Pharmacotherapy A Pathophysiologic Approach, 9th Ed.

57. Diabetes Mellitus

Curtis L. Triplitt, Thomas Repas, and Carlos A. Alvarez


 Images Diabetes mellitus (DM) is a group of metabolic disorders of fat, carbohydrate, and protein metabolism that results from defects in insulin secretion, insulin action (sensitivity), or both.

 Images The incidence of type 2 DM is increasing. This has been attributed in part to a Western-style diet, increasing obesity, sedentary lifestyle, and an increasing minority population.

 Images The two major classifications of DM are type 1 (insulin deficient) and type 2 (combined insulin resistance and relative deficiency in insulin secretion). They differ in clinical presentation, onset, etiology, and progression of disease. Both are associated with microvascular and macrovascular disease complications.

 Images Diagnosis of diabetes is made by four criteria: fasting plasma glucose ≥126 mg/dL (≥7 mmol/L), a 2-hour value from a 75-g oral glucose tolerance test ≥200 mg/dL (≥11.1 mmol/L), a casual plasma glucose level of ≥200 mg/dL (≥11.1 mmol/L) with symptoms of diabetes, or a hemoglobin A1c [HbA1c] ≥6.5% (≥0.065; ≥48 mmol/mol Hb). The diagnosis should be confirmed by repeat testing if obvious hyperglycemia is not present.

 Images Goals of therapy in DM are directed toward attaining normoglycemia (or appropriate glycemic control based on the patient’s comorbidities), reducing the onset and progression of retinopathy, nephropathy, and neuropathy complications, intensive therapy for associated cardiovascular risk factors, and improving quality and quantity of life.

 Images Metformin should be included in the therapy for all type 2 DM patients, if tolerated and not contraindicated, as it is the only oral antihyperglycemic medication proven to reduce the risk of total mortality, according to the United Kingdom Prospective Diabetes Study (UKPDS).

 Images Intensive glycemic control is paramount for reduction of microvascular complications (neuropathy, retinopathy, and nephropathy) as evidenced by the Diabetes Control and Complications Trial (DCCT) in type 1 DM and the UKPDS in type 2 DM. The UKPDS also reported that control of hypertension in patients with diabetes will not only reduce the risk of retinopathy and nephropathy but also reduce cardiovascular risk.

 Images Short-term (3 to 5 years), intensive glycemic control does not lower the risk of macrovascular events as reported by the Action in Diabetes and Vascular Disease, Action to Control Cardiovascular Risk in Diabetes, and Veterans Administration Diabetes Trial trials. Microvascular event reduction may be sustained, and macrovascular events reduced by improved early glycemic control, as evidenced by the UKPDS and DCCT follow-up studies. Significant reductions in macrovascular risk may take 15 to 20 years. This sustained reduction in microvascular risk and new reduction in macrovascular risk has been coined metabolic memory.

 Images Knowledge of the patient’s quantitative and qualitative meal patterns, activity levels, pharmacokinetics of insulin preparations, and pharmacology of oral and injected antihyperglycemic agents is essential to individualize the treatment plan and optimize blood glucose control while minimizing risks for hypoglycemia and other adverse effects of pharmacologic therapies.

 Images Type 1 DM treatment necessitates insulin therapy. Currently, the basal–bolus insulin therapy or pump therapy in motivated individuals often leads to successful glycemic outcomes. Basal–bolus therapy includes a basal insulin for fasting and postabsorptive control, and rapid-acting bolus insulin for mealtime coverage. Addition of mealtime pramlintide in patients with uncontrolled or erratic postprandial glycemia may be warranted.

 Images Type 2 DM treatment often necessitates use of multiple therapeutic agents (combination therapy), including oral and/or injected antihyperglycemics and insulin, to obtain glycemic goals due to the persistent reduction in β-cell function over time. Slowing, but not arresting, β-cell failure has been shown with thiazolidinediones and the glucagon-like peptide-1 (GLP-1) agonist class of medications.

 Images Aggressive management of cardiovascular disease risk factors in type 2 DM is necessary to reduce the risk for adverse cardiovascular events or death. Smoking cessation, use of antiplatelet therapy as a secondary prevention strategy and in select primary prevention situations, aggressive management of dyslipidemia—primary goal to lower low-density lipoprotein cholesterol (<100 mg/dL [<2.59 mmol/L]) and secondarily to raise high-density lipoprotein cholesterol to ≥40 mg/dL (≥1.03 mmol/L)—and treatment of hypertension (again often requiring multiple drugs) to <130/80 mm Hg are vital.

 Images Prevention strategies for type 1 DM have been unsuccessful. Prevention strategies for type 2 DM are established. Lifestyle changes, dietary restriction of fat, aerobic exercise for 30 minutes five times a week, and weight loss form the backbone of successful prevention. No medication is currently FDA approved for prevention of diabetes, although several, including metformin, acarbose, pioglitazone, and rosiglitazone, have clinical trials demonstrating a delay of diabetes onset.

 Images Patient education and ability to demonstrate self-care and adherence to therapeutic lifestyle and pharmacologic interventions are crucial to successful outcomes. Multidisciplinary teams of healthcare professionals including physicians (primary care, endocrinologists, ophthalmologists, and vascular surgeons), podiatrists, dietitians, nurses, pharmacists, social workers, behavioral health specialists, and certified diabetes educators are needed, as appropriate, to optimize these outcomes in persons with DM.

Images Diabetes mellitus (DM) is a heterogeneous group of metabolic disorders characterized by hyperglycemia. It is associated with abnormalities in carbohydrate, fat, and protein metabolism and may result in chronic complications including microvascular, macrovascular, and neuropathic disorders. It is estimated that in 2010, 26 million Americans ≥20 years old have DM, with as many as one fourth of these patients being undiagnosed, and an additional 79 million at high risk for the development of diabetes. The economic burden of DM approximated $218 billion in 2007, for diabetes and prediabetes. This is representative of an annual cost for each citizen of the United States of $700. DM is the leading cause of blindness in adults aged 20 to 74 years and the leading cause of end-stage renal disease in the United States. It also accounts for approximately 65,000 lower extremity amputations annually. Finally, a cardiovascular event is responsible for two thirds of deaths in individuals with type 2 DM and is the leading cause of death in type 1 DM of long duration.1

Optimal management of the patient with DM will reduce or prevent complications, decreasing morbidity and mortality while improving quality of life. Research, clinical trials, and drug development efforts over the past several decades have provided valuable information that applies directly to improving outcomes in patients with DM and have expanded the therapeutic armamentarium. Additionally, interventions in an attempt to prevent complications and the onset of diabetes have been reported for type 1 and 2 DM.


Type 1 DM accounts for 5% to 10% of all cases of DM and is most often due to autoimmune destruction of the pancreatic β cells.2 Although type 1 DM most frequently develops in childhood or early adulthood, new cases occur at any age.

Type 1 DM is thought to be initiated by the exposure of a genetically susceptible individual to an environmental agent. The development of β-cell autoimmunity occurs in less than 10% of the genetically susceptible population and progresses to type 1 DM in less than 1% of that population.3 There is a direct relation to the prevalence of β-cell autoimmunity and the incidence of type 1 DM in various populations. The countries of Sweden, Sardinia, and Finland have the highest prevalence of islet cell antibody (ICA; 3% to 4.5%) and are associated with the highest incidence of type 1 DM: 22 to 35 per 100,000.4 The prevalence of type 1 DM is increasing, but the cause of this increase is not fully understood.

Markers of β-cell autoimmunity are detected in 14% to 33% of persons with adult-onset diabetes. This type of DM is referred to as latent autoimmune diabetes in adults (LADA) and presents with early failure of oral agents and need for insulin therapy.4

Idiopathic type 1 DM is a nonautoimmune form of diabetes frequently seen in minorities, especially Africans and Asians, with intermittent insulin requirements.2

Secondary forms of DM occur due to a variety of causes.2 Maturity onset diabetes of youth (MODY) is due to one of six genetic defects. Endocrine disorders such as acromegaly and Cushing’s syndrome may also cause diabetes. Any disease of the exocrine pancreas such as cystic fibrosis, pancreatitis, and hereditary hemochromatosis can damage β cells and impair insulin secretion. These unusual causes, however, only cause 1% to 2% of the total cases of DM. Please see Other Specific Types of Diabetes (<5% of Diabetes) below for further discussion.

Type 2 DM accounts for up to 90% of all cases of DM. Overall the prevalence of type 2 DM in the United States is about 11.3% in persons aged 20 or older; this prevalence is increasing. It is estimated that for every four persons who are diagnosed with DM, one person remains undiagnosed.1

There are multiple risk factors for the development of type 2 DM, including family history (i.e., parents or siblings with diabetes); obesity (i.e., ≥20% over ideal body weight, or body mass index [BMI] ≥25 kg/m2); chronic physical inactivity; race or ethnicity (see list below); history of impaired glucose tolerance (IGT), impaired fasting glucose (IFG), or hemoglobin A1c (HbA1c) 5.7% to 6.4% (0.057 to 0.064; 39 to 46 mmol/mol Hb) (see Diagnosis of Diabetes below); hypertension (≥140/90 mm Hg in adults); high-density lipoprotein (HDL) cholesterol (HDL-C) ≤35 mg/dL (≤0.91 mmol/L) and/or a triglyceride level ≥250 mg/dL (≥2.83 mmol/L); history of gestational diabetes mellitus (GDM) (see Classification of Diabetes below) or delivery of a baby weighing >9 lb (>4 kg); history of vascular disease; presence of acanthosis nigricans; and polycystic ovary disease.5

Images The prevalence of type 2 DM increases with age and varies widely among various racial and ethnic populations. The prevalence of type 2 DM is especially increased in Native Americans, Hispanic Americans, Asian Americans, African Americans, and Pacific Islanders. While the prevalence of type 2 DM increases with age, the disorder is increasingly being diagnosed in adolescence. Much of the rise in adolescent type 2 DM is related to an increase in overweight/obesity and sedentary lifestyle, in addition to genetic predisposition.6 Most cases of type 2 DM are polygenetic; the underlying pathophysiology remains uncertain7 (Figs. 57-1 and 57-2).


FIGURE 57-1 National Health and Nutrition Evaluation Survey (NHANES) prevalence of diabetes by age among adults ≥20 years of age: United States, 2005–2008. (Centers for Disease Control and Prevention, 2011 National Diabetes Fact Sheet at


FIGURE 57-2 Rate of new cases of type 1 and type 2 diabetes among youth aged <20 years, by race/ethnicity, 2002–2005. (NHW, non-Hispanic whites; NHB, non-Hispanic blacks; H, Hispanics; API, Asians/Pacific Islanders; AI, American Indians.) (Centers for Disease Control and Prevention, 2011 National Diabetes Fact Sheet at

GDM complicates approximately 7% of all pregnancies in the United States.8 Most women become normoglycemic after pregnancy; however, 30% to 50% may develop prediabetes or type 2 DM later in life.


Classification of Diabetes

Diabetes is a metabolic disorder characterized by resistance to the action of insulin, insufficient insulin secretion, or both.2 The clinical manifestation of these disorders is hyperglycemia. The vast majority of diabetic patients are classified into one of two broad categories: type 1 diabetes caused by an absolute deficiency of insulin or type 2 diabetes defined by the presence of insulin resistance with an inadequate compensatory increase in insulin secretion. Women who develop diabetes due to the stress of pregnancy are classified as having gestational diabetes. Finally, uncommon types of diabetes caused by infections, drugs, endocrinopathies, pancreatic destruction, and known genetic defects are classified separately (Table 57-1).

TABLE 57-1 Etiologic Classification of Diabetes Mellitusa


Type 1 Diabetes

Images This form of diabetes results from autoimmune destruction of the β cells of the pancreas. Evidence of β-cell autoimmunity, including ICAs, antibodies to glutamic acid decarboxylase, islet protein tyrosine phosphatase-like molecule IA2, and/or antibodies to insulin, is present at the time of diagnosis in 90% of individuals. Type 1 diabetes is often thought to most commonly present in children and adolescents; however, it can occur at any age. Younger individuals typically have a more rapid rate of β-cell destruction and often present with ketoacidosis. Adults may maintain sufficient insulin secretion to prevent ketoacidosis for many years; this is referred to as latent autoimmune diabetes in adults.3,4

Type 2 Diabetes

Images Type 2 DM is characterized by a combination of some degree of insulin resistance and a relative lack of insulin secretion (being insufficient to normalize plasma glucose levels), with progressively lower insulin secretion over time. Most individuals with type 2 diabetes exhibit abdominal obesity, which itself causes insulin resistance. In addition, hypertension, dyslipidemia (high triglyceride levels and low HDL-C levels), and elevated plasminogen activator inhibitor-1 (PAI-1) levels, which contribute to a hypercoagulable state, are often present in these individuals. Due in part to these factors, patients with type 2 diabetes are at increased risk of developing macrovascular complications in addition to microvascular complications. Type 2 diabetes has a strong genetic predisposition and is more common in all ethnic groups other than those of European ancestry.4,5

Gestational Diabetes Mellitus

GDM is defined as glucose intolerance that is first recognized during pregnancy. Hormone changes during pregnancy result in increased insulin resistance, and GDM may ensue when the mother cannot adequately compensate with increased insulin secretion to maintain normoglycemia. In most, glucose intolerance occurs near the beginning of the third trimester, although risk assessment and intervention when appropriate should begin from the first prenatal visit due to the risk of undiagnosed diabetes. If DM is diagnosed prior to pregnancy, this is not GDM, but rather pregnancy with preexisting DM. Clinical detection is important, as therapy will reduce perinatal morbidity and mortality.2

Other Specific Types of Diabetes (<5% of Diabetes)

Genetic Defects MODY is characterized by impaired insulin secretion in response to a glucose stimulus with minimal or no insulin resistance. Patients typically exhibit mild hyperglycemia at an early age, but diagnosis may be delayed, depending on the severity of presentation. The disease is inherited in an autosomal dominant pattern with at least six different loci identified to date (MODY 2 and 3 are most common). The production of mutant insulin molecules has been identified in a few families and results in mild glucose intolerance.2

Several genetic mutations have been described in the insulin receptor and are associated with insulin resistance. Type A insulin resistance refers to the clinical syndrome of acanthosis nigricans, virilization in women, polycystic ovaries, and hyperinsulinemia. In contrast, anti-insulin receptor antibodies may block the binding of insulin. This was referred to in the past as type B insulin resistance. Endocrinopathies, pancreatic exocrine dysfunction, drugs, and infections, among others, may also result in hyperglycemia (Table 57-1).


Type 1 Diabetes Mellitus

The prevalence of type 1 DM is low in the general population. Due to the acute onset of symptoms in most individuals at time of diagnosis, screening for type 1 DM in the asymptomatic general population is not recommended.5Screening for β-cell autoantibody status in high-risk family members may be appropriate; however, such screening is most often recommended in the context of clinical trials for the prevention of type 1 DM.

