time of the onset and peak of these insulins. Humulin N, Iletin NPH, and Novolin N are examples of NPH insulins; Humulin L, Iletin L, and Novolin L are examples of Lente insulins.

Long-acting insulins, called Ultralente insulin, are sometimes referred to as peakless insulins because they tend to have a long, slow, sustained action rather than sharp, definite peaks in action. The onset of long-acting human insulin is 6 to 8 hours; peak, 12 to 16 hours; and duration, 20 to 30 hours.

“Peakless” basal insulin, insulin glargine (Lantus), is approved for use as a basal insulin—that is, the insulin is absorbed very slowly over 24 hours and can be given once a day. Because the insulin is in a suspension with a pH of 4, it cannot be mixed with other insulins because this would cause precipitation. It is given once a day at bedtime.

In the future, “inhaled insulin” may be approved for use. This type of insulin is in the form of a very fine powder, which is in-haled through a device similar to that used to administer asthma medications. The patient’s program would consist of a “basal” rate of insulin such as glargine supplemented by an inhaled dose before each meal.

The nurse may find that different sources list differing num-bers of hours for the onset, peak, and duration of action of the main types of insulin, and patients’ responses may vary (ie, larger doses prolong onset, duration, and peak). The nurse should focus on which meals—and snacks—are being “covered” by which in-sulin doses. In general, the rapid- and short-acting insulins are ex-pected to cover the rise in glucose levels after meals, immediately after the injection; the intermediate-acting insulins are expected to cover subsequent meals; and the long-acting insulins provide a relatively constant level of insulin and act as a basal insulin.

SPECIES (SOURCE)

In the past, all insulins were obtained from beef (cow) and pork (pig) pancreases. “Human insulins” are now widely available. They are produced by recombinant DNA technology and have largely replaced insulin from animal sources (ADA, Insulin Administration, 2003).

MANUFACTURER

The two manufacturers of insulin in the United States are Eli-Lilly and Novo Nordisk. The insulins made by the different companies are usually interchangeable, provided the concentration (eg, U-100), species (eg, human), and type (eg, NPH) of insulin are the same. Human insulins made by different companies have different brand names. Therefore, a patient taking 20 units human NPH insulin may be using either Humulin N or Novolin N.

Insulin Regimens

Insulin regimens vary from one to four injections per day. Usually there is a combination of a short-acting insulin and a longer-acting insulin. The normally functioning pancreas continuously secretes small amounts of insulin during the day and night. In ad-dition, whenever blood glucose rises after ingestion of food, there is a rapid burst of insulin secretion in proportion to the glucose-raising effect of the food. The goal of all but the simplest, one-injection insulin regimens is to mimic this normal pattern of insulin secretion in response to food intake and activity patterns. Table 41-4 describes several insulin regimens and the advantages and disadvantages of each.

Patients can learn to use SMBG results and carbohydrate counting to vary the insulin doses. This allows patients more flex-ibility in the timing and content of meals and exercise periods. However, complex insulin regimens require a strong level of commitment, intensive education, and close follow-up by the health care team. In addition, patients aiming for normal blood glucose levels run the risk of more hypoglycemic reactions.

The type of regimen used by any particular patient varies. For example, patient knowledge, willingness, goals, health status, and finances all may affect decisions regarding insulin treatment. In addition, the physician’s philosophy about blood glucose control and the availability of equipment and support staff may influence decisions regarding insulin therapy. There are two general ap-proaches to insulin therapy: conventional and intensive.

CONVENTIONAL REGIMEN

One approach is to simplify the insulin regimen as much as pos-sible, with the aim of avoiding the acute complications of diabetes (hypoglycemia and symptomatic hyperglycemia). With this type of simplified regimen (eg, one or more injections of a mixture of short- and intermediate-acting insulins per day), patients may fre-quently have blood glucose levels well above normal. The excep-tion is the patient who never varies meal patterns and activity levels. This approach would be appropriate for the terminally ill, the frail elderly with limited self-care abilities, or any patient who is completely unwilling or unable to engage in the self-management activities that are part of a more complex insulin regimen.

INTENSIVE REGIMEN

The second approach is to use a more complex insulin regimen to achieve as much control over blood glucose levels as is safe and practical. The results of the landmark DCCT study (1993) and the UKPDS study (1998) have demonstrated that maintaining blood glucose levels as close to normal as possible prevents or slows the progression of long-term diabetic complications. Another reason for using a more complex insulin regimen is to allow pa-tients more flexibility to change their insulin doses from day to day in accordance with changes in their eating and activity pat-terns, with stress and illness, and as needed for variations in the prevailing glucose level.