Type 2 Diabetes Mellitus

The American Diabetes Association (ADA) recommends screening for type 2 DM at any age in individuals who are overweight (BMI ≥25 kg/m2) and have at least one other risk factor for the development of type 2 DM. Risk factors, in addition to being overweight or obese, include physical inactivity, first-degree relative with diabetes or high-risk ethnicity/race, women who have delivered a baby >9 lb (>4 kg) or a history of GDM, hypertension, high triglycerides, low HDL, women with polycystic ovary syndrome, diagnosed with prediabetes, acanthosis nigricans, or a history of cardiovascular disease (CVD; see also Epidemiology above). The recommended screening test is the fasting plasma glucose (FPG), HbA1c, or 2-hour oral glucose tolerance test (OGTT). Adults without risk factors should be screened starting at age 45 years, as age itself is a risk factor for type 2 DM. The optimal time between screenings is not known, and the index of suspicion for the presence of diabetes should guide the clinician. Repeat testing every 3 to 5 years is cost-effective.5

Children and Adolescents

Despite a lack of clinical evidence to support widespread testing of children for type 2 DM, it is clear that more children and adolescents are developing type 2 DM. The ADA, by expert opinion, recommends that overweight (defined as BMI >85th percentile for age and sex, weight for height >85th percentile, or weight >120% of ideal) youths with at least two of the following risk factors: a family history of type 2 diabetes in first- and second-degree relatives; Native Americans, African Americans, Hispanic Americans, and Asians/South Pacific Islanders; those with signs of insulin resistance or conditions associated with insulin resistance (acanthosis nigricans, hypertension, dyslipidemia, polycystic ovary syndrome, or small-for-gestational-age birth weight); or maternal history of diabetes or GDM during the child’s gestation be screened. Screening should be done every 3 years starting at 10 years of age or at the onset of puberty if it occurs at a younger age.5

Gestational Diabetes

Risk assessment for GDM should occur at the first prenatal visit. Due to the increasing incidence of obesity and undiagnosed DM, it is reasonable to screen women with risk factors for the development of diabetes as soon as feasible. If the initial screening is negative, they should undergo retesting at 24 to 28 weeks’ gestation. Screening for GDM is done with a standard 75-g OGTT. The diagnosis of GDM is confirmed when any one plasma glucose value measured at baseline (fasting), 1 hour, or 2 hours meets the diagnostic criteria. These criteria are unique to GDM (Table 57-2).2,5,8

TABLE 57-2 Screening for and Diagnosis of Gestational Diabetes Mellitus with a 75-g Glucose Load2


Diagnosis of Diabetes

Images The diagnosis of diabetes requires the identification of a glycemic cut point, which discriminates normals from diabetic patients. The cut points are meant to reflect the level of glucose above which microvascular complications have been shown to increase. Cross-sectional studies have shown a consistent increase in the risk of developing retinopathy at a fasting glucose level above 99 to 116 mg/dL (5.5 to 6.4 mmol/L), a 2-hour postprandial level above 125 to 185 mg/dL (6.9 to 10.3 mmol/L), and a HbA1c above 5.9% to 6.0%. (0.059 to 0.060; 41 to 42 mmol/mol Hb). Current diagnostic criteria are slightly above these cut points (Table 57-3).2

TABLE 57-3 Criteria for the Diagnosis of Diabetes Mellitusa


The HbA1c was not recommended in the past due to many nonstandardized assays. Most laboratories now use a method that is National Glycohemoglobin Standardization Program (NGSP) certified and standardized to the Diabetes Control and Complications Trial (DCCT) assay, which allows for cross-application of their results. If standardized, the HbA1c is logical for the diagnosis of diabetes as it measures glycemic exposure over the past 2 to 3 months, in contrast to a single-day, single-point glucose evaluation. In addition, patients do not have to be fasting and the test is easily monitored. An HbA1c of 6% to 6.4% (0.060 to 0.066; 42 to 46 mmol/mol Hb) denotes a 10-fold increase in risk of diabetes, yet does not consistently identify patients with IFG or IGT. In addition, there are slight race differences in normal HbA1c levels. One-third fewer individuals with diabetes are identified using the A1C ≥6.5% (≥0.065; ≥48 mmol/mol Hb) versus a FPG ≥126 mg/dL (≥7 mmol/L), yet more providers may be more likely to diagnose diabetes from an A1C than from an obviously elevated FPG level. The ADA continues to recommend three other glucose criteria for the diagnosis of DM in nonpregnant adults (Table 57-3). If the patient has obvious hyperglycemia and diabetes, reconfirming the diagnosis by one of the above criteria is not required.2

Increased Risk of Diabetes or Prediabetes

As shown in Table 57-4, the ADA identified a HbA1c value of 5.7% to 6.4% (0.057 to 0.064; 39 to 46 mmol/mol Hb) to define an increased risk for diabetes. The HbA1c lower limit of 5.7% (0.057; 39 mmol/mol Hb) was chosen due to its good specificity, although it has a low sensitivity, to identify patients at increased risk for diabetes. IFG continues to be defined as a plasma glucose of at least 100 mg/dL (5.6 mmol/L) but less than 126 mg/dL (7 mmol/L). IGT is defined as a 2-hour glucose value ≥140 mg/dL (≥7.8 mmol/L), but less than 200 mg/dL (11.1 mmol/L) during a 75-g OGTT.2,5

TABLE 57-4 Categorizations of Abnormal Glucose Status


Serial measurements, at clinician-defined intervals, can help to identify patients moving toward diabetes, and those who are stable. Patients who have even minor increases in glucose or HbA1c values over time should be followed closely. Also, the HbA1c measurement can be affected by anemias and several hemoglobinopathies, which necessitates the use of one of the plasma glucose criterion in these individuals.


Type 1 Diabetes Mellitus

Type 1 DM results from pancreatic β-cell failure with “absolute” deficiency of insulin secretion. Most often this is due to immune-mediated destruction of pancreatic β cells, but rare unknown or idiopathic processes may also contribute. There often is a long preclinical period of immune-mediated β-cell destruction later followed by onset of hyperglycemia when 80% to 90% of the β cells have been destroyed. Occasionally there is a period of transient remission called the “honeymoon” phase, before established disease develops along with the requirement for lifelong insulin therapy and the potential risk of diabetes-related complications (Fig. 57-3).


FIGURE 57-3 Scheme of the natural history of the β-cell defect in type 1 diabetes mellitus. (Copyright © 2008 American Diabetes Association. From Medical Management of Type 1 Diabetes, 5th ed. Reprinted with permission from the American Diabetes Association.)

It is thought that in order for type 1 DM to develop, there must be a trigger in a genetically susceptible individual. However, it is unknown whether there are one or more inciting factors such as cow’s milk (or lack of breast-feeding), or viral, dietary, or other environmental exposures that initiate the autoimmune process.2,3 Vitamin D deficiency has been observed to be more prevalent in patients who develop type 1 DM; however, further study is needed to confirm a role in causation.9

The autoimmune process is mediated by macrophages and T lymphocytes with circulating autoantibodies to various β-cell antigens. The most commonly detected antibody associated with type 1 DM is the ICA. Other autoantibodies include insulin, glutamic acid decarboxylase 65, and zinc transporter 8 (ZnT8). These antibodies are generally considered markers of disease rather than mediators of β-cell destruction. They have been used to identify individuals at risk for type 1 DM and in evaluating disease prevention strategies.3

More than 90% of newly diagnosed persons with type 1 DM have one of these antibodies, as will up to 4% of unaffected first-degree relatives. Once insulin autoantibodies are detected, there is an increased risk of development of additional autoantibodies and progression to diabetes. β-Cell autoimmunity may precede the diagnosis of type 1 DM by up to 9 to 13 years. Autoimmunity may remit in some individuals, or can progress to absolute β-cell failure in others. Other autoimmune disorders frequently associated with type 1 DM include Hashimoto’s thyroiditis, Graves’ disease, Addison’s disease, vitiligo, and celiac sprue. The extent of involvement can range from no other associated disorders to autoimmune polyglandular failure.

There are strong genetic linkages to the DQA and B genes and certain human leukocyte antigens (HLAs). Some are associated with increased risk (DR3 and DR4) while others are protective (DRB1*04008-DQB1*0302 and DRB1*0411-DQB1*0302) on chromosome 6.10 Additional candidate gene regions have been identified on other chromosomes as well. Because twin studies do not show 100% concordance, environmental factors such as infectious, chemical, and dietary agents likely also contribute to the expression of the disease.

The autoimmune destruction of pancreatic β-cell function results in hyperglycemia due to an absolute deficiency of insulin. Insulin lowers blood glucose (BG) by a variety of mechanisms, including stimulation of tissue glucose uptake, suppression of glucose production by the liver, and suppression of free fatty acid (FFA) release from fat cells.11 The suppression of FFAs plays an important role in glucose homeostasis. Increased levels of FFAs inhibit the uptake of glucose by muscle and stimulate hepatic gluconeogenesis.12

Type 2 Diabetes Mellitus

Normal Metabolism In the fasting state 75% of total body glucose disposal takes place in non–insulin-dependent tissues such as the brain, neurons, and others. Brain glucose uptake occurs at the same rate during fed and fasting periods. The remaining 25% of glucose metabolism takes place in the liver and muscle, which is dependent on insulin. In the fasting state approximately 85% of glucose production is derived from the liver, and the remaining amount is produced by the kidney. Glucagon, produced by pancreatic α cells, is secreted in the fasting state to oppose the action of insulin and stimulate hepatic glucose production and glycogenolysis. Glucagon and insulin secretion are closely linked; one increases while the other decreases to keep plasma glucose levels normal. In the fed state, carbohydrate ingestion increases the plasma glucose concentration and stimulates insulin release from the pancreatic βcells. The resultant hyperinsulinemia (a) suppresses hepatic glucose production, (b) stimulates glucose uptake by peripheral tissues, and (c) suppresses glucagon release (in conjunction with incretin hormones). The majority (~80% to 85%) of glucose is taken up by muscle, with only a small amount (~4% to 5%) being metabolized by adipocytes.7,13,14

Although fat tissue is responsible for only a small amount of total body glucose disposal, it plays a very important role in the maintenance of total body glucose homeostasis. Small increments in the plasma insulin concentration exert a potent antilipolytic effect, leading to a marked reduction in the plasma FFA levels. The decline in plasma FFA concentrations results in an increased glucose uptake in muscle and reduces hepatic glucose production indirectly.

Type 2 Diabetes Individuals are characterized by multiple defects including (a) defects in insulin secretion; (b) insulin resistance involving muscle, liver, and the adipocyte; (c) excess glucagon secretion; (d) glucagon-like peptide-1 (GLP-1) deficiency and possibly resistance.7

Impaired Insulin Secretion The pancreas in people with a normal-functioning β cell is able to adjus its secretion of insulin to maintain normal plasma glucose levels. In nondiabetic individuals, insulin increases in proportion to the severity of the insulin resistance and plasma glucose remains normal. Impaired insulin secretion is a hallmark finding in T2DM. In early β-cell dysfunction, first-phase insulin release, seen with an IV bolus of glucose, is deficient. First-phase insulin is released if there is stored insulin in the β cell and acts to “prime” the liver to nutrient intake. Absent first-phase insulin necessitates an increase in second-phase insulin to compensate for hyperglycemia. When the insulin released can no longer normalize plasma glucose, dysglycemia, including prediabetes and diabetes, can ensue. Both β-cell mass and function in the pancreas are reduced. β-Cell failure is progressive, and starts years prior to the diagnosis of diabetes. People with T2DM lose ~5% to 7% of β-cell function per year of diabetes. The reasons for this loss are likely multifactorial including (a) glucose toxicity; (b) lipotoxicity; (c) insulin resistance; (d) age; (e) genetics; and (f) incretin deficiency. Age results in declining β-cell responsiveness and possibly mass. β-Cell failure predisposition is also present in high-risk ethnicity/races. Glucotoxicity involves glucose levels chronically exceeding 140 mg/dL (7.8 mmol/L). The β cell is unable to maintain elevated rates of insulin secretion, and releases less insulin as glucose levels increase (Fig. 57-4).7,13,14


FIGURE 57-4 The relationship between fasting plasma insulin and fasting plasma glucose in 177 normal-weight individuals. Plasma insulin and glucose increase together up to a fasting glucose of 140 mg/dL (7.8 mmol/L). When the fasting glucose exceeds 140 mg/dL (7.8 mmol/L), the β cell makes progressively less insulin, which leads to an overproduction of glucose by the liver and results in a progressive increase in fasting glucose. (Reprinted from DeFronzo RA. Pathogenesis of type 2 diabetes mellitus. Med Clin N Am 2004;88:787–835, Copyright © 2004, with permission from Elsevier.)

Incretins In the type 2 diabetic patient, decreased postprandial insulin secretion is due to both impaired pancreatic β-cell function and a reduced stimulus for insulin secretion from gut hormones. The role gut hormones play in insulin secretion is best shown by comparing the insulin response to an oral glucose load versus an isoglycemic IV glucose infusion. In nondiabetic control individuals 73% more insulin is released in response to an oral glucose load compared with reproducing the oral glucose load’s plasma glucose curve by giving IV glucose. This increased insulin secretion in response to an oral glucose stimulus is referred to as “the incretin effect” and suggests that gut-derived hormones when stimulated by glucose lead to an increase in pancreatic insulin secretion. In type 2 diabetic patients, this “incretin effect” is blunted, with the increase in insulin secretion only being 50% of that seen in nondiabetic control individuals. It is now known that two hormones, GLP-1 and glucose-dependent insulinotropic polypeptide (GIP), are responsible for over 90% of the increased insulin secretion seen in response to an oral glucose load. Patients with type 2 diabetes remain sensitive to GLP-1 while GIP levels are normal or elevated in T2DM.7

GLP-1 is secreted from the L cells, with the highest L-cell concentration in the distal intestinal mucosa, in response to mixed meals. Since GLP-1 levels rise within minutes of food ingestion, neural signals and possibly proximal GI tract receptors stimulate GLP-1 secretion. The insulinotropic action of GLP-1 is glucose dependent, and for GLP-1 to enhance insulin secretion, glucose concentrations must be higher than 90 mg/dL (5 mmol/L). In addition to stimulating insulin secretion, GLP-1 suppresses glucagon secretion, slows gastric emptying, and reduces food intake by increasing satiety. These effects of GLP-1 combine to limit postprandial glucose excursions. GIP is secreted by K cells in the intestine and may have a role with insulin secretion during near-normal glucose levels and may act as an insulin sensitizer in adipocytes. However, GIP has no effect on glucagon secretion, gastric motility, or satiety. The half-lives of GLP-1 and GIP are short (<10 minutes). Both hormones are rapidly inactivated by removal of two N-terminal amino acids by the enzyme dipeptidyl peptidase-4 (DPP-4). GLP-1 levels appear to decrease as glucose values increase from normal to type 2 DM, and it is unlikely to be a primary defect that causes diabetes in the majority of T2DM. Genetically a minority may have the TCF7L2 gene defect, which is associated with a decreased response to GLP-1.7

Insulin Resistance

Liver In type 2 diabetic subjects with mild to moderate fasting hyperglycemia (140 to 200 mg/dL, 7.8 to 11.1 mmol/L), basal hepatic glucose production is increased by ~0.5 mg/kg/min. Consequently, during the overnight sleeping hours the liver of an 80-kg diabetic individual with modest fasting hyperglycemia adds an additional 35 g of glucose to the systemic circulation. This increase in fasting hepatic glucose production is the cause of fasting hyperglycemia.13,14

Following glucose ingestion, insulin is secreted into the portal vein and carried to the liver, where it reduces hepatic glucose output. T2DM patients also fail to suppress glucagon in response to a meal and may even have a paradoxical rise in glucagon levels. Thus, hepatic insulin resistance and hyperglucagonemia result in continued production of glucose by the liver. Therefore, T2DM patients have two sources of glucose in the postprandial state: one from the diet and one from continued glucose production from the liver. These sources of glucose may result in marked hyperglycemia.

Peripheral (Muscle) Muscle is the major site of postprandial glucose disposal in humans, and approximately 80% of total body glucose uptake occurs in skeletal muscle. In response to a physiologic increase in plasma insulin concentration, muscle glucose uptake increases linearly, reaching a plateau value of 10 mg/kg/min. Even in lean T2DM, the onset of insulin action is delayed for ~40 minutes, and the ability of insulin to stimulate leg glucose uptake is reduced by 50%. Impaired intracellular insulin signaling is a well-established abnormality, with notable impairments at almost every step of activation due to insulin resistance, lipotoxicity, and glucotoxicity. The compensatory hyperinsulinemia required to overcome impaired insulin signaling (insulin resistance) can activate an alternative pathway through MAP kinase, which may be involved in atherosclerosis. Mitochondrial dysfunction may also play a role in muscle insulin resistance. Mitochondrial function and/or density appear to be lower in type 2 DM. This may result in less energy expenditure and an increased risk of dysfunction with high-fat diets (Fig. 57-5).13,14


FIGURE 57-5 Whole-body glucose disposal, a measure of insulin resistance, is reduced 40% to 50% in obese nondiabetic and lean type 2 diabetic individuals. Obese diabetic individuals are slightly more resistant than lean diabetic patients. (From DeFronzo RA. Diabetes Reviews 1977;5:177–269.)

Peripheral (Adipocyte) In obese nondiabetic and T2DM, basal plasma FFA levels are increased and fail to suppress normally after glucose ingestion. FFAs are stored as triglycerides in adipocytes and serve as an important energy source during conditions of fasting. Insulin is a potent inhibitor of lipolysis, and restrains the release of FFAs from the adipocyte by inhibiting the hormone-sensitive lipase enzyme. It is now recognized that chronically elevated plasma FFA concentrations can lead to insulin resistance in muscle and liver, and impair insulin secretion. In addition to FFAs that circulate in plasma in increased amounts, T2DM patients have increased stores of intracellular fat products in muscle and liver, and the increased fat content correlates closely with the presence of insulin resistance in these tissues. Excess lipolysis from fat can also contribute to gluconeogenesis indirectly through glycerol and FFAs.7,13,14

Cellular Mechanisms of Insulin Resistance

Obesity and Insulin Resistance Weight gain leads to insulin resistance in most, and obese nondiabetic individuals with risk factors often have the same degree of insulin resistance as lean T2DM patients. Subsets of obese, but metabolically normal patients (6% to 30%) do exist, as well as nonobese, but metabolically abnormal patients, so broad categorization of risk for a patient needs to be confirmed by further examination.

The term visceral adipose tissue (VAT) refers to fat cells located within the abdominal cavity and includes omental, mesenteric, retroperitoneal, and perinephric adipose tissue. VAT has been shown to correlate with insulin resistance and explain much of the variation in insulin resistance seen. It represents 20% of fat in men and 6% of fat in women. Central obesity can be easily assessed using waist circumference, which is a good surrogate marker for VAT. This fat tissue has been shown to have a higher rate of lipolysis than subcutaneous fat, resulting in an increase in FFA production. These fatty acids are released into the portal circulation and drain into the liver, where they stimulate the production of very-low-density lipoproteins and decrease insulin sensitivity in peripheral tissues.13,14

VAT also produces a number of adipocytokines, such as TNF-α, interleukin 6, angiotensinogen, PAI-1, and resistin, which contribute to insulin resistance, hypertension, and hypercoagulability. These factors drain into the portal circulation and reduce insulin sensitivity in peripheral tissues. The fat cell also has the capability of producing at least one adipocytokine that improves insulin sensitivity: adiponectin. This factor is made in decreasing amounts as an individual becomes more obese. In animal models, adiponectin decreases hepatic glucose production and increases fatty acid oxidation in muscle.

The Metabolic Syndrome The metabolic syndrome is a risk indicator, but not an absolute risk indicator, because it does not specifically account for all risk factors, such as age, sex, and low-density lipoprotein cholesterol (LDL-C) levels, or directly measure hypercoagulability of the proinflammatory condition. Patients with metabolic syndrome do have a higher risk for CVD, and at least a fivefold increase in their risk of type 2 DM, if they do not already have type 2 DM. The metabolic syndrome does not identify synergism among identified risk factors, but rather additive risk, leading many to question its relevance above adequate risk factor identification and aggressive treatment. It may be useful to certain clinicians to “package” risk factors into the metabolic syndrome to encourage aggressive management.

The most recent definition of metabolic syndrome was adopted by multiple organizations in 2009 (Table 57-5).15,16

TABLE 57-5 Defining the Metabolic Syndrome


Clinical Controversy…

Metabolic Syndrome: Fact or Fiction?

The term metabolic syndrome was first coined in the late 1970s and associated with CVD. This disease has been studied extensively and also referred to as the “insulin resistance syndrome,” “dysmetabolic syndrome,” and “syndrome X.” In 2001, the National Cholesterol Education Program (NCEP) Adult Treatment Panel (ATP) III formally defined the metabolic syndrome as a clustering of risk factors that include at least three of the following: elevated blood pressure, abdominal obesity, elevated triglycerides, low high-density lipoproteins, and elevated BG.15 Since the metabolic syndrome was formally defined, several other organizations including the International Diabetes Federation (IDF), American Heart Association (AHA), National Heart, Lung, and Blood Institute (NHLBI), and the World Heart Federation, among others, have released statements to further refine the proposed definition.16

It is estimated that 34% of adults in the United States have the metabolic syndrome17 and are considered to be at high risk of developing CVD and DM.18 This epidemic has also been described in children and adolescents19 and is expected to expand as the US obesity rates continue to climb.

However, in 2005 the ADA in conjunction with the European Association for the Study of Diabetes released a joint statement critical of the clinical utility of the metabolic syndrome.20 They concluded in their statement that there is no doubt that CVD risk factors cluster in certain individuals, although they assert the definition is imprecise and its use as a CVD marker is questionable. They also state that there is critical information missing to warrant its designation as a syndrome. A swift rebuttal from the AHA and NHLBI was released weeks later encouraging the continued use of the metabolic syndrome concept.21In their statement, the authors stressed that the metabolic syndrome is not considered a singular entity and that it is a syndrome with no single pathogenesis. The authors of the statement also argue that the distinction of the metabolic syndrome should allow clinicians to approach patients as a whole with an emphasis on intensive lifestyle management. Those in the diabetes community and who authored the ADA/EASD statement contend that there is no additional benefit from identifying these clusters of CVD risk factors over measuring and treating them individually. They maintain the lack of predictive capabilities limits the metabolic syndrome’s utility as a CVD marker.

The use of the metabolic syndrome in clinical practice remains a controversy. Clinicians should always take care to individualize therapy for patients who present with high CVD risk. Patient’s individual goals, values, and resources should be considered when tailoring a treatment plan.


The clinical presentations of type 1 DM and type 2 DM are very different. Autoimmune type 1 DM can occur at any age. Approximately 75% will develop the disorder before age 20 years, but the remaining 25% develop the disease as adults. Individuals with type 1 DM are often thin and are prone to develop diabetic ketoacidosis (DKA) if insulin is withheld, or under conditions of severe stress with an excess of insulin counterregulatory hormones.2,3,5Symptoms in patients with type 1 DM such as polyuria, polydipsia, polyphagia, weight loss, and lethargy accompanied by hyperglycemia are the most common initial presentation. In the outpatient setting, many patients initially present with vague complaints such as weight loss and fatigue. Polyuria, polydipsia, and polyphagia may not be apparent unless a comprehensive history is taken. Twenty percent to 40% of patients with type 1 DM present with DKA after several days of polyuria, polydipsia, polyphagia, and weight loss. This presentation is common in patients from a low socioeconomic background. Rarely, type 1 DM patients are diagnosed without multiple symptoms or DKA when they have blood tests drawn for other reasons. This rare presentation typically occurs when patients have a first-degree family member with type 1 DM and are closely monitored.



Patients with type 2 DM often present without symptoms, even though complications tell us that they may have been hyperglycemic for several years.10 Often these patients are diagnosed secondary to unrelated blood testing. Lethargy, polyuria, nocturia, and polydipsia can be seen at diagnosis in type 2 diabetes, but significant weight loss at diagnosis is less common. More often, patients with type 2 DM are overweight or obese. Clinically, DM is a spectrum of diseases ranging from absolute insulin deficiency to relative insulin deficiency, and patients can have normal to grossly abnormal insulin sensitivity. Classical clinical presentation characteristics should be used in conjunction with other definitive laboratory data to properly classify patients (see also Classical Clinical Presentation of Diabetes Mellitus below).


Diabetes Mellitus

Desired Outcome

Images The primary goals of DM management are to reduce the risk for microvascular and macrovascular disease complications, to ameliorate symptoms, to reduce mortality, and to improve quality of life.5Early treatment with near-normal glycemia will reduce the risk for development of microvascular disease complications, but aggressive management of traditional cardiovascular risk factors (i.e., smoking cessation, treatment of dyslipidemia, intensive blood pressure control, and antiplatelet therapy) is needed to reduce the likelihood of development of macrovascular disease. Hyperglycemia not only increases the risk for microvascular disease but also contributes to poor wound healing, compromises white blood cell function, alters capillary function, and leads to classic symptoms of DM. DKA and hyperosmolar hyperglycemic state (HHS) are severe manifestations of poor diabetes control, almost always requiring hospitalization. Reducing the potential for microvascular complications is targeted by adherence to therapeutic lifestyle intervention (i.e., diet and exercise programs) and drug therapy regimens, as well as attaining blood pressure goals. Minimizing weight gain and hypoglycemia, especially severe hypoglycemia, and altering the glycemic goal to match the patient’s morbidities are necessary. Evidence-based guidelines, as published by the ADA, may help in the attainment of these goals (Table 57-6).5

TABLE 57-6 Selected American Diabetes Association Evidence-Based Recommendationsa



General Approach to Treatment

Appropriate care requires goal setting for glycemia, blood pressure, lipid levels (goals described later in chapter; see also Chaps. 3 and 11), regular monitoring for complications, dietary and exercise modifications, medications, appropriate self-monitoring of blood glucose (SMBG), and laboratory assessment of the aforementioned parameters.5 Glucose control alone does not sufficiently reduce the risk of macrovascular complications in persons with DM.22

Glycemic Goal Setting and Hemoglobin A1c

Controlled clinical trials provide ample evidence that glycemic control is paramount in reducing microvascular complications in both type 1 DM23 and type 2 DM.24 HbA1c measurements are the gold standard for following long-term glycemic control for the previous 2 to 3 months.5 Hemoglobinopathies, anemia, red cell membrane defects, transfusions, and substantial increase or decrease of red blood cell life span in a patient can affect HbA1cmeasurements. Identification of potential problems and then ensuring the test is performed in a laboratory using a method that is NGSP certified and standardized to the DCCT assay (see will minimize issues. Other strategies such as measurement of fructosamine, which measures glycated plasma proteins or glycated albumin, may be necessary to assess diabetes control in patients with altered red blood cell life span, although they are less standardized, and not correlated to risk of complications.

The A1C-Derived Average Glucose study correlated multiple HbA1c and glucose readings to term the phrase estimated average glucose (eAG). The eAG better correlates with HbA1c readings, and now is regularly reported below HbA1c values on laboratory results. For example, a HbA1c of 6% or 7% (0.060 to 0.070; 42 to 53 mmol/mol Hb) correlates with an average glucose of 126 or 154 mg/dL (7 or 8.5 mmol/L), respectively, and online calculators and graphs are easily found.25

Less stringent HbA1c goals (>7% [>0.070; >53 mmol/mol Hb]) may be appropriate in patients with a history of severe hypoglycemia, limited life expectancy, advanced microvascular/macrovascular complications or comorbidities, at-risk elderly, dementia, or in younger children. A HbA1c target of <7% (<0.070; <0.53 mmol/mol Hb) is appropriate for others (Table 57-7), and lower values should be targeted if significant hypoglycemia, weight gain, and other adverse effects can be avoided.5 Glycemic control recommendations for different age groups of type 1 DM patients are based on the risk of hypoglycemia, the relatively low risk of complications prior to puberty, and psychological and/or developmental issues (Table 57-7).

TABLE 57-7 Glycemic Goals of Therapy by Organization8


Initial Evaluation of Diabetes Mellitus

On initial evaluation, a thorough medical history and identification of specific type of diabetes, including duration of diabetes, characteristics of onset (e.g., DKA or asymptomatic), dietary and weight history, education history, medication history including current and past medications for DM, current regimen including medications, diet, physical activity, and adherence, should be performed. Hospitalization history, hypoglycemia (frequency, cause, timing), and diabetes-related complications should be documented. Laboratory evaluation should include at a minimum an A1C, lipid profile, liver function tests, thyroid-stimulating hormone level, serum creatinine and electrolytes, and a urine analysis for microalbuminuria. In type 1 DM, consider screening for celiac disease by measuring tissue transglutaminase or antiendomysial antibodies. The physical examination and pertinent data should include all vital signs, weight and/or BMI, blood pressure assessment, thyroid palpation, cardiovascular and carotid auscultation, skin integrity, assessment for acanthosis nigricans, and a foot examination, including screening for impaired sensation detection with a 10-g force monofilament.5

Monitoring for Complications

The ADA recommends initiation of complications monitoring at the time of diagnosis of DM.8 Current recommendations continue to advocate yearly dilated eye examinations in type 2 DM, and an initial dilated eye examination in the first 3 to 5 years in type 1 DM, and then yearly thereafter. Less frequent testing (every 2 to 3 years) can be implemented on the advice of an eye care specialist. The blood pressure should be assessed at each visit. The feet should be examined at each visit for distal pulses, skin integrity, calluses, and deformities, and yearly screening should be done for loss of protective sensation with a distal polyneuropathy tool, such as the 10-g force Semmes-Weinstein monofilament. A urine test for microalbumin to screen for nephropathy once yearly from diagnosis is appropriate in type 2 DM, and initiated 5 years after diagnosis if the patient has type 1 DM. Yearly testing for lipid abnormalities, and more frequently if needed to achieve lipid goals, is recommended. It is generally accepted that a yearly thyroid-stimulating hormone level may be appropriate in type 1 DM, LADA, and select type 2 DM patients.5

Self-Monitored Blood Glucose and Continuous Glucose Monitoring

The advent of SMBG in the early 1980s revolutionized the treatment of DM, enabling patients to know their BG concentration at any moment easily and relatively inexpensively. At its core, SMBG is a tool to provide structure for a change and/or safety: change, in that the patient has an opportunity to intervene when a SMBG value is obtained, and safety, as hypoglycemia and hyperglycemia need to be avoided and/or identified and treated. In general, SMBG frequency should match how complicated the regimen is for glycemic control and minimally allow testing to avoid hypoglycemia.

Frequent SMBG is necessary to achieve near-normal BG concentrations if hypoglycemic agents are used. Assessment for hypoglycemia and hyperglycemia, adjustment of prandial doses of insulin, administration of corrective doses of insulin, change in diet and exercise, and checking accuracy of continuous glucose monitors are but a few of the reasons a patient may need SMBG at a given time. This is particularly true in patients with type 1 DM, as most will be intensively managed with insulin. The more intense the pharmacologic regimen is, the more intense the SMBG needs to be (before meals, at bedtime, occasionally after meals, and middle of sleep cycle in patients on multiple insulin injections or pump therapy whether type 1 or type 2 DM). The optimal frequency of SMBG for patients with type 2 DM on oral agents is unresolved.26 Frequency of monitoring in type 2 DM should be sufficient to facilitate reaching glucose goals and to test for hypoglycemia. The role of SMBG in improving glycemic control in type 2 DM patients is controversial, but has shown to reduce the HbA1c ~0.4% (~0.004; ~4 mmol/mol Hb) to no improvement. What is clear is that patients must be empowered to change their therapeutic regimen (lifestyle and medications) in response to test results, or no meaningful glycemic improvement is likely to be effected.5

Alternate site testing may improve adherence to SMBG recommendations, but only SMBG meters that can “sip” blood onto the strip will accommodate such testing. Alternate site glucose testing is performed on the palm, forearm, or the thigh. These areas tend to have less nerve endings and may be more comfortable for a patient, but several cautions must be observed. Interstitial glucose readings identified with alternative site testing will lag behind fingertip capillary blood, as the capillary flow/density is often less in the alternate testing sites when compared with that in the fingertip. Alternate site testing is discouraged in any situation where immediate action will be needed based on the glucose reading, such as testing for hypoglycemia or in patients with hypoglycemia unawareness, wide fluctuations in SMBG, or when the BG is known to be fluctuating, such as postprandially.

Choosing a meter for your patient depends most importantly on his or her dexterity, eye acuity, strip cost, and features that may be important to him or her. Demonstrate to and then have the patient confirm the monitoring technique to minimize problems. Each meter has specifications on hematocrit, elevation, whole blood versus plasma, and heat/cold tolerance. In addition, acetaminophen, ascorbate, dopamine, mannitol, and sugar-based products may alter testing results. Consult the manufacturer materials for specifics.

Continuous glucose monitoring (CGM) may be useful in select patients. CGM measures interstitial glucose, which lags behind capillary SMBG, and the same cautions as alternate site testing should be followed. CGM can be useful in patients with frequent hypoglycemia or hypoglycemic unawareness, nocturnal hypoglycemia, and for identification of fluctuating glucose patterns and/or previously unknown problems in patients with higher or lower than expected HbA1c results. CGM still needs to be calibrated after insertion of a new sensor and minimally every 12 hours with SMBG readings, alarms need to be properly set, and a new sensor must be placed every 3 to 7 days. The ADA currently recommends that CGM can be considered in type 1 DM adults ≥25 years of age, and those <25 years of age, if adherent to its use, and in others with the above issues noted.5

Nonpharmacologic Therapy


Medical nutrition therapy is recommended for all persons with DM and, along with activity, is a cornerstone of treatment.5 Paramount for all medical nutrition therapy is the attainment of optimal metabolic outcomes and the prevention and treatment of complications. It is imperative that patients understand the connection between carbohydrate intake, medications, and glucose control. For individuals with type 1 DM, the focus is on physiologically regulating insulin administration with a balanced diet to achieve and maintain a healthy body weight. A healthy meal plan that is moderate in carbohydrates and low in saturated fat (<7% of total calories), with a focus on balanced meals delivering all of the essential vitamins and minerals, is recommended in DM. The amount (grams) and type (via the glycemic index, though controversial) of carbohydrates, whether accounted for by exchanges or carbohydrate counting, should be considered. All foods can be fit into a healthy meal plan, and the days of recommending no sweets are in the past. If a healthy weight and normal glucose goals can be maintained, there is no reason to deny food choices. Overweight/obese patients with type 2 DM often require caloric restriction to promote weight loss, and portion size and frequency are often issues. The specific diet appears to be less important than if the patient will adhere to the diet, although low-fat diet for CVD or avoiding a high-protein diet in nephropathy may be appropriate. Rather than a set diabetic diet, advocate a diet using foods that are within the financial reach and cultural milieu of the patient. Discourage bedtime and between-meal snacks, set realistic goals for changes based on what the patient can/will change, and follow up to see how and if those changes occurred.27


In general, most patients with DM can benefit from increased activity.28 Aerobic exercise improves insulin sensitivity and modestly improves glycemic control in the majority of individuals, and reduces cardiovascular risk factors, contributes to weight loss or maintenance, and improves well-being. The patient should choose an activity that he or she is likely to continue. Start exercise slowly in previously sedentary patients. It remains unclear which asymptomatic patients should be screened for CVD prior to the beginning of an exercise regimen. Patients with long-standing disease (age >35 years, or >25 years old with DM ≥10 years), patients with multiple cardiovascular risk factors, presence of microvascular disease (especially renal disease), and patients with previous evidence of atherosclerotic disease should have a cardiovascular evaluation, probably including an electrocardiogram, with further workup related to CVD risk. In addition, several complications (uncontrolled hypertension, autonomic neuropathy, insensate feet, and retinopathy) may require restrictions on the activities recommended. Physical activity goals include at least 150 min/wk of moderate (50% to 70% maximal heart rate) intensity exercise spread over at least 3 days a week with no more than 2 days between activity. In addition, resistance/strength training, in patients without retinal contraindications, is recommended to be added into this exercise regimen at least two times a week.5

Pharmacologic Therapy

From the late 1970s to 1995, only two options for pharmacologic treatment were available for patients with diabetes: sulfonylureas (for type 2 DM only) and insulin (for type 1 or 2). Since 1995, a number of new oral agents, injectables, and insulins have been introduced in the United States.

The Look Action for Health in Diabetes (Look AHEAD) trial recently reported that no decrease in cardiovascular outcomes from intensive lifestyle changes in type 2 DM subjects was noted after 10 years of follow-up. In addition, intensive lifestyle was not able to obtain intensive glycemic control in the majority of subjects, reiterating the need for early diabetes medication use in conjunction with diet and exercise interventions.