Although the DCCT found that intensive treatment (three or four injections of insulin per day) reduced the risk of complica-tions, not all people with diabetes are candidates for very tight control of blood glucose. The risk for severe hypoglycemia was increased threefold in patients receiving intensive treatment in the DCCT (ADA, Implications of the Diabetes Control and Complications Trial, 2003). Those who may not be candidates include patients with:

•        Nervous system disorders rendering them unaware of hypo-glycemic episodes (eg, those with autonomic neuropathy)

•        Recurring severe hypoglycemia

•        Irreversible diabetic complications, such as blindness or end-stage renal disease

•        Cerebrovascular and/or cardiovascular disease

•        Ineffective self-care skills

An exception is the patient who has received a kidney trans-plant because of nephropathy and chronic renal failure; this pa-tient should be on an intensive regimen to preserve function of the new kidney.

The patient needs to be involved in the decision regarding which insulin regimen to use. Patients need to compare the po-tential benefits of different regimens with the potential costs (eg, time involved, number of injections or finger sticks for glu-cose testing, amount of record-keeping). There are no set guide-lines as to which insulin regimen should be used for which patients. It must not be assumed that an elderly patient or a patient with visual impairment should automatically be given a simplified regimen. Likewise, it must not be assumed that all people will want to be involved in a complex treatment regimen. Nurses play an important role in educating patients about the different ap-proaches to insulin therapy. Nurses should refer patients to dia-betes specialists or diabetes education centers, when available, for further training and education in the various insulin treatment regimens.

Complications of Insulin Therapy

LOCAL ALLERGIC REACTIONS

A local allergic reaction (redness, swelling, tenderness, and in-duration or a 2- to 4-cm wheal) may appear at the injection site 1 to 2 hours after the insulin administration. These reactions, which usually occur during the beginning stages of therapy and disappear with continued use of insulin, are becoming rare be-cause of the increased use of human insulins. The physician may prescribe an antihistamine to be taken 1 hour before the injection if such a local reaction occurs.

SYSTEMIC ALLERGIC REACTIONS

Systemic allergic reactions to insulin are rare. When they do occur, there is an immediate local skin reaction that gradually spreads into generalized urticaria (hives). The treatment is de-sensitization, with small doses of insulin administered in gradu-ally increasing amounts using a desensitization kit. These rare reactions are occasionally associated with generalized edema or anaphylaxis.

INSULIN LIPODYSTROPHY

Lipodystrophy refers to a localized reaction, in the form of either lipoatrophy or lipohypertrophy, occurring at the site of insulin injections. Lipoatrophy is loss of subcutaneous fat and appears as slight dimpling or more serious pitting of subcutaneous fat. The use of human insulin has almost eliminated this disfiguring complication.

Lipohypertrophy, the development of fibrofatty masses at the injection site, is caused by the repeated use of an injection site. If insulin is injected into scarred areas, absorption may be delayed.

This is one reason that rotation of injection sites is so important. The patient should avoid injecting insulin into these areas until the hypertrophy disappears.

INSULIN RESISTANCE

Most patients at one time or another have some degree of insulin resistance. This may occur for various reasons, the most common being obesity, which can be overcome by weight loss. Clinical in-sulin resistance has been defined as a daily insulin requirement of 200 units or more. In most diabetic patients taking insulin, im-mune antibodies develop and bind the insulin, thereby decreas-ing the insulin available for use. All animal insulins, as well as human insulins to a lesser degree, cause antibody production in humans.

Very few resistant patients develop high levels of antibodies. Many of these patients have a history of insulin therapy inter-rupted for several months or more. Treatment consists of admin-istering a more concentrated insulin preparation, such as U500, which is available by special order. Occasionally, prednisone is needed to block the production of antibodies. This may be fol-lowed by a gradual reduction in insulin requirement. Therefore, patients need to monitor themselves for hypoglycemia.

MORNING HYPERGLYCEMIA

An elevated blood glucose level upon arising in the morning may be caused by an insufficient level of insulin due to several causes: the dawn phenomenon, the Somogyi effect, or insulin waning. The dawn phenomenon is characterized by a relatively normal blood glucose level until approximately 3 a.m., when blood glucose levels begin to rise. The phenomenon is thought to result from nocturnal surges in growth hormone secretion that create a greater need for insulin in the early morning hours in patients with type 1 diabetes. It must be distinguished from insulin waning (the progressive increase in blood glucose from bedtime to morning) or the Somogyi effect (nocturnal hypo-glycemia followed by rebound hyperglycemia). Insulin waning is frequently seen if the evening NPH dose is administered be-fore dinner and is prevented by moving the evening dose of NPH insulin to bedtime.

It may be difficult to tell from the patient’s history which of these causes is responsible for morning hyperglycemia. To deter-mine the cause, the patient must be awakened once or twice during the night to test blood glucose levels. Testing the blood glucose level at bedtime, at 3 a.m., and on awakening provides information that can be used in making adjustments in insulin to avoid morning hyperglycemia caused by the dawn phenomenon. Table 41-5 summarizes the differences among insulin waning, the dawn phenomenon, and the Somogyi effect.

Alternative Methods of Insulin Delivery

INSULIN PENS

These devices use small (150- to 300-unit) prefilled insulin car-tridges that are loaded into a penlike holder. A disposable needle is attached to the device for insulin injection. Insulin is delivered by dialing in a dose or pushing a button for every 1- or 2-unit in-crement administered. People using these devices still need to in-sert the needle for each injection; however, they do not need to carry insulin bottles or to draw up insulin before each injection. These devices are most useful for patients who need to inject only one type of insulin at a time (eg, premeal regular insulin three times a day and bedtime NPH insulin) or who can use the pre-mixed insulins. These pens are convenient for those who administer insulin before dinner if eating out or traveling. They are also useful for patients with impaired manual dexterity, vision, or cog-nitive function that makes the use of traditional syringes difficult.

JET INJECTORS

As an alternative to needle injections, jet injection devices deliver insulin through the skin under pressure in an extremely fine stream. These devices are more expensive than other alternative devices mentioned above and require thorough training and su-pervision when first used. In addition, patients should be cau-tioned that absorption rates, peak insulin activity, and insulin levels may be different when changing to a jet injector. (Insulin administered by jet injector is usually absorbed faster.) Bruising has occurred in some patients with use of the jet injector.

INSULIN PUMPS

Continuous subcutaneous insulin infusion involves the use ofsmall, externally worn devices that closely mimic the functioning of the normal pancreas (ADA, Continuous Subcutaneous Insulin Infusion, 2003). Insulin pumps contain a 3-mL syringe attached to a long (24- to 42-in), thin, narrow-lumen tube with a needle or Teflon catheter attached to the end (Figs. 41-4 and 41-5). The patient inserts the needle or catheter into the subcutaneous tissue (usually on the abdomen) and secures it with tape or a transparent dressing. The needle or catheter is changed at least every 3 days. The pump is then worn either on a belt or in a pocket. Some women keep the pump tucked into the front or side of the bra or wear it on a garter belt on the thigh.

The rapid-acting lispro insulin is used in the insulin pump and is delivered at a basal rate and as a bolus with meals. A con-tinuous basal rate of insulin is typically 0.5 to 2.0 units/hour, depending on the patient’s needs. A bolus dose of insulin is delivered before each meal when the patient activates the pump (by pushing buttons). The patient determines the amount of in-sulin to infuse based on blood glucose levels and anticipated food intake and activity level. Advantages of insulin pumps include increased flexibility in lifestyle (in terms of timing and amount of meals, exercise, and travel) and, for many patients, improved blood glucose control.

A disadvantage of insulin pumps is that unexpected disrup-tions in the flow of insulin from the pump may occur if the tubing or needle becomes occluded, if the supply of insulin runs out, or if the battery is depleted, increasing the risk of DKA. Effective teaching and a knowledgeable patient can minimize this risk. Another disadvantage is the potential for infection at needle in-sertion sites. Hypoglycemia may occur with insulin pump ther-apy; however, this is usually related to the lowered blood glucose levels many patients achieve rather than to a specific problem with the pump itself. The tight diabetic control associated with using an insulin pump may increase the incidence of hypoglycemia un-awareness because of the very gradual decline in serum glucose level from levels greater than 70 mg/dL (3.9 mmol/L) to those less than 60 mg/dL (3.3 mmol/L).

Some patients find that wearing the pump for 24 hours each day is an inconvenience. However, the pump can easily be dis-connected, per patient preference, for limited periods (eg, for showering, exercise, or sexual activity).

Insulin pump candidates must be willing to assess blood glu-cose levels multiple times daily while on pump therapy. In addi-tion, they must be psychologically stable and open about having diabetes, because the insulin pump is often a visible sign to others and a constant reminder to the patient that he or she has diabetes. Most important, patients using insulin pumps must have exten-sive education in the use of the insulin pump and in self-management of blood glucose and insulin doses. They must work closely with a team of health care professionals who are experi-enced in insulin pump therapy—specifically, a diabetologist/ endocrinologist, a dietitian, and a certified diabetes educator.