Currently, nine classes of oral agents are approved for the treatment of type 2 diabetes: α-glucosidase inhibitors, biguanides, meglitinides, peroxisome proliferator–activated receptor γ (PPAR-γ) agonists (which are also commonly identified as thiazolidinediones [TZDs] or glitazones), DPP-4 inhibitors, dopamine agonists, bile acid sequestrants, sodium-glucose cotransporter 2 inhibitors, and sulfonylureas. Oral antidiabetic agents are often grouped according to their glucose-lowering mechanism of action. Biguanides and TZDs are often categorized as insulin sensitizers due to their ability to reduce insulin resistance. Sulfonylureas and meglitinides are often categorized as insulin secretagogues because they enhance endogenous insulin release. Three injectable classes, including insulin, GLP-1 receptor agonists, and amylinomimetics, are also available.

Drug Class Information


Pharmacology Insulin is an anabolic and anticatabolic hormone. It plays major roles in protein, carbohydrate, and fat metabolism. Endogenously produced insulin is cleaved from the larger proinsulin peptide in the β cell to the active peptide of insulin and inactive C-peptide. All commercially available insulin preparations contain only the active insulin peptide.

Characteristics Characteristics that are commonly used to categorize insulin preparations include source, strength, onset, and duration of action. Additionally, insulin may be characterized as analog, defined as insulin preparations that had amino acids within the insulin molecule modified and/or “modifiers” added to impart particular physiochemical and pharmacokinetic advantages. Table 57-8 summarizes available insulin preparations.

TABLE 57-8 Available Injectable and Insulin Preparations


U-100 and U-500, 100 and 500 units/mL, respectively, are the strengths of injectable insulin currently available in the United States. U-500 regular insulin is available for individuals who may require large doses of insulin to control their diabetes. In the United States, all other insulin preparations are available only in U-100 strength. For some patients with type 1 diabetes who require extremely low doses of insulin, dilution of U-100 insulin to obtain accurate insulin doses may be necessary. Diluents, instructions on dilution, and empty bottles can be obtained from the manufacturers for dilution.

Historically, insulin came from either beef or pork sources. Manufacturers in the United States have discontinued production of beef and pork source insulin preparations as of December 2003, and now exclusively use recombinant DNA technology to manufacture insulin. Eli Lilly and Sanofi-Aventis currently use a non–disease-producing strain of Escherichia coli for synthesis of insulin, whereas Novo Nordisk uses Saccharomyces cerevisiae, or bakers’ yeast, for synthesis.

Purity of insulin refers to the amount of proinsulin and other impurities present in a given insulin product. Prior to 1980, most insulin contained enough impurities (300 to 10,000 ppm) to cause local reactions on injection, as well as systemic adverse effects from antibody production. Modern technology has provided less expensive techniques to purify insulin. As a result, all insulin products contain ≤10 ppm of proinsulin, with purified preparations (all recombinant DNA human insulin and insulin analogs) containing <1 ppm of proinsulin.

Regular crystalline insulin naturally self-associates into a hexameric (six insulin molecules) structure when injected subcutaneously. Before absorption through a blood capillary can occur, the hexamer must dissociate first to dimers, and then to monomers. This principle is the premise for additives such as protamine and zinc described below, and modification of amino acids for insulin analogs. Lispro, aspart, and glulisine insulin preparations dissociate rapidly to monomers; thus, absorption is rapid. Lispro (B-28 lysine and B-29 proline human insulin; monomeric) insulin with two amino acids transposed, aspart (B-28 aspartic acid human insulin; monomeric and dimeric) insulin with replacement of one amino acid, and glulisine (B-3 lysine and B-29 glutamic acid) are rapidly absorbed, peak faster, and have shorter durations of action when compared with regular insulin. Proteins tend to be insoluble near their isoelectric point, and glargine insulin uses this to prolong absorption. In comparison to human insulin, with an isoelectric point of 5.4, the analog glargine insulin (A-21 glycine, B-30a-arginine, B-30a L-arginine, and B-30b L-arginine human insulin) has an isoelectric point of 6.8. In the bottle, glargine is buffered to a pH of 4, a level at which it is completely soluble, resulting in a clear colorless solution. When injected into the neutral pH of the body, it rapidly forms microprecipitates that slowly dissolve into monomers and dimers that are then subsequently absorbed. The result is a long-acting, approximately 24-hour duration insulin analog. Detemir, in contrast, attaches a C14 fatty acid (a 14-carbon fatty acid) at the B-29 position and removes the B-30 amino acid. This allows the fatty acid side chain to bind to interstitial albumin at the SQ injection site. Also, the formulation allows stronger hexamer self-associations, which prolong absorption. Once detemir dissociates from the interstitial albumin, it is free to enter a capillary, where it is again bound to albumin, which can further prolong action. It then travels to a site of action and interacts, after dissociation from albumin, with insulin receptors.

Insulin analogs are modified human insulin molecules, and safety is paramount for FDA approval. Key factors that should be considered in the approval process include local injection reactions, antigenicity, efficacy compared with human insulin, insulin receptor binding affinity, and insulin-like growth factor 1–receptor affinity (which is compared with that of human insulin to determine mitogenic potential).

Pharmacokinetics Subcutaneous injection kinetics is dependent on onset, peak, and duration of action, and is summarized in Table 57-9. Absorption of insulin from a subcutaneous depot is dependent on several factors, including source of insulin, concentration of insulin, additives to the insulin preparations (e.g., zinc, protamine), blood flow to the area (rubbing of injection area, increased skin temperature, and exercise in muscles near the injection site may enhance absorption), and injection site. Regular or neutral protamine Hagedorn (NPH) insulin is commonly injected in (from most rapid to slowest absorption): abdominal fat, posterior upper arms, lateral thigh area, and superior buttocks area. Insulin analogs, unlike regular or NPH insulin, appear to retain their kinetic profile at all sites of injection. U-500 regular insulin has a delayed onset and peak, and a longer duration of action when compared with U-100 regular insulin; the pharmacokinetic profile of U-500 is more similar to NPH.

TABLE 57-9 Pharmacokinetics of Various Insulins Administered Subcutaneously


Addition of protamine (NPH, NPL, and aspart protamine suspension) or excess zinc (historically lente or ultralente insulin) will delay onset, peak, and duration of the insulin’s effect. Variability in absorption, inconsistent suspension of the insulin by the patient or healthcare provider when drawing up a dose, and inherent insulin action based on the pharmacokinetics of the products may all contribute to a labile glucose response. NPH insulin and all suspension-based insulin preparations should be inverted or rolled gently at least 10 times to fully suspend the insulin prior to each use.

As detemir insulin has a unique mechanism to prolong absorption, it should not be surprising that the pharmacokinetics is unique. Detemir insulin reported less intrapatient variability between injections when compared with NPH or glargine insulin. This may be advantageous when variability in the insulin level may make a large difference in glycemic excursions, as in type 1 DM. It should be noted that at low dose (0.2 unit/kg) the duration of action is approximately 14 to 16 hours, while at doses above 0.3 unit/kg, it is close to 24 hours. In type 1 DM, 30% to 50% of patients may require twice-daily use of detemir insulin to cover 24-hour basal insulin needs, but this is unlikely to be an issue in type 2 DM patients, as they tend to use more units per day to attain glycemic goals. Direct comparative data between glargine insulin and detemir insulin are difficult to interpret, as detemir insulin was allowed to be dosed twice daily. Equivalent glycemic control was attained with either insulin. It is possible that glargine insulin in a minority of type 1 DM patients may require twice-daily dosing, but this is poorly documented in the literature.

The half-life of an IV injection of regular insulin is about 9 minutes. Thus, the effective duration of action of a single IV injection is short, and changes in IV insulin rates will reach steady state in approximately 45 minutes. IV pharmacokinetics of other soluble insulin preparations (lispro, aspart, glulisine, and even glargine) is similar to IV regular insulin, but they have no advantages over IV regular insulin and are more expensive. For completeness, aspart, lispro, and glulisine are FDA approved for IV use.

Insulin is degraded in the liver, muscle, and kidney. Liver deactivation is 20% to 50% in a single passage. Approximately 15% to 20% of insulin metabolism occurs in the kidney. This may partially explain the lower insulin dosage requirements in patients with end-stage renal disease.

Currently, insulin must be injected to retain its glycemic lowering properties. Alternative absorption pathways, including pulmonary, topical, GI, and even nasal, are being explored. The first inhalation insulin (Exubera) was discontinued due to poor sales and subsequent reports of lung cancer. Technosphere inhaled insulin (Afrezza) was rejected by the FDA due to concerns regarding the redesign of their delivery device. Additional trials are underway to assess the safety of the redesigned delivery device in patients with type 1 and type 2 DM. The onset of action is similar to IV insulin, which is unique.

Microvascular Complications Insulin has been shown to be as efficacious as any oral agent for treating DM. The United Kingdom Prospective Diabetes Study (UKPDS), which used sulfonylureas or insulin, showed equal efficacy in lowering the risk of microvascular events in newly diagnosed type 2 DM.24 Similarly, in type 1 DM, the DCCT showed efficacy in reducing microvascular complications.23

Macrovascular Complications The connection between high insulin levels (hyperinsulinemia), insulin resistance, and cardiovascular events incorrectly leads some clinicians to believe that insulin therapy may cause macrovascular complications. Endogenous hyperinsulinemia in the setting of insulin resistance has been linked to increased cardiovascular events; however, this is not the case with hyperinsulinemia due to exogenous injectable insulin preparations. The UKPDS and DCCT found no differences in macrovascular outcomes with intensive insulin therapy. One study, the Diabetes Mellitus, Insulin Glucose Infusion in Acute Myocardial Infarction study,29 reported reductions in mortality with insulin therapy. This group assessed the effect of an insulin–glucose infusion in type 2 DM patients who had experienced an acute myocardial infarction (MI). Those randomized to insulin infusion followed by intensive insulin therapy lowered their absolute mortality risk by 11% over a mean follow-up period of approximately 3 years. This was most evident in subjects who were insulin-naïve or had a low cardiovascular risk prior to the acute MI.29 The importance of glycemic control in hospitalized patients is covered later in the chapter.

Adverse Effects The most common adverse effects reported with insulin are hypoglycemia and weight gain. Hypoglycemia is more common in patients on intensive insulin therapy regimens versus those on less-intensive regimens. Also, patients with type 1 DM tend to have more hypoglycemic events compared with type 2 DM patients. In the UKPDS study, performed over 10 years, the percentage of diabetic patients who needed assistance (third-party or hospitalization) due to a hypoglycemic reaction was 2.3%. The UKPDS reported a rate of 36.5% for risk of any hypoglycemic event, including mild, self-treated events. In the DCCT, tighter control produced a risk three times higher for severe hypoglycemia compared with conventional therapy. Moreover, insulin was associated with 14% of emergency hospitalizations in older Americans using nationally representative public health surveillance data.30 Glycemic goals should incorporate hypoglycemic risk versus the benefit of lowering the glucose when HbA1c levels are near normal, especially in type 1 DM.

Hypoglycemia Minimization of risk for patients on insulin should include education about the signs and symptoms of hypoglycemia, proper treatment of hypoglycemia, and BG monitoring. BG monitoring is essential for those on insulin, and is particularly of value in patients with hypoglycemia unawareness. Patients with hypoglycemia unawareness do not experience the normal sympathetic symptoms of hypoglycemia (tachycardia, tremulousness, and, often, sweating). Initial hypoglycemia symptoms are neuroglycopenic in nature (confusion, agitation, loss of consciousness, and/or progression to coma). Patients with hypoglycemia unawareness should at least temporarily raise their glycemic goals (requiring a reduction in insulin dose) and check their BG level prior to any activities that may be dangerous with a low blood sugar (e.g., driving and certain sports, among others). Proper treatment of hypoglycemia dictates ingestion of carbohydrates, with glucose being preferred. Unconsciousness is an indication for either IV glucose or glucagon injection, which increases glycogenolysis in the liver. Glucagon use would be appropriate in any situation in which the patient does not have or cannot have ready IV access for glucose administration. Education for reconstitution and injection of glucagon is recommended for close friends and family of a patient who has recurrent neuroglycopenic events. The patient and close contacts should be informed that it can take 10 to 15 minutes for the injection to start raising glucose levels, and patients often vomit during this time. Proper positioning to avoid aspiration should be emphasized.

Weight gain is predominantly from increased truncal fat, and tends to be related to daily dose and plasma insulin levels present. It is undesirable in most type 2 DM patients, but may be seen as beneficial in underweight patients with type 1 DM. Weight gain appears to be related to intensive insulin therapy, and can be somewhat minimized by physiologic replacement of insulin.

Two forms of lipodystrophy, although much less common today in people with diabetes, still occur. Lipohypertrophy is caused by many injections into the same injection site. Due to insulin’s anabolic actions, a raised fat mass is present at the injection site with resultant variable insulin absorption. Lipoatrophy, in contrast, is thought to be due to insulin antibodies or allergic-type reactions with destruction of fat at the site of injection. In both cases, injection away from the site with more purified insulin is recommended, although reports of lipoatrophy have been reported with most insulin preparations. Anecdotal evidence has shown that specially formulated cromolyn may help to stabilize the allergic type of reaction.

One large study using administrative data found an association between insulin glargine and cancer. However, several other large database studies and meta-analysis have shown no such association. Glargine in vitro has a higher affinity for IGF-1 than regular human insulin, which could theoretically explain the increased risk of cancer, yet in vivo the metabolite of glargine is mostly present. The metabolite has similar affinity for IGF-1 as regular insulin. However, in the observational retrospective study, confounding by indication, selection, or detection bias in older patients may have played a greater role in the detection of cancer than the insulin glargine therapy. Supporting this premise, when glargine was used in intensive insulin therapy regimens in healthier populations, no such association was seen. Recently, the prospective, randomized Outcome Reduction with Initial Glargine Intervention trial reported no difference in cancer risk or cardiovascular events with low-dose insulin glargine use over approximately 6 years.31 These data are not definitive, but encouraging.

Drug–Drug Interactions There are no significant drug–drug interactions with injected insulin, although other medications that may affect glucose control can be considered. Detemir does not have albumin binding interactions, as it occupies only a small percent of albumin binding sites. Table 57-10 lists common medications known to affect BG levels.

TABLE 57-10 Medications that may Affect Glycemic Controla


Dosing and Administration The dose of insulin for any person with altered glucose metabolism must be individualized. In type 1 DM, the average daily requirement for insulin is 0.5 to 0.6 unit/kg, with approximately 50% being delivered as basal insulin, and the remaining 50% dedicated to meal coverage. During the honeymoon phase it may fall to 0.1 to 0.4 unit/kg. During acute illness or with ketosis or states of relative insulin resistance, higher dosages are warranted. In type 2 DM a higher dosage is required for those patients with significant insulin resistance. Dosages vary widely depending on underlying insulin resistance and concomitant oral insulin sensitizer use. Strategies on how to initiate and monitor insulin therapy will be described later in Therapeutics below.

U-500 regular insulin is reserved for use in patients with extreme insulin resistance and most often is given two or three times a day. Caution must be used, however, in order to avoid errors in prescribing and dispensing U-500. In the inpatient setting, the prescription of U-500 is often written in volume (mL) and administered using a tuberculin syringe. In an individual prescribed 50 units three times a day before meals, this prescription would be written as follows: “U-500 regular insulin, inject 50 units (0.1 mL) subcutaneously three times daily before meals.” In outpatients, however, it is often easier for patients to use U-100 insulin syringes. One unit of U-500 insulin drawn up using the markings of a U-100 equals 5 units of insulin. The same prescription as described above would be written as follows: “U-500 regular insulin: inject 50 units (10 units as measured by the unit markings of a U-100 syringe) subcutaneously three times daily before meals.”

Storage It is recommended that unopened injectable insulin be refrigerated (2°C to 8°C [36°F to 46°F]) prior to use. The manufacturer’s expiration date printed on the insulin is used for unopened, refrigerated insulin. Once the insulin is in use, the manufacturer-recommended expiration dates will vary based on the insulin and delivery device. Table 57-8 outlines manufacturer-recommended expiration dates for room temperature (15°C to 30°C [59°F to 86°F]) insulin. For financial reasons, patients may attempt to use insulin preparations longer than their expiration dates, but careful attention must be paid to monitoring for glycemic control deterioration and signs of insulin decay (clumping, precipitates, discoloration, etc.) if this is attempted.

Glucagon-Like Peptide-1 Agonists


Pharmacology Exendin 4 is a 39–amino acid peptide isolated from the saliva of the Gila monster (Heloderma suspectum) and shares 53% amino acid sequence with human GLP-1. Exenatide is the synthetic version of naturally occurring exendin 4. Exenatide (Byetta, Bydureon) has been shown to bind to GLP-1 receptors in many parts of the body including the brain and pancreas but is more resistant to DPP-4 degradation than endogenous GLP-1. Exenatide and GLP-1 have common glucoregulatory actions. The GLP-1 receptor activity of exenatide is pharmacologic, however, and is approximately three to four times more than the normal peak physiologic GLP-1 activity. Exenatide enhances insulin secretion in a glucose-dependent manner, suppressing inappropriately high postprandial glucagon secretion resulting in decreased hepatic glucose production. It increases satiety, slows gastric emptying, and promotes weight loss.

Pharmacokinetics There are two formulations of exenatide: exenatide injected twice daily (Byetta) and extended-release exenatide injected once weekly (Bydureon).

The concentration of twice-daily exenatide is detectable in plasma within 10 to 15 minutes after subcutaneous injection, and the drug has a tmax of ~2 hours and a plasma half life of ~3.3 to 4 hours. Plasma concentrations increase in a dose-dependent manner and concentrations are detectable for up to 10 hours postinjection, although pharmacodynamically, effects last for approximately 6 hours.

Extended-release once-weekly exenatide has a prolonged duration of action due to the exenatide being contained in a suspension of microspheres and gradually released over time. Following a single dose, exenatide is released from the microspheres over approximately 10 weeks. After initiation of once-weekly injections of 2 mg exenatide suspension, there is a gradual increase in plasma exenatide concentration over 6 to 7 weeks, after which steady state is achieved.

Bioavailability of exenatide after injection in the abdomen, upper arm, or the thigh is similar. Elimination of exenatide is primarily by glomerular filtration with subsequent proteolytic degradation. When exenatide is administered to subjects with worsening degrees of renal insufficiency, there is a progressive prolongation of the half-life, and in dialysis patients, plasma clearance of exenatide is markedly reduced. The incidence of GI side effects appears to be increased in individuals with impaired renal function, possibly due to higher plasma levels; thus, caution is advised.

No significant differences in exenatide pharmacokinetics have been observed with obesity, race, gender, or advancing age.

Efficacy The average HbA1c reduction is approximately 0.9% (0.009; 10 mmol/mol Hb) with twice-daily exenatide, similar to oral agents, but HbA1c lowering is dependent on baseline values. Some patients will have greater or lesser reduction in HbA1c. Similar HbA1c reduction is seen in patients on oral agents. Once-weekly extended-release exenatide resulted in significantly greater changes from baseline compared with twice-daily exenatide in HbA1c (–1.6% vs. –0.9% [–0.016 vs. –0.009; –18 mmol/mol Hb vs. –10 mmol/mol Hb) and FPG (–35 mg/dL vs. –12 mg/dL [–1.9 mmol/L vs. –0.7 mmol/L]).32

Exenatide significantly decreases postprandial glucose excursions, but has only a modest effect on FPG values. If a patient has significant elevations in FPG levels, these should be corrected with other agents and then exenatide added later. It is recommended to lower the sulfonylurea dose only if GLP-1 agonists are started with near-normal glucose levels. Sulfonylureas release insulin in a non–glucose-dependent fashion and can cause hypoglycemia.

Exenatide may aid some patients’ efforts to lose weight. The average weight loss in controlled trials of twice-daily exenatide was 1 to 2 kg over 30 weeks, without dietary advice being given to the patients. Long-term, open-label follow-up on 10 mcg twice a day shows continued and sustained weight loss for at least 3 years. Approximately 84% of patients on exenatide lost some weight. Exenatide, through decreasing appetite and slowing gastric emptying, may reduce the number of calories a patient eats at a meal. If a patient does not decrease calorie intake, no weight loss is likely to occur, as exenatide does not increase caloric expenditure.

Microvascular Complications Exenatide reduces the HbA1c level, which has been shown to be related to the risk of microvascular complications.

Macrovascular Complications No randomized clinical trials have examined the effect of exenatide on long-term cardiovascular outcomes. However, improvements in several cardiovascular risk factors have been reported. In an open-label study of exenatide 10 mcg twice a day, triglycerides (–37 ± 10 mg/dL [–0.42 ± 0.11 mmol/L]) decreased, and HDL cholesterol (+4.5 ± 0.4 mg/dL [0.12 ± 0.01 mmol/L]) increased. Once-weekly extended-release exenatide resulted in greater reduction in total cholesterol and LDL cholesterol compared with twice-daily exenatide.33 Nonsignificant reductions in systolic and diastolic blood pressure have been observed; a significant reduction was seen in subjects with above-normal systolic blood pressure. The greatest improvement in cardiovascular risk factors was, in general, seen in subjects who had the greatest weight loss.

Adverse Effects The most common adverse effects associated with exenatide are GI. Nausea is more likely with twice-daily exenatide (>35%) compared with once-weekly extended-release exenatide (~14%). Vomiting or diarrhea occurs in approximately 10% of patients on twice-daily exenatide. As these adverse effects appear to be dose related, the patient on twice-a-day exenatide should be started on 5 mcg twice a day and titrated to 10 mcg twice a day only if the adverse effects have resolved. When the patient is increased to the 10 mcg twice a day dose, these adverse effects may recur for a short period of time. GI adverse effects appear to decrease over time. However, approximately 1 in 20 patients on twice-daily exenatide have prolonged problems with side effects, possibly requiring discontinuation or transition to once-weekly extended-release exenatide.

Many episodes of nausea are better characterized as stomach fullness. Patients should be instructed to eat slowly and stop eating when full, or risk nausea/vomiting. Weight loss does not appear to be related to adverse effects, but rather to a reduction in calories consumed. Exenatide provides glucose-dependent insulin secretion; thus, hypoglycemic rates when combined with metformin or a TZD are not substantially increased. However, when combined with a sulfonylurea or insulin, hypoglycemia may occur. Although exenatide reduces glucagon when the glucose is high, there is no suppression of counterregulatory hormones during hypoglycemia. Exenatide antibodies can occur, but generally decrease over time and usually do not affect glycemic control. In approximately 5% of patients, titers may increase over time, potentially resulting in a deterioration of glycemic control.

Exenatide has been associated with cases of acute pancreatitis, but this has not been shown to be causal. Further study is needed, however, and several important points should be noted: (a) patients with type 2 DM often have risk factors for pancreatitis such as gallstones, hypertriglyceridemia, obesity, and concomitant medication use; (b) GLP-1 agonists could mask initial signs of pancreatitis, including nausea, vomiting, and abdominal pain; and (c) large database studies have not linked exenatide to a higher rate of acute pancreatitis. In a patient with a history of pancreatitis, the benefits of using exenatide must be weighed against potential risks. If a patient with abdominal pain, nausea, and/or vomiting presents, it is best to discontinue exenatide temporarily and confirm that the symptoms are not a sign of a more serious underlying problem. Exenatide given twice daily does not change the risk of thyroid C-cell tumors in rats and does not have a black box warning; no increased risk of C-cell tumors has been reported in humans. Extended-release exenatide has a black box warning in regards to thyroid C-cell tumors due to rat data. The difference appears to be that the extended release continually stimulates the GLP-1 receptor on the thyroid of rodents, increasing the risk of thyroid C-cell tumors. No tumors have been reported in humans.

There have been reports of injection site reactions with extended-release once-weekly exenatide. Nodule injection site reactions are not painful and are often not visible, but can be felt at the injection site, which may have been injected 2 to 4 weeks prior. These nodules are an aggregation of the microspheres subcutaneously, not an immune reaction, and they may last 6 to 8 weeks. Injection site erythema, which can be severe in some cases, is related to exenatide antibody status (potentially worse if very high titers) or may be due to the platform, as this reaction is well described with the poly(D,L-lactide-co-glycolide) microsphere material.

Drug Interactions Exenatide delays gastric emptying; if the patient has gastroparesis, exenatide is not recommended. Exenatide can also delay the absorption of other medications. Examples of medications that may be effected include oral pain medications and antibiotics dependent on concentration-dependent efficacy. If rapid absorption of the medication is necessary, it is best to take the mediation 1 hour before, or at least 3 hours after, the injections of twice-daily exenatide. There have been postmarketing reports of increased INR in patients on warfarin on exenatide, sometimes associated with bleeding. It is advised that INR be monitored frequently until stable on initiation of exenatide.

Dosing and Administration Dosing of twice-daily exenatide (Byetta) should begin with 5 mcg twice a day, and titrated to 10 mcg twice a day in 1 month or when tolerability allows and if warranted for glycemic control. Twice-daily exenatide should be injected subcutaneously 0 to 60 minutes before the morning and evening meals. If the patient does not eat breakfast, he or she may take the first injection of the day at lunch. The peak effect of twice-daily exenatide is at approximately 2 hours, so anecdotally the patient may get better appetite suppression if injected an hour prior to the meal.

The dosing of extended-release exenatide (Bydureon) is 2 mg suspension injected subcutaneously every 7 days, at any time of day, with or without meals. Extended-release exenatide is injected immediately after the powder is suspended in the diluent. The process of extended-release once-weekly exenatide injections is more complex than using the twice-daily exenatide pen. Patients must be instructed on self-administration.

Exenatide may be injected in abdomen, thigh, or upper arm region, but patients are advised to use a different injection site when injecting into the same region.

Storage and dosage availability information can be found in Table 57-8.


Pharmacology Liraglutide (Victoza) is a GLP-1 receptor agonist that has 97% amino acid sequence homology to endogenous GLP-1. The only alteration is an arginine substituted for lysine at position 34. A C-16 fatty acid (palmitic acid) is attached at position 26 (with a glutamic amino acid spacer to optimize GLP-1 receptor interaction) so that liraglutide can bind noncovalently to albumin, prolonging the half-life.

Liraglutide enhances glucose-dependent insulin secretion while suppressing inappropriately high glucagon secretion in the presence of elevated glucose concentrations, resulting in a reduction in hepatic glucose production. Liraglutide reduces food intake, which may result in weight loss, and slows gastric emptying so that the rate of glucose appearance into the plasma better matches the glucose disposition. During hypoglycemia, liraglutide does not stimulate insulin secretion and does not inhibit the release of the counterregulatory hormone glucagon.

Pharmacokinetics After injection of liraglutide, there is self-association into a heptameric structure, binding to albumin first in the interstitial space, then in the blood, and then in the interstitial space around the GLP-1 receptor that prolongs the half-life. In healthy individuals, the half-life is 13 hours, making it suitable for daily administration. Injection into the abdomen, upper arm, and thigh gives clinically similar pharmacokinetics. Maximum concentrations are reached approximately 8 to 12 hours after injection, with steady state reached after approximately 3 days. Liraglutide is extensively plasma protein bound (mostly to albumin as previously stated) with an elimination half-life of 10 to 18 hours. The absolute bioavailability is approximately 50%.

The metabolism of liraglutide appears to be by degradation, similar to other large proteins, and several small minor metabolites (total of 3% to 5% of the dose) may be found. The DPP-4 enzyme in vitro has been shown to slowly metabolize liraglutide, and this may be the case in vivo as well.

The pharmacokinetics of liraglutide does not appear to be affected by age, race, and gender. Severe renal or mild to severe hepatic impairment may actually lower the AUC by approximately 25%, although the clinical significance of this is not known.

Efficacy The average HbA1c reduction is approximately 1.1% (0.0011; 12 mmol/mol Hb) with liraglutide. Similar to other agents, the reduction in HbA1c is dependent on the baseline values. Liraglutide lowers FPG level by approximately 25 to 40 mg/dL (1.4 to 2.2 mmol/L), and postprandial plasma glucose levels are reduced similarly. Due to the longer half-life, liraglutide can suppress glucagon overnight, which improves the FPG. Similar to exenatide, liraglutide-treated patients may lose weight. The average weight loss in controlled trials was 1 to 3 kg over 26 weeks, and weight loss achieved appeared to be sustained through 2 years. Liraglutide, through decreasing appetite and slowing gastric emptying, may reduce the number of calories a patient eats at a meal.

Microvascular Complications Liraglutide reduces the HbA1c level, which has been shown to be related to the risk of microvascular complications.

Macrovascular Complications There are no published clinical trials examining the effect of liraglutide on long-term cardiovascular outcomes; however, no signal of cardiovascular harm was noted on FDA approval.

Adverse Effects The most common adverse effects associated with liraglutide are GI. Nausea occurs in ~11% to 29% of subjects on 1.2 mg, and 14% to 40% of subjects on 1.8 mg daily. Vomiting occurs in approximately 5% of subjects, and diarrhea occurs in approximately 8% to 15% of patients placed on liraglutide. GI adverse effects appear to decrease over time, but approximately 5% to 10% of subjects withdrew due to GI side effects. As these adverse effects appear to be dose related, the patient should be titrated from 0.6 to 1.2 mg, and to 1.8 mg as tolerated. Randomized trials did not allow for individualized titration, and likely had worse tolerability that can be obtained clinically by individualization of titration.

Many episodes of nausea would be better characterized as stomach fullness. To minimize GI side effects, patients should be instructed to eat slowly and stop eating when full, or risk nausea/vomiting. Liraglutide provides glucose-dependent insulin secretion, and hypoglycemic rates when combined with metformin ± a TZD are not substantially increased, but when combined with a sulfonylurea or insulin, significant hypoglycemia may occur. When combined with a sulfonylurea, the rates of hypoglycemia were similar between addition of liraglutide and that of glargine insulin. Liraglutide antibodies can occur (4% to 13%), but the rates are generally low and do not affect glycemic control or risk of side effects.

Liraglutide has been associated with the serious adverse event of acute pancreatitis, but causality has not been proven. Further study is needed, but type 2 DM patients have many risk factors for pancreatitis and the common GI side effects of GLP-1 agonists could mask initial signs of pancreatitis. If a patient with abdominal pain, nausea, and/or vomiting presents, it is best to discontinue liraglutide temporarily and if symptoms persist, evaluate for other potential causes, including pancreatitis. Clinicians must weigh the benefits of liraglutide against the potential risks in a patient with a history of pancreatitis.

A boxed warning about thyroid C-cell tumors (as with extended-release exenatide) is listed in the package insert of liraglutide. Rodent models reported a higher risk of C-cell tumors of the thyroid, including medullary thyroid carcinoma. Rodents may not be the ideal model to study this effect as they express a high number of GLP-1 receptors on thyroid C-cells, whereas in humans the expression of GLP-1 receptors in the thyroid is minimal. Rodents also have a higher baseline prevalence of C-cell tumors compared with humans. In addition, calcitonin, a marker used to screen for C-cell tumors, may increase by a non–clinically significant amount in select patients. No signal for C-cell tumors in humans or nonhuman primates has been noted thus far. As clinical use increases, however, this will continue to be examined. Currently no specific additional monitoring of patients is recommended. Nonetheless, liraglutide is contraindicated in patients with a personal or family history of medullary thyroid cancer, and in those with multiple endocrine neoplasia syndromes.

Drug Interactions Liraglutide delays gastric emptying; thus, it can delay the absorption of other medications. Examples of medications that may be effected include oral pain medications and antibiotics dependent on threshold levels for efficacy. If rapid absorption of the medication is necessary, it is best to take the mediation 1 hour before, or at least 3 hours after, the injection. Liraglutide may worsen gastroparesis and clinically it may not be prudent to use in this patient population.

Dosing and Administration The dosing of liraglutide should begin with 0.6 mg daily for ≥1 week, and then increased to 1.2 mg daily for ≥1 week. Patients may be maintained on the 1.2-mg dose, or increased to the maximum dose of 1.8 mg daily after ≥1 week. The 0.6-mg dose is considered a titration dose, and does not reduce the HbA1c substantially in the majority of patients. This titration is recommended to improve GI tolerability. Titration should be individualized based on side effects and clinical response. Liraglutide is dosed once daily, and may be given independent of meals. As with exenatide, a reduction in insulin secretagogues and insulin may be necessary if the patient is near glycemic goal or hypoglycemia occurs.

Storage and dosage availability information can be found in Table 57-8.



Pharmacology Pramlintide (Symlin) is an antihyperglycemic agent used in patients currently treated with insulin. Pramlintide is a synthetic analog of amylin (amylinomimetic), a neurohormone cosecreted from the β cells with insulin. Amylin is very low or absent in type 1 DM, and lower than normal in type 2 DM patients requiring insulin therapy. Pramlintide is provided as a 37–amino acid polypeptide, which differs in amino acid sequence from human amylin by replacement positions 25 (alanine), 28 (serine), and 29 (serine) with proline. Pramlintide suppresses inappropriately high postprandial glucagon secretion, increases satiety, which may result in weight loss, and slows gastric emptying so that the rate of glucose appearance into the plasma better matches the glucose disposition.

Pharmacokinetics The absolute bioavailability of pramlintide after subcutaneous injection is 30% to 40%. The tmax is approximately 20 minutes, but the Cmax is dose dependent. The t1/2 is approximately 45 minutes; thus, the pharmacodynamic duration of action is about 3 to 4 hours. Pramlintide does not extensively bind to albumin, and should not have significant binding interactions. Metabolism is primarily by the kidneys, and one active metabolite (2-37 pramlintide) has a similar half-life as the parent compound. No accumulation has been seen in renal insufficiency, but caution is advised. Injection into the arm may increase exposure and variability of absorption, so injection into the abdomen or thigh is recommended. Moderate to severe renal insufficiency does not affect exposure.