Many insurance policies cover the cost of pump therapy; if it is not covered, the extra expense of the pump and associated sup-plies may be a deterrent for some patients. Medicare now covers insulin pump therapy for the patient with type 1 diabetes.

IMPLANTABLE AND INHALANT INSULIN DELIVERY

Research into mechanical delivery of insulin has involved im-plantable insulin pumps that can be externally programmed ac-cording to blood glucose test results. Clinical trials with these devices are continuing. In addition, there is research into the de-velopment of implantable devices that both measure the blood glucose level and deliver insulin as needed. Methods of adminis-tering insulin by the oral route (oral spray or capsule), skin patch, and inhalation are undergoing intensive study.

TRANSPLANTATION OF PANCREATIC CELLS

Transplantation of the whole pancreas or a segment of the pan-creas is being performed on a limited population (mostly diabetic patients receiving kidney transplantations simultaneously). One main issue regarding pancreatic transplantation is weighing the risks of antirejection medications against the advantages of pancreas transplantation. Another approach under investigation is the implantation of insulin-producing pancreatic islet cells (ADA, Pancreas Transplantation for Patients With Type 1 Dia-betes, 2003). This latter approach involves a less extensive surgi-cal procedure and a potentially lower incidence of immunogenic problems. However, thus far, independence from exogenous in-sulin has been limited to 2 years after transplantation of islet cells. A recent study of patients with islet cell transplants using less toxic antirejection drugs has shown promise (Shapiro et al., 2000).

Oral Antidiabetic Agents

Oral antidiabetic agents may be effective for patients who have type 2 diabetes that cannot be treated by diet and exercise alone; however, they cannot be used during pregnancy. In the United States, oral antidiabetic agents include the sulfonylureas, biguanides, alpha glucosidase inhibitors, thiazolidinediones, and meglitinides (Table 41-6). Sulfonylureas and meglitinides are considered in-sulin secretagogues because their action increases the secretion of insulin by the pancreatic beta cells.

SULFONYLUREAS

The sulfonylureas exert their primary action by directly stim-ulating the pancreas to secrete insulin. Therefore, a functioning pancreas is necessary for these agents to be effective, and they cannot be used in patients with type 1 diabetes. These agents improve insulin action at the cellular level and may also directly decrease glucose production by the liver. The sulfonylureas can be divided into first- and second-generation categories (see Table 41-6).

The most common side effects of these medications are GI symptoms and dermatologic reactions. Hypoglycemia may occur when an excessive dose of a sulfonylurea is used or when the patient omits or delays meals, reduces food intake, or in-creases activity. Because of the prolonged hypoglycemic effects of these agents (especially chlorpropamide), some patients need to be hospitalized for treatment of oral agent-induced hypo-glycemia. Another side effect of chlorpropamide is a disulfiram (Antabuse) type of reaction when alcohol is ingested (see section on alcohol consumption for more information). Some medica-tions may directly interact with sulfonylureas, potentiating their hypoglycemic effects (eg, sulfonamides, chloramphenicol, clofi-brate, phenylbutazone, and bishydroxycoumarin). In addition, certain medications may independently affect blood glucose lev-els, thereby indirectly interfering with these agents. Medications that may increase glucose levels include potassium-losing di-uretics, corticosteroids, estrogen compounds, and diphenylhy-dantoin (Dilantin). Medications that may cause hypoglycemia include salicylates, propranolol, monoamine oxidase inhibitors, and pentamidine.

Second-generation sulfonylureas have the advantage of a shorter half-life and excretion by both the kidney and the liver. This makes these medications safer to use in the elderly, in whom accumulation of the medication can cause recurring hypoglycemia.

BIGUANIDES

The biguanides are other kinds of oral antidiabetic agents. Metformin (Glucophage) produces its antidiabetic effects by fa-cilitating insulin’s action on peripheral receptor sites. Therefore, it can be used only in the presence of insulin. Biguanides have no effect on pancreatic beta cells. Biguanides used with a sulfonyl-urea may enhance the glucose-lowering effect more than either medication used alone. Lactic acidosis is a potential and serious complication of biguanide therapy; the patient must be monitored closely when therapy is initiated or when dosage changes. 

Medications that may interact with biguanides include anticoagulants, corticosteroids, diuretics, and oral contraceptives. Metformin is contraindicated in patients with renal impairment (serum creati-nine level more than 1.4) or those at risk for renal dysfunction (eg, those with acute myocardial infarction). Renal function studies should be performed periodically to ensure that function is not impaired. Metformin should not be administered for 2 days be-fore any diagnostic testing that may require use of a contrast agent. These situations increase the risk for lactic acidosis.