Efficacy The average HbA1c reduction is approximately 0.6% (0.006; 7 mmol/mol Hb) with pramlintide, although optimization of the insulin and pramlintide doses may result in further drops in HbA1c. If the 120-mcg dose is used in type 2 DM patients on insulin, it may also result in 1.5-kg weight loss. In type 1 DM patients, the average reduction in HbA1c was 0.4% to 0.5% (0.004 to 0.005; 5 to 6 mmol/mol Hb). Prandial pramlintide added versus rapid-acting insulin in type 2 DM subjects uncontrolled on basal insulin reported similar efficacy, but with no weight gain, compared with ~5-kg weight gain with rapid-acting insulin. Pramlintide decreases prandial glucose excursions, but has little effect on the FPG concentration. When pramlintide is injected before the meal, gastric emptying may delay absorption of mealtime nutrients, necessitating delay of rapid-acting insulin. This may be overcome by injecting the mealtime insulin at the conclusion of the meal, or whenever the BG starts to rise. The average weight loss in controlled trials was 1 to 2 kg, without dietary advice being given to the patients. Pramlintide, through decreasing appetite, may reduce the number of calories a patient eats at a meal.

Microvascular Complications Pramlintide reduces the HbA1c level, which has been shown to be related to the risk of microvascular complications.

Macrovascular Complications No published clinical trials have examined the effect of pramlintide on cardiovascular outcomes.

Adverse Effects The most common adverse effects associated with pramlintide are GI in nature. Nausea occurs in ~20% of type 2 DM patients, and vomiting or anorexia occurs in approximately 10% of type 1 or type 2 DM patients. Nausea is more common in type 1 DM, occurring in ~40% to 50% of patients. The higher rates in type 1 DM related to GI adverse effects appear to decrease over time and are dose related; thus, starting at a low dose and slowly titrating as tolerated is recommended. Pramlintide alone does not cause hypoglycemia, but it is indicated for use in patients on insulin; thus, hypoglycemia can occur. The risk of severe hypoglycemia early in therapy is higher in type 1 DM than in type 2 DM patients. A twofold increase in severe hypoglycemic reactions in type 1 DM patients has been reported.

Drug Interactions Pramlintide delays gastric emptying; thus, it can delay the absorption of other medications. Examples of medications that may be effected include oral pain medications and antibiotics dependent on threshold levels for efficacy. If rapid absorption of the medication is necessary, it is best to take the mediation 1 hour before, or at least 3 hours after, the injection of pramlintide.

Dosing and Administration Pramlintide dosing varies in type 1 and type 2 DM. It is imperative that the prandial insulin dose, if used, be reduced 30% to 50% when pramlintide is started to minimize severe hypoglycemic reactions or delayed until postprandial glucose levels rise. Basal insulin may need to be adjusted only if the FPG is close to normal. In type 2 DM, the starting dose is 60 mcg prior to meals, and may be titrated to the maximally recommended 120-mcg dose as tolerated and warranted based on postprandial plasma glucose concentrations. At least one clinical trial started at the 120-mcg dose without significantly more intolerability. In type 1 DM, dosing starts at 15 mcg prior to meals, and can be titrated up in 15-mcg increments to a maximum of 60 mcg prior to each meal if tolerated and warranted. Snacks may or may not need to be covered with pramlintide (recommended if ≥250 kcal [≥1,046 kJ] or ≥30 g of carbohydrate is eaten). Storage information can be found in Table 57-8.


Pharmacology The primary mechanism of action of sulfonylureas is enhancement of insulin secretion. Sulfonylureas bind to a specific sulfonylurea receptor (SUR) on pancreatic β cells. Binding closes an adenosine triphosphate–dependent K+ channel, leading to decreased potassium efflux and subsequent depolarization of the membrane. Voltage-dependent Ca2+ channels open and allow an inward flux of Ca2+. Increases in intracellular Ca2+ bind to calmodulin on insulin secretory granules, causing translocation of secretory granules of insulin to the cell surface and resultant exocytosis of the granule of insulin. Elevated secretion of insulin from the pancreas travels via the portal vein and subsequently suppresses hepatic glucose production.

Classification Sulfonylureas are classified as first-generation and second-generation agents. The classification scheme is largely derived from differences in relative potency, potential for selective side effects, and differences in binding to serum proteins (i.e., risk for protein-binding displacement drug interactions). First-generation agents consist of acetohexamide, chlorpropamide, tolazamide, and tolbutamide. Each of these agents is lower in potency relative to the second-generation drugs: glimepiride, glipizide, and glyburide (Table 57-11). It is important to recognize that all sulfonylureas are equally effective at lowering BG when administered in equipotent doses.

TABLE 57-11 Oral Agents for the Treatment of Type 2 Diabetes Mellitus




Pharmacokinetics All sulfonylureas are metabolized in the liver, some to active and others to inactive metabolites. Glyburide metabolites are active, whereas glipizide and glimepiride do not have active metabolites. Cytochrome P450 (CYP450) 2C9 is involved with the hepatic metabolism of the majority of sulfonylureas. Agents with active metabolites or parent drug that are renally excreted require dosage adjustment or use with caution in patients with compromised renal function. The half-life of the sulfonylurea also relates directly to the risk for hypoglycemia. The hypoglycemic potential is therefore higher with chlorpropamide and glyburide. The long duration of effect of chlorpropamide may be particularly problematic in elderly individuals, whose renal function declines with age, and therefore it has great potential for accumulation, resulting in severe and protracted hypoglycemia. Individuals at high risk for hypoglycemia (e.g., elderly individuals and those with renal insufficiency or advanced liver disease) should be started at a very low dose of a sulfonylurea with a short half-life. Hypoglycemia on low-dose sulfonylureas may dictate a therapy without the risk of hypoglycemia.

Efficacy As mentioned earlier, when given in equipotent doses, all sulfonylureas are equally effective at lowering BG. On average, HbA1c will fall 1.5% to 2% (0.015 to 0.020; 17 to 22 mmol/mol Hb) in drug-naïve patients, with FPG reductions of 60 to 70 mg/dL (3.3 to 3.9 mmol/L), but is dependent on baseline values and duration of diabetes. A majority of patients will not reach glycemic goals with sulfonylurea monotherapy. Patients with inadequate control on a sulfonylurea usually fall into two groups: those with low C-peptide levels and high (>250 mg/dL [>13.9 mmol/L]) FPG levels. These patients are often primary failures on sulfonylureas (<30 mg/dL [<1.7 mmol/L] drop of FPG) and have significant glucose toxicity or LADA. The other group is those with a good initial response (>30 mg/dL [>1.7 mmol/L] drop of FPG), but which is insufficient to reach their glycemic goals. Over 75% of patients fall into the second group. Factors that portend a positive response include newly diagnosed patients with no indicators of type 1 DM, high fasting C-peptide levels, and moderate fasting hyperglycemia (<250 mg/dL [<13.9 mmol/L]).

Microvascular Complications Sulfonylureas showed a reduction of microvascular complications in type 2 DM patients in the UKPDS.24 A more in-depth discussion follows later in the chapter.

Macrovascular Complications The UKPDS reported no significant benefit or harm in newly diagnosed type 2 DM patients given sulfonylureas over 10 years. The University Group Diabetes Program study documented higher rates of coronary artery disease in type 2 patients given tolbutamide, when compared with patients given insulin or placebo, although this study has been widely criticized.34 Some sulfonylureas bind to the SUR-2A receptor that is found in cardiac tissue. Binding to the SUR-2A receptor has been implicated in blocking ischemic preconditioning via K+ channel closure in the heart. Ischemic preconditioning is the premise that prior ischemia in cardiac tissue can provide greater tolerance of subsequent ischemia. Thus, patients with heart disease potentially have one compensatory mechanism to protect the heart from ischemia blocked. Conclusions are controversial, but alternative treatments are available if questioned.

Adverse Effects The most common side effect of sulfonylureas is hypoglycemia. The pretreatment FPG is a strong predictor of hypoglycemic potential. The lower the FPG is on initiation, the higher the potential for hypoglycemia. Also, in addition to the high-risk individuals outlined in Pharmacokinetics below, those who skip meals, exercise vigorously, or lose substantial amounts of weight are also more likely to experience hypoglycemia.

Hyponatremia (serum sodium <129 mEq/L [129 mmol/L]) is reportedly associated with tolbutamide, but it is most common with chlorpropamide and occurs in as many as 5% of individuals treated. An increase in antidiuretic hormone secretion is the mechanism for hyponatremia. Risk factors include age >60 years, female gender, and concomitant use of thiazide diuretics.

Weight gain is common with sulfonylureas. In essence, patients who are no longer glycosuric and who do not reduce caloric intake with improvement of BG will store excess calories. Other notable, although much less common, adverse effects of sulfonylureas are skin rash, hemolytic anemia, GI upset, and cholestasis. Disulfiram-type reactions and flushing have been reported with tolbutamide and chlorpropamide when alcohol is consumed.

Drug Interactions Several drugs are thought to interact with sulfonylureas, most likely through the CYP450 system or altered renal excretion. Protein-binding changes should occur shortly after the interacting medication is given, as the concentration of free (thus active) sulfonylurea will acutely increase. First-generation sulfonylureas, which bind to proteins ionically, are more likely to cause drug–drug interactions than second-generation sulfonylureas, which bind nonionically. The clinical importance of protein-binding interactions has been questioned, as the majority of these drug interactions have been found to be truly due to hepatic metabolism. Drugs that are inducers or inhibitors of CYP450 2C9 should be monitored carefully when used with a sulfonylurea.35 Additionally, other drugs known to alter BG should be considered (Table 57-10).

Dosing and Administration The usual starting dose and maximum dose of sulfonylureas are summarized in Table 57-11. Lower dosages are recommended for most agents in elderly patients and those with compromised renal or hepatic function. The dosage can be titrated as soon as every 2 weeks based on FPG values (use a longer interval with chlorpropamide) to achieve glycemic goals. This is possible due to the rapid increase of insulin secretion in response to the sulfonylurea. Of note, immediate-release glipizide’s maximal dose is 40 mg/day, but its maximal effective dose is about 10 to 15 mg/day. The maximal effective dose of sulfonylureas tends to be about 60% to 75% of their stated maximum dose.

Short-Acting Insulin Secretagogues

Pharmacology Although the binding site is adjacent to the binding site of sulfonylureas, nateglinide and repaglinide stimulate insulin secretion from the β cells of the pancreas, similarly to sulfonylureas. Both repaglinide (a benzoic acid derivative) and nateglinide (a phenylalanine amino acid derivative) require the presence of glucose to stimulate insulin secretion. As glucose levels diminish to normal, stimulated insulin secretion diminishes.

Pharmacokinetics Both nateglinide and repaglinide are rapid-acting insulin secretagogues that are rapidly absorbed (~0.5 to 1 hour) and have a short half-life (1 to 1.5 hours). Nateglinide is highly protein bound, primarily to albumin, but also to α1-acid glycoprotein. It is predominantly metabolized by CYP2C9 (70%) and CYP3A4 (30%) to less active metabolites. Glucuronide conjugation then allows rapid renal elimination. No dosage adjustment is needed in moderate to severe renal insufficiency. Repaglinide is highly protein bound, and is mainly metabolized by oxidative metabolism and glucuronidation. The CYP3A4 and 2C8 systems have been shown to be involved with metabolism. Approximately 90% of repaglinide is eliminated in the feces, with only 10% found in the urine. Moderate to severe renal insufficiency does not appear to affect repaglinide, but moderate to severe hepatic impairment may prolong exposure.

Efficacy In monotherapy, both significantly reduce postprandial glucose excursions and reduce HbA1c levels. Repaglinide, dosed 4 mg three times a day, when compared with glyburide in diet-treated, drug-naïve patients reduced HbA1c levels less (1% vs. 2.4% [0.01 vs. 0.024; 11 mmol/mol Hb vs. 26 mmol/mol Hb], from baseline, respectively). Nateglinide, dosed 120 mg three times a day, in a similar population reduced HbA1c values by 0.8% (0.008; 9 mmol/mol Hb). The lower efficacy of these agents versus sulfonylureas should be considered when patients are >1% (>0.01; >11 mmol/mol Hb) above their HbA1c goal. These agents can be used to provide increased insulin secretion during meals, when it is needed, in patients close to glycemic goals. Also, it should be noted that addition of either agent to a sulfonylurea will not result in any improvement in glycemic parameters.

Adverse Effects Hypoglycemia is the main side effect noted with both agents. Hypoglycemic risk appears to be less versus sulfonylureas. In part, this is due to the glucose-sensitive release of insulin. If the glucose concentration is normal, less glucose-stimulated release of insulin will occur. In two separate studies, nateglinide rates of hypoglycemia were 3% and repaglinide 15% versus glyburide and glipizide rates of 15% and 19%, respectively. Weight gain of 2 to 3 kg has been noted with repaglinide, whereas weight gain with nateglinide appears to be <1 kg.

Drug Interactions Glycemic control and hypoglycemia should be closely monitored when glucuronidation inhibitors are given with repaglinide. Gemfibrozil more than doubles the half-life of repaglinide and has resulted in prolonged hypoglycemic reactions. It is a potent glucuronidation inhibitor and CYP2C8 inhibitor. Trimethoprim, a CYP2C8 inhibitor, increased repaglinide levels by 60%. Nateglinide appears to be a weak inhibitor of CYP2C9 based on tolbutamide metabolism. Although no significant drug–drug interactions have been reported, caution should be used with strong CYP2C9 and CYP3A4 inhibitors.

Dosing and Administration Nateglinide and repaglinide should be dosed prior to each meal (up to 30 minutes prior). The recommended starting dose for repaglinide is 0.5 mg in subjects with HbA1c <8% (<0.08; <64 mmol/mol Hb) or treatment-naïve patients, increased weekly to a total maximum daily dose of 16 mg (see Table 57-11). The maximal effective dose of repaglinide is likely 2 mg with each meal, as a dose of 1 mg prior to each meal provides approximately 90% of the maximal glucose-lowering effect. Nateglinide should be dosed at 120 mg prior to meals, and does not require titration. A 60-mg dose is available, but the HbA1c decrement is small (0.3% to 0.5% [0.003 to 0.005; 3 to 6 mmol/mol Hb]). If a meal is skipped, the medication can be skipped, and meals extremely low in carbohydrate content may not need a dose. Both agents may be used in patients with renal insufficiency, and may fit into therapy in patients in need of an insulin secretagogue but having hypoglycemia to sulfonylureas, moderate to severe renal insufficiency, and well-controlled diabetes, but with erratic meal schedules.


Pharmacology Metformin is the only biguanide available in the United States. It has been used clinically for more than 50 years, and has been approved in the United States since 1995. Metformin enhances insulin sensitivity of mainly hepatic but also peripheral (muscle) tissues. This allows for an increased uptake of glucose into these insulin-sensitive tissues. All the mechanisms of how metformin accomplishes glucose reduction are still being investigated, although adenosine 5′-monophosphate–activated protein kinase activity, tyrosine kinase activity enhancement, increased adenosine 5′-monophosphate, and partial inhibition of the mitochondrial respiratory chain are involved. Metformin has no direct effect on the β cells, although insulin levels are reduced, reflecting increases in insulin sensitivity.

Pharmacokinetics Metformin has approximately 50% to 60% oral bioavailability, low lipid solubility, and a volume of distribution that approximates body water. It is not metabolized and does not bind to plasma proteins. Metformin is eliminated by renal tubular secretion and glomerular filtration. The average plasma half-life of metformin is 6 hours, although pharmacodynamically, metformin’s antihyperglycemic effects last more than 24 hours. Red blood cells are a second compartment of distribution for metformin, delivering an effective half-life of 17 hours.

Efficacy Metformin consistently reduces HbA1c levels by 1.5% to 2% (0.015 to 0.020; 17 to 22 mmol/mol Hb) and FPG levels by 60 to 80 mg/dL (3.3 to 4.4 mmol/L) in drug-naïve patients, and retains the ability to reduce FPG levels when they are extremely high (>300 mg/dL [>16.7 mmol/L]). The sulfonylureas’ ability to stimulate insulin release from β cells at extremely high glucose levels is often impaired, a concept commonly referred to as glucose toxicity. Metformin also has positive effects on several components of the insulin resistance syndrome. It decreases plasma triglycerides and LDL-C by approximately 8% to 15%, in addition to increasing HDL-C very modestly (2%). Metformin reduces levels of PAI-1 and causes a modest reduction in weight (2 to 3 kg). In preliminary findings, metformin may also lower the risk of pancreatic, colon, and breast cancer in type 2 DM patients. Metformin, potentially through multiple mechanisms including adenosine 5′-monophosphate–activated protein kinase activity, may act as a growth inhibitor in some cancers and help to kill cancer “stem cells” which are resistant to chemotherapy, and liver kinase B1, which is an upstream kinase of adenosine 5′-monophosphate–activated protein kinase. More controlled studies are needed.

Microvascular Complications Metformin (n = 342) was compared with intensive glucose control with insulin or sulfonylureas in the UKPDS. No significant differences were seen between therapies with regard to reducing microvascular complications, but the power of the study is questionable.36

Images Macrovascular Complications Metformin reduced macrovascular complications in obese subjects in the UKPDS.36 It significantly reduced all-cause mortality and risk of stroke versus intensive treatment with sulfonylureas or insulin. Metformin also reduced diabetes-related death and MIs versus the conventional treatment arm of the UKPDS. It should be noted that the UKPDS had very few people on lipid-lowering therapy, antihypertensives, or aspirin. Metformin is logical in overweight/obese patients, if tolerated and not contraindicated, as it is the only oral antihyperglycemic medication potentially proven to reduce the risk of total mortality and is generic.

Adverse Effects Metformin causes GI side effects, including abdominal discomfort, stomach upset, and/or diarrhea, in approximately 30% of patients. Anorexia and stomach fullness is likely part of the reason loss of weight is noted with metformin. These side effects are usually mild and can be minimized by slow titration. GI side effects also tend to be transient, lessening in severity over several weeks. If encountered, make sure patients are taking metformin with or right after meals, and reduce the dose to a point at which no GI side effects are encountered. Increases in the dose may be tried again in several weeks. Anecdotally, extended-release metformin (Glucophage XR) may lessen some of the GI side effects. Metallic taste, interference with vitamin B12 absorption, and hypoglycemia during intense exercise have been documented, but are clinically uncommon.

Metformin therapy rarely (3 to 9 cases per 100,000 patient-years) causes lactic acidosis. Metformin partially blocks the mitochondrial respiratory chain. In addition, any disease state that may increase lactic acid production or decrease lactic acid removal may predispose to lactic acidosis. Tissue hypoperfusion, such as that due to congestive heart failure, severe lung disease, hypoxic states, shock, or septicemia, via increased production of lactic acid, and severe liver disease or alcohol, via reduced removal of lactic acid in the liver, all increase the risk of lactic acidosis. The clinical presentation of lactic acidosis is often nonspecific flu-like symptoms; thus, the diagnosis is usually made by laboratory confirmation of high lactic acid levels and acidosis. Metformin use in renal insufficiency, defined as a serum creatinine of 1.4 mg/dL (124 μmol/L) in women and 1.5 mg/dL (133 μmol/L) in men or greater, is contraindicated, as it is renally eliminated. Elderly patients, who often have reduced muscle mass, should have their glomerular filtration rate estimated by a 24-hour urine creatinine collection. If the estimated glomerular filtration rate is less than 60 mL/min (1 mL/s), metformin use should be carefully evaluated. Recent evidence has reported that metformin may be fairly safe in moderate renal insufficiency. Metformin use can be modified based on the estimated glomerular filtration rate, at <60, <45 to ≥30, and <30 mL/min/1.73 m2 (<0.58, <0.43 to ≥0.29, and <0.29 mL/s/m2); corresponding actions are to monitor renal function every 3 to 6 months, then limit dose to 50% of maximal dose, and then stop metformin, respectively. Due to the risk of acute renal failure during IV dye procedures, metformin therapy should be withheld starting the day of the procedure and resumed in 2 to 3 days, after normal renal function has been documented.

Drug Interactions Cimetidine competes for renal tubular secretion of metformin and concomitant administration leads to higher metformin serum concentrations. At least one case report of lactic acidosis with metformin therapy implicates cimetidine. Theoretically other cationic drugs may interact, but none have been reported to date.

Dosing and Administration Immediate-release metformin is usually dosed 500 mg twice a day with the largest meals to minimize GI side effects. Metformin may be increased by 500 mg as tolerated until glycemic goals or 2,500 mg/day is achieved (see Table 57-11). Metformin 850 mg may be dosed daily, and then increased every 1 to 2 weeks to the maximum dose of 850 mg three times a day (2,550 mg/day). Approximately 80% of the glycemic-lowering effect may be seen at 1,500 mg, and 2,000 mg/day is the maximal effective dose.

Extended-release metformin can be initiated at 500 mg a day with the evening meal and titrated by 500 mg as tolerated to a single evening dose of 2,000 mg/day. Extended-release metformin 750-mg tablets may be titrated as tolerated to the maximum dose of 2,250 mg/day, although, as stated above, 1,500 mg/day provides the majority of the glycemic-lowering effect. Twice-daily to three-times-a-day dosing of extended-release metformin may help to minimize GI side effects and improve glycemic control, but will not change the glycemic reduction.


Pharmacology TZDs are also referred to as glitazones. Pioglitazone (Actos) and rosiglitazone (Avandia) are the two currently approved TZDs for the treatment of type 2 DM (see Table 57-11). TZDs work by binding to the PPAR-γ, which are primarily located on fat cells and vascular cells. The concentration of these receptors in the muscle is very low, but improvement in mitochondrial function through changes in lipotoxicity, glucotoxicity, and possibly binding of proteins outside the mitochondrial membrane may occur. TZDs enhance insulin sensitivity at muscle, liver, and fat tissues indirectly. They cause preadipocytes to differentiate into mature fat cells in subcutaneous fat stores. Small fat cells are more sensitive to insulin and more able to store FFAs. The result is a flux of FFAs out of the plasma, visceral fat, and liver into subcutaneous fat, a less insulin-resistant storage tissue. Muscle intracellular fat products, which contribute to insulin resistance, also decline. TZDs also affect adipokines (e.g., angiotensinogen, tissue necrosis factor-α, interleukin 6, PAI-1), which can positively affect insulin sensitivity, endothelial function, and inflammation. Of particular note, adiponectin is reduced with obesity and/or diabetes, but is increased with TZD therapy, which improves endothelial function and insulin sensitivity, and has a potent antiinflammatory effect. Lastly, TZDs appear to improve mitochondrial function through a reduction in FFAs. Cyclin-dependent kinase 5 has also recently been purposed as an important activator of PPAR-γ.

Pharmacokinetics Pioglitazone and rosiglitazone are well absorbed with or without food. Both are highly (>99%) bound to albumin. Pioglitazone is primarily metabolized by CYP2C8, to a lesser extent by CYP3A4 (17%), and by hydroxylation/oxidation. The majority of pioglitazone is eliminated in the feces with 15% to 30% appearing in urine as metabolites. Two active metabolites (M-III and M-IV) are present. Rosiglitazone is metabolized by CYP2C8, and to a lesser extent by CYP2C9, and also by N-demethylation and hydroxylation. Two thirds is found in urine and one third in feces. The half-lives of pioglitazone and rosiglitazone are 3 to 7 and 3 to 4 hours, respectively. The two active metabolites of pioglitazone, with longer half-lives, deliver the majority of activity at steady state. Pioglitazone requires no dosage adjustment in moderate to severe renal disease for pharmacokinetic reasons. Interestingly, with pioglitazone the AUC in women is 20% to 60% higher, which is not seen with rosiglitazone, but no dosage adjustment is recommended. Both medications have a duration of antihyperglycemic action of over 24 hours.

Efficacy Pioglitazone and rosiglitazone reduce HbA1c values ~1% to 1.5% (~0.010 to 0.015; 11 to 17 mmol/mol Hb) and reduce FPG levels by ~60 to 70 mg/dL (~3.3 to 3.9 mmol/L) at maximal doses. Glycemic-lowering onset is slow, and maximal glycemic-lowering effects may not be seen until 3 to 4 months of therapy. It is important to inform patients of this fact and that they should not stop therapy even if minimal glucose lowering is initially encountered. The efficacy of both drugs is dependent on sufficient insulinemia. If there is insufficient endogenous insulin production (β-cell function) or exogenous insulin delivery via injections, neither will lower glucose concentrations efficiently. Interestingly, patients who are more obese or who gain weight on either medication tend to have a larger reduction in HbA1cvalues. Pioglitazone consistently decreases plasma triglyceride levels by 10% to 20%, whereas rosiglitazone tends to have a neutral effect. LDL-C concentrations tend to increase with rosiglitazone 5% to 15%, but do not significantly increase with pioglitazone. Both appear to convert small, dense LDL particles, which have been shown to be highly atherogenic, to large, fluffy LDL particles that are less dense. Large, fluffy LDL particles may be less atherogenic, but any increase in LDL must be of concern. Both drugs increase HDL, although pioglitazone may raise it more than rosiglitazone. TZDs also affect several components of the insulin resistance syndrome. PAI-1 levels are decreased, and many other adipocytokines are affected, endothelial function improves, and blood pressure may decrease slightly.

Microvascular Complications TZDs reduce HbA1c levels, which have been shown to be related to the risk of microvascular complications.

Macrovascular Complications Macrovascular complications with TZDs are controversial. In PROactive, the prospective pioglitazone clinical trial in macrovascular events, pioglitazone 45 mg was added to standard therapy in patients who had experienced a macrovascular event or had peripheral vascular disease.37 The two groups were well matched at baseline and the reported average observation time period was about 3 years. The primary end point (reduction in death, MI, stroke, acute coronary syndrome, coronary revascularization, leg amputation, and leg revascularization) was reduced 10% (P = 0.095). The main secondary end point (all-cause mortality, nonfatal MI, or stroke) was reduced 16% (P = 0.027). The seemingly dichotomous results relate to the inclusion of leg revascularization as a primary end point, which were increased in the pioglitazone group. Reasons for the increase are speculative, but may relate to more testing/inspection due to peripheral edema. Also of note, the pioglitazone group had 209 nonadjudicated admissions for heart failure occur versus 153 in the placebo group (P = 0.007), although fatal heart failure was not increased. Several published meta-analyses of rosiglitazone reported higher MI rates with rosiglitazone, but none have reported a higher risk of mortality. A hazard ratio (HR) of 1.43 (95% confidence interval [CI], 1.03 to 1.98; P = 0.03) for the risk of an MI with rosiglitazone versus other oral agents was reported.38

A prospective, multicenter, open-label noninferiority trial in 4,447 patients of rosiglitazone added to background metformin or sulfonylurea versus the active comparator metformin + sulfonylurea was recently reported (Rosiglitazone Evaluated for Cardiovascular Outcomes in Oral Agent Combination Therapy for Type 2 Diabetes [RECORD]). Rosiglitazone was noninferior to the comparator for all CV outcomes except for heart failure. A nonsignificant increase in risk for MI (HR, 1.14; 95% CI, 0.80 to 1.63) as well as a nonsignificant reduction in stroke (HR, 0.72; 95% CI, 0.49 to 1.05) was reported. On subset analysis, previous ischemic heart disease trended toward a higher risk (HR, 1.26; CI, 0.95 to 1.68; P = 0.055).39 Most studies with rosiglitazone trend toward, but do not reach, statistically significant increases in ischemic events. The FDA has placed rosiglitazone under a strict risk evaluation and mitigation program, limiting access to patients and prescribers who acknowledge and consent to knowing its macrovascular risks.

Adverse Effects Troglitazone, the first TZD approved, caused idiosyncratic hepatotoxicity and had deaths from liver failure, which prompted removal from the U.S. market. Newer TZDs do not have the same propensity, but have had postmarketing reports of liver injury. Patients with abnormal alanine aminotransferase (ALT) levels should be started with caution, and if the ALT is >3 times the upper limit of normal, especially if the total bilirubin is also >2 times the upper limit of normal, the medication should be discontinued. Pioglitazone has been shown in one well-designed trial to reduce hepatic steatosis, which may improve abnormal ALT levels in many patients with diabetes.

Retention of fluid leads to many different possible side effects with TZDs. The etiology of the fluid retention has not been fully elucidated, but appears to include peripheral vasodilation and/or improved insulin sensitization at the kidney with a resultant increase in renal sodium and water retention. A reduction in plasma hemoglobin (2% to 4%), attributed to a 10% increase in plasma volume, may result in a dilutional anemia that does not require treatment. Peripheral edema is also commonly (4% to 5% in monotherapy or combination therapy) reported. When a TZD is used in combination with insulin, the incidence of edema (~15%) is increased. TZDs are contraindicated in patients with New York Heart Association Class III and IV heart failure, and great caution should be exercised when given to patients with Class I and II heart failure or other underlying cardiac disease, as pulmonary edema and heart failure have been reported. Edema tends to be dose related and if not severe, a reduction in the dose as well as use of diuretics, anecdotally hydrochlorothiazide with triamterene, amiloride, or spironolactone instead of loop diuretics, will allow the continuation of therapy in the majority of patients. Rarely, TZDs have been reported to worsen macular edema of the eye.

Weight gain, which is also dose related, can be seen with both rosiglitazone and pioglitazone. Mechanistically, both fluid retention and fat accumulation play a part in explaining the weight gain. TZDs, besides stimulating fat cell differentiation, also reduce leptin levels, which stimulate appetite and food intake. Average weight gain varies, but a 1.5- to 4-kg weight gain is not uncommon. Rarely, a patient will gain large amounts of weight in a short period of time, and this may necessitate discontinuation of therapy. Weight gain positively predicts a larger HbA1c reduction, but must be balanced with the well-documented effects of long-term weight gain.

TZDs have also been associated with an increased fracture rate in the upper and lower limbs in women and men, although women appear to have a higher risk. These fractures are not osteoporitic in the classic sense, and do not occur in common osteoporosis fracture sites such as spine or hip. Most occur in wrists, forearms, ankles, or feet. Versus comparative diabetes therapy, TZDs may increase the risk of a fracture by 25%. The underlying pathophysiology is speculative, but may relate to TZD effects on the pluripotent stem cell and shunting of new cells to fat instead of osteocytes as well as altering osteoblasts/osteoclasts. It would be prudent to consider a patient’s risk factors for fractures if a TZD is being considered as antidiabetic therapy.

The risk of bladder cancer is slightly increased with pioglitazone, and likely rosiglitazone. Bladder tumors have been noted in rodent models using TZDs. An ongoing 10-year observational study reported an excess of 3 in 10,000 patient-years (from 7 to 10 in 10,000) risk of bladder cancer with pioglitazone at 5 years. Excess risk appears to be mostly in men and smokers, and is dose and duration associated. Mechanisms are speculative, but may involve microcrystals of the drug in the bladder that cause chronic irritation.

As a caution, premenopausal anovulatory patients may resume ovulation on TZDs. Adequate pregnancy and contraception precautions should be explained to all women capable of becoming pregnant, as both agents are pregnancy category C.

Drug Interactions Significant drug interactions that can cause clinical sequelae have not been noted with either medication. Neither pioglitazone nor rosiglitazone appears to be an inhibitor or inducer of CYP3A4/2C8 or CYP2C8/CYP2C9, respectively, although drugs that are strong inhibitors or inducers of these pathways (e.g., gemfibrozil or rifampin) may increase or decrease levels of active drug significantly. The package insert recommends limiting the dose of pioglitazone to 15 mg in combination with gemfibrozil.

Dosing and Administration The recommended starting dosages of pioglitazone and of rosiglitazone are 15 to 30 mg once daily and 2 to 4 mg once daily, respectively. Dosages may be increased slowly based on therapeutic goals and side effects. The maximum dose and maximum effective dose of pioglitazone is 45 mg, and rosiglitazone is 8 mg once daily, although 4 mg twice a day may reduce HbA1c by 0.2% to 0.3% (0.002 to 0.003; 2 to 3 mmol/mol Hb) more versus 8 mg once daily.

Rosiglitazone’s availability is limited for now, and an active risk evaluation and mitigation program has been implemented due to the risk of ischemic events. Patients and prescribers must sign up through the rosiglitazone website in order to receive the medication from a central mail-order pharmacy, as local pharmacies no longer carry rosiglitazone. Patients and prescribers must agree either that it is continuation of therapy and the risks versus benefits are known to both or that it is a new prescription and that the patient has been fully informed of the risk, including MI, and of the alternatives available, including pioglitazone. Pioglitazone is not included in this particular risk evaluation and mitigation program.

α-Glucosidase Inhibitors

Pharmacology Currently, there are two α-glucosidase inhibitors available in the United States: acarbose (Precose) and miglitol (Glyset). α-Glucosidase inhibitors competitively inhibit enzymes (maltase, isomaltase, sucrase, and glucoamylase) in the small intestine, delaying the breakdown of sucrose and complex carbohydrates. They do not cause any malabsorption of these nutrients. The net effect from this action is to reduce the postprandial BG rise. GLP-1 may also be increased. Distal intestinal degradation of undigested carbohydrate by the gut flora results in gas (CO2 and methane) and production of short-chain fatty acids, which may stimulate GLP-1 release from intestinal L cells.

Pharmacokinetics The mechanism of action of α-glucosidase inhibitors is limited to the luminal side of the intestine. Some metabolites of acarbose are systemically absorbed and renally excreted, whereas the majority of miglitol is absorbed and renally excreted unchanged.

Efficacy Postprandial glucose concentrations are reduced (40 to 50 mg/dL [2.2 to 2.8 mmol/L]), while fasting glucose levels are relatively unchanged (~10% reduction). Efficacy on glycemic control is modest (average reductions in HbA1c of 0.3% to 1% [0.003 to 0.010; 3 to 11 mmol/mol Hb]), affecting primarily postprandial glycemic excursions. Thus, patients near target HbA1c levels with near-normal FPG levels, but high postprandial levels, may be candidates for therapy.

Microvascular Complications α-Glucosidase inhibitors modestly reduce HbA1c levels, which has been shown to be related to the risk of microvascular complications.

Macrovascular Complications The STOP-NIDDM study, in subjects with IGT, reported a significant reduction in the risk of cardiovascular events, although the total number of events was very small.40,41 No large cardiovascular study confirming these preliminary results has been done in prediabetes or diabetes patients.

Adverse Effects The GI side effects, such as flatulence, bloating, abdominal discomfort, and diarrhea, are very common and greatly limit the use of α-glucosidase inhibitors. Mechanistically, these side effects are caused by distal intestinal degradation of undigested carbohydrate by the microflora, which results in gas (CO2 and methane) production. Microflora convert the carbohydrate to short-chain fatty acids that are mostly absorbed; thus, there is not a large calorie loss. α-Glucosidase inhibitors should be initiated at a low dose and titrated slowly to reduce GI intolerance. Beano, an α-glucosidase enzyme, may help to decrease GI side effects, but may decrease efficacy slightly, and it is better to decrease carbohydrate or the dose of the α-glucosidase inhibitor.