An extended-release form and a combination form (Gluco-vance) combines metformin with a sulfonylurea, such as glyburide. The combination provides two mechanisms of action and im-proved patient compliance. Hypoglycemia is a risk.

ALPHA GLUCOSIDASE INHIBITORS

Acarbose (Precose) and miglitol (Glyset) are oral alpha glucosi-dase inhibitors used in type 2 diabetes management. They workby delaying the absorption of glucose in the intestinal system, re-sulting in a lower postprandial blood glucose level. As a conse-quence of plasma glucose reduction, hemoglobin A_1C levels drop. In contrast to the sulfonylureas, acarbose and miglitol do not enhance insulin secretion. They can be used alone with dietary treatment as monotherapy or in combination with sulfonylureas, thiazolidinediones, or meglitinides. When these medications are used in combination with sulfonylureas or meglitinides, hypo-glycemia may occur. The patient must be advised that if hypo-glycemia occurs, sucrose absorption will be blocked and treatment for hypoglycemia should be in the form of glucose, such as glu-cose tablets. The advantage of oral alpha glucosidase inhibitors is that they are not systemically absorbed and are safe to use. Their side effects are diarrhea and flatulence. These effects may be min-imized by starting at a very low dose and increasing the dose grad-ually. Because acarbose and miglitol affect food absorption, they must be taken immediately before a meal, making therapeutic adherence a potential problem.

THIAZOLIDINEDIONES

Rosiglitizone (Avandia) and pioglitozone (Actos) are oral diabetes medications categorized as thiazolidinediones. They are indi-cated for patients with type 2 diabetes who take insulin injections and whose blood glucose control is inadequate (hemoglobin A_1C level greater than 8.5%). They have also been approved as first-line agents to treat type 2 diabetes, in combination with diet. Thi-azolidinediones enhance insulin action at the receptor site without increasing insulin secretion from the beta cells of the pancreas. These medications may affect liver function; therefore, liver func-tion studies must be performed at baseline and at frequent inter-vals (monthly for the first 12 months of treatment, and quarterly thereafter). Women should be informed that thiazolidinediones can cause resumption of ovulation in perimenopausal anovulatory women, making pregnancy a possibility.

MEGLITINIDES

Repaglinide (Prandin), an oral glucose-lowering agent of the class of oral agents called meglitinides, lowers the blood glucose level by stimulating insulin release from the pancreatic beta cells. Its effectiveness depends on the presence of functioning beta cells. Therefore, repaglinide is contraindicated in patients with type 1 diabetes. Repaglinide has a fast action and a short duration. It should be taken before each meal to stimulate the release of in-sulin in response to that meal. It is also indicated for use in com-bination with metformin in patients whose hyperglycemia cannot be controlled by exercise, diet, and either metformin or repaglinide alone. The principal side effect of repaglinide is hypo-glycemia; however, this side effect is less severe and frequent than for a sulfonylurea because repaglinide has a short half-life (approximately 1 hour). Patients must be taught the signs and symptoms of hypoglycemia and should understand that the medication should not be taken unless the patient eats a meal. Repaglinide is supplied in 0.5-, 1-, and 2-mg tablets.

Naglitinide (Starlix), another meglitinide, has a very rapid onset and short duration. It should be taken with meals and not taken if the meal is skipped. Hypoglycemia risk is low if taken correctly.

General Considerations for Oral Agents

Patients need to understand that oral agents are prescribed as an ad-dition to (not as a substitute for) other treatment modalities, such as diet and exercise. Use of oral antidiabetic medications may need to be halted temporarily and insulin prescribed if hyperglycemia develops that is attributable to infection, trauma, or surgery.

In time, oral antidiabetic agents may no longer be effective in controlling the patient’s diabetes. In such cases, the patient is treated with insulin. Approximately half of all patients who ini-tially use oral antidiabetic agents eventually require insulin. This is referred to as a secondary failure. Primary failure occurs when the blood glucose level remains high a month after initial med-ication use.

Because the mechanisms of action vary (Fig. 41-6), the ef-fect may be enhanced using multidose, multiple medications (Inzucchi et al., 1998). Use of multiple medications with dif-ferent mechanisms of action is very common today (Quinn, 2001b). Using a combination of oral agents with insulin has been proposed as a treatment for some patients with type 2 diabetes. However, the effectiveness of this approach has not yet been demonstrated.