If a patient develops hypoglycemia within several hours of ingesting an α-glucosidase inhibitor, oral glucose is advised because the drug will inhibit the breakdown of more complex sugar molecules. Milk, with lactose sugar, may be used as an alternative when no glucose is available, as acarbose only slightly (10%) inhibits lactase. Fructose may also work, if the others are not available.

Rarely, elevated serum aminotransferase levels have been reported with the highest doses of acarbose. It appeared to be dose and weight related, and is the premise for the weight-based maximum doses.

Dosing and Administration Dosing for both miglitol and acarbose are similar. Initiate with a very low dose (25 mg with one meal a day); increase very gradually (over several months) to a maximum of 50 mg three times a day for patients ≤60 kg or 100 mg three times a day for patients >60 kg (see Table 57-11). Titration speed should be varied based on GI side effects to the target dose. Both α-glucosidase inhibitors should be taken with the first bite of the meal so that drug may be present to inhibit enzyme activity. Only patients consuming a diet high in complex carbohydrates will have significant reductions in glucose levels. α-Glucosidase inhibitors are contraindicated in patients with short-bowel syndrome or inflammatory bowel disease, and neither should be administered in patients with serum creatinine >2 mg/dL (>177 μmol/L), as this population has not been studied.

Dipeptidyl Peptidase-4 Inhibitors Sitagliptin (Januvia), saxagliptin (Onglyza), linagliptin (Tradjenta), and alogliptin (Nesina) are DPP-4 inhibitors currently approved in the United States.

Pharmacology DPP-4 inhibitors prolong the half-life of endogenously produced GLP-1 and GIP that normally is only minutes. GIP levels are normal in type 2 DM, and may contribute a minor amount of insulin secretion but have no effect on glucagon. However, levels of GLP-1 are deficient in type 2 DM. As these agents block nearly 100% of the DPP-4 enzyme activity for at least 12 hours, normal physiologic, nondiabetic GLP-1 levels are achieved. DPP-4 inhibitors significantly reduce the inappropriately elevated glucagon postprandially, although not back to nondiabetic levels, and improve insulin response to hyperglycemia. This results in reduction of glucose levels without increase in hypoglycemia when used as monotherapy. These drugs do not alter gastric emptying and do not have significant satiety effects. DPP-4 inhibitors also appear to be weight neutral.

Pharmacokinetics Sitagliptin has rapid absorption, with a tmax of approximately 1.5 hours. Absolute bioavailability after oral intake is approximately 87%. Only ~40% is bound to plasma proteins; the volume of distribution is approximately 200 L. The t1/2 of sitagliptin is approximately 12 hours. Seventy-nine percent of the dose of sitagliptin is excreted unchanged in the urine by active tubular secretion; however, the organic anion transporter 3 or p-glycoprotein transport may be involved as well. Sitagliptin exposure is increased by approximately 2.3-, 3.8-, and 4.5-fold relative to healthy subjects for patients with moderate renal insufficiency (creatinine clearance [CLcr] 30 to <50 mL/min [0.50 to <0.83 mL/s]), severe renal insufficiency (CLcr <30 mL/min [<0.50 mL/s]), and ESRD (on dialysis), respectively. This is not a safety or adverse reaction issue; however, reduction of the dose based on renal function is appropriate, as only 100% of the enzyme can be inhibited, and long-term exposure to higher levels in humans has not been extensively studied. Pharmacodynamically, DPP-4 inhibition appeared to mirror directly the plasma concentration of sitagliptin. Doses of 50 mg produce at least 80% inhibition of DPP-4 enzyme activity at 12 hours, and those of 100 mg produce 80% inhibition of DPP-4 enzyme activity at 24 hours. Food has no effect on absorption kinetics of sitagliptin. Hepatic impairment, age, gender, or race has no effect on the pharmacokinetics of sitagliptin.

When saxagliptin is administered with a high-fat meal, the tmax increases about 20 minutes and the AUC increases about 27%. However, this is not clinically significant and saxagliptin may be given with or without meals. The oral bioavailability of saxagliptin is approximately 67%. Distribution is similar to the body water compartment. There is negligible protein binding, and one active metabolite, 5-hydroxy saxagliptin, is half as potent a DPP-4 inhibitor as the parent compound, and contributes to activity. Metabolism is by the CYP3A4/5 system, and strong inhibitors or inducers will have an effect on activity. The half-lives of saxagliptin and its active metabolite are 2.5 and 3.1 hours, respectively. Approximately 25% of the dose is found in feces representing unabsorbed drug and bile excretion. Females have ~25% more exposure to 5-hydroxy saxagliptin, and exposure is increased 25% to 50% in elderly, likely due to renal clearance. The majority (75%) of saxagliptin and 5-hydroxy saxagliptin is renally eliminated, and some renal excretion is seen. In moderate (CLcr 30 to 50 mL/min [0.50 to 0.83 mL/s]) or severe (CLcr <30 mL/min [<0.50 mL/s]) renal impairment, saxagliptin and its active metabolite exposure are increased 2.1- and 4.5-fold, respectively.

The peak plasma concentrations of linagliptin occur at approximately 1.5 hours after oral administration of a single 5-mg dose to healthy subjects. The half-life of linagliptin is approximately 12 hours. The absolute bioavailability of linagliptin is approximately 30%. A high-fat meal reduces Cmax by 15% and increases AUC by 4%. However, this effect is not clinically relevant and linagliptin may be taken with or without food. Linagliptin distributes extensively in tissues and at high concentrations. Seventy percent to 80% is bound to plasma proteins while 20% to 30% is unbound in plasma. Plasma binding is not altered in patients who have renal or hepatic impairment. Following oral administration, metabolism is minimal and about 90% of linagliptin is excreted unchanged. Renal excretion is less than 5% of the administered dose and is not affected by decreased renal function.

Alogliptin has a bioavailability of approximately 100% and can be administered with or without food. It is only 20% bound to plasma proteins and approximately 75% of the dose is found unchanged in the urine. Less than 1% is metabolized to an active metabolite, and <6% to an inactive metabolite.

Efficacy The average reduction in HbA1c is approximately 0.7% to 1% (0.007 to 0.010; 8 to 11 mmol/mol Hb) at maximum dose. The decrease in HbA1c at different baseline values is very small. As these drugs are well tolerated, adjustment in the dose due to adverse effects is unlikely. They tend to have a shallow dose–response curve.

Microvascular and Macrovascular Complications HbA1c levels are reduced, which has been related to a reduction in microvascular complications, but no outcome data are available to date.

Drug–Drug Interactions Significant drug–drug interactions with sitagliptin are unlikely. Sitagliptin is metabolized approximately 20% by CYP450 3A4 with some CYP450 2C8 involvement, but is neither an inhibitor nor an inducer of any CYP450 enzyme system. It is a p-glycoprotein substrate, but had negligible effects on digoxin and cyclosporine A, increasing the AUC by only 30%.

Saxagliptin is metabolized by CYP3A4/5, and is a p-glycoprotein substrate, but is neither an inhibitor nor an inducer. Rifampin, an inducer, can decrease active levels by 50%. However, moderate to strong inhibitors or inducers of CYP3A4/5, such as diltiazem or ketoconazole, can increase the AUC of saxagliptin by approximately twofold, with a corresponding decrease in the formation of the active metabolite 5-hydroxy saxagliptin. In such situations, it is recommended that the dose of saxagliptin be limited to 2.5 mg daily.

Linagliptin is a weak to moderate inhibitor of CYP3A4, and a p-glycoprotein substrate. Thus, efficacy of linagliptin may be reduced when administered in combination with inducers of p-glycoprotein and CYP3A4 (e.g., with rifampin). If patients require the use of such drugs, the use of alternative treatments is recommended. No other significant drug interactions have been reported.

No significant drug–drug interactions with alogliptin have been noted to date.

Adverse Effects DPP-4 inhibitors are very well tolerated, are weight neutral, and do not cause GI side effects. Mild hypoglycemia may occur, but in monotherapy or in combination with medications that have a low incidence of hypoglycemia, DPP-4 inhibitors do not increase the risk of hypoglycemia. Headache and nasopharyngitis, potentially related to the drug, may be slightly more common with DPP-4 inhibitors, but no significant increases in peripheral edema, hypertension, or cardiac outcomes have been noted to date.

Urticaria and/or facial edema may be seen in approximately 1% of patients, and discontinuation is warranted. Rare cases of Stevens-Johnson syndrome have been reported.

In regard to long-term safety, DPP-4 enzymes metabolize a wide variety of peptides (PYY, neuropeptide Y, growth hormone–releasing hormone, vasoactive intestinal polypeptide, and others), potentially affecting other regulatory systems. DPP-4 (also known as CD26) plays an important role for T-cell activation. Theoretically the inhibition of DPP-4 could be associated with adverse immunologic reactions. Saxagliptin results in a dose-related reduction in absolute lymphocyte count in up to 0.5% to 1.5% of patients. In most, recurrence is not observed with reexposure. However, it may recur with rechallenge in some patients. The clinical relevance is not known, but if prolonged infection is encountered, it is logical to measure lymphocyte counts and consider discontinuation.

Dosing and Administration Sitagliptin is dosed orally at 100 mg daily unless renal insufficiency is present. The 50-mg dose is recommended if the CLcr is 30 to <50 mL/min (0.50 to <0.83 mL/s), or 25 mg if <30 mL/min (<0.50 mL/s). Saxagliptin is dosed orally 5 mg daily, unless the CLcr is <50 mL/min (<0.83 mL/s), or strong CYP3A4/5 inhibitors are used; then the recommended daily dose is 2.5 mg. Linagliptin is available only in one dose: 5 mg daily, and does not require dose adjustment in renal insufficiency or related to concomitant drug therapy. Alogliptin, similar to sitagliptin, has a two-step dosing adjustment in renal insufficiency. Alogliptin 25 mg daily should be decreased to 12.5 mg when the CrCl <60 mL/min (<1 mL/s), and 6.25 mg when <30 mL/min (<0.50 mL/s). Because of their excellent tolerability profile and flat dose–response curve, these drugs should be maximally dosed, unless noted above.

Bile Acid Sequestrants Currently, the only bile acid sequestrant approved for the treatment of type 2 DM is colesevelam (Welchol).

Pharmacology Colesevelam is a bile acid sequestrant that acts in the intestinal lumen to bind bile acid, decreasing the bile acid pool for reabsorption. Whether colesevelam’s mechanism of action to lower plasma glucose levels is in the intestinal lumen, a systemic effect due to the intestinal lumen effect or some combination of these two is unknown. Possible mechanisms include effects on the farnesoid X and TGR5 receptors within the intestine as well as effects on farnesoid X receptor within the liver. There is evidence that colesevelam may affect the secretion of GLP-1 and GIP. See also Chapter 11.

Pharmacokinetics Colesevelam is not absorbed from the intestinal lumen; thus, there is no absorption, distribution, or metabolism.

Efficacy HbA1c reductions from baseline (~8% [~0.08; ~64 mmol/mol Hb]) were approximately 0.4% (0.004; 5 mmol/mol Hb) when a dose of 3.8 g/day was added to stable metformin, sulfonylureas, or insulin. The FPG was modestly reduced about 5 to 10 mg/dL (0.3 to 0.6 mmol/L). Colesevelam also reduces LDL-C cholesterol in patients with type 2 DM. A 12% to 16% reduction in LDL-C was reported from baseline LDL-C concentrations of ~105 mg/dL (~2.72 mmol/L). Triglycerides increased when combined with sulfonylureas or insulin, but not with metformin. Colesevelam is weight neutral. Pediatric patients (10 to 17 years of age) have been studied for cholesterol reduction, but not for type 2 DM.

Microvascular Complications Bile acid sequestrants modestly reduce HbA1c levels, which have been shown to be related to the risk of microvascular complications.

Macrovascular Complications Although colesevelam lowers plasma glucose and LDL-C, it has not been proven to prevent cardiovascular morbidity or mortality.

Drug–Drug Interactions There are multiple absorption-related drug–drug interactions with colesevelam. The most important include levothyroxine, glyburide, and oral contraceptives. In addition, phenytoin, warfarin, digoxin, and fat-soluble vitamins have postmarketing reports of altered absorption. It is recommended that medications suspected of an interaction should be moved at least 4 hours prior to dosing the colesevelam. Colesevelam has also been implicated in the malabsorption of fat-soluble vitamins (A, E, D, K). In addition to the obvious fat-soluble vitamin supplementation, this may have implications for associated conditions. Other drugs that are very fat-soluble such as cyclosporine A, drugs that may be affected by a change in fat-soluble vitamin status such as warfarin and vitamin K, or conditions that may be potentially worsened by fat-soluble vitamin status such as some bleeding disorders or dermatologic conditions should be monitored.

Adverse Effects The most common side effects are GI. Constipation (11%) and dyspepsia (8%) are more common with colesevelam than placebo. Because of the constipating effects of colesevelam, it is not recommended in patients with gastroparesis, bowel obstruction, or a history of major GI surgery. It also should not be used in patients with significant swallowing or esophageal issues, as it may worsen the underlying condition or cause obstruction. Colesevelam should be taken with a large amount of water to lower the risk of the above issues. Hypoglycemia rates were low, although caution with insulin or sulfonylureas is prudent.

Colesevelam, similar to all bile acid sequestrants, may raise triglyceride levels. The increase in triglycerides is proportional to baseline triglyceride levels, and colesevelam is contraindicated in patients with a triglyceride >500 mg/dL (>5.65 mmol/L). Close monitoring is recommended if the baseline triglycerides >300 mg/dL (3.39 mmol/L). Colesevelam is contraindicated in patients with a history of triglyceridemia-induced pancreatitis, and prudent clinical judgment should be used in any patient with a history of pancreatitis and elevated triglycerides.

Dosing and Administration Dosing for type 2 DM is six 625-mg tablets daily (total dose/day = 3.75 g), which may be split into three tablets two times a day if desired. A 3.75-g oral suspension packet, dosed daily, or a 1.875-g oral suspension packet dose twice daily is also available. Suspension packets must be diluted in a minimum of one half to one cup of water. Take tablets and suspension with a large amount of water, if possible. All dosage forms should be administered with meals as colesevelam binds to bile released during the meal.

Dopamine Agonists Bromocriptine mesylate (Cycloset) is currently approved for the treatment of type 2 DM.

Pharmacology Bromocriptine is a dopamine agonist, but the exact mechanism of how bromocriptine improves glycemic control is unknown. Low hypothalamic dopamine levels, especially on waking, are augmented, which may decrease sympathetic tone and output. These effects are speculated to improve hepatic insulin sensitivity.

Pharmacokinetics Bioavailability is 65% to 95% after an orally administered dose; bioavailability may be increased ~50% if given with a meal. Peak plasma concentration is about 1 hour if taken without food, but with food it is 90 to 120 minutes. Bromocriptine is highly protein bound, and has a volume of distribution of 61 L. Only ~7% reaches the systemic circulation due to GI-based metabolism and first-pass metabolism. Bromocriptine is extensively metabolized by the CYP3A4 pathway, and the majority (~95%) is excreted in the bile. The half-life is approximately 6 hours. Plasma exposure is increased in females by approximately 18% to 30%, but no dosage adjustment is currently recommended.

Efficacy In clinical trials, bromocriptine mesylate reduced HbA1c by 0.3% to 0.6% (0.003 to 0.006; 3 to 7 mmol/mol Hb) from baseline.

Microvascular and Macrovascular Complications There is no study examining microvascular disease. Macrovascular event reduction has not been proven. In just over 3,000 subjects, bromocriptine decreased a composite cardiovascular outcome over 1 year. The composite outcome occurred in 37 (1.8%) bromocriptine-treated subjects versus 32 (3.2%) subjects not given bromocriptine (HR 0.6 [95% two-sided CI, 0.35 to 0.96]).

Drug–Drug Interactions Bromocriptine is extensively metabolized by CYP3A4 and strong inhibitors or inducers may change bromocriptine levels. As bromocriptine is highly protein bound, it may increase the unbound fraction of other highly protein-bound drugs. Several drug–drug and potential drug–disease interactions are present including antipsychotics in psychotic disorders as they decrease dopamine activity, atypical antipsychotics as they may decrease the effectiveness of bromocriptine, and ergot-based therapy for migraines as bromocriptine may increase migraine and ergot-related nausea and vomiting. There are case reports of hypertension and tachycardia when administered with sympathomimetic drugs in postpartum women, and bromocriptine should not be given to this group of potential patients. The effectiveness in other disease states where dopamine agonism may be indicated is unknown.

Adverse Effects Adverse reactions leading to discontinuation occurred in 24% of bromocriptine patients compared with 9% in the placebo comparator group. Nausea, rhinitis, headache, asthenia, dizziness, constipation, and sinusitis all occurred in over 10% of subjects. Nausea occurred in 25% to 35% of patients, and vomiting, which tended to be more common in women, occurred in 5% to 6% of patients. Nausea, vomiting, fatigue, headache, and dizziness were common adverse events during the titration phase of phase 3 studies, and only 70% of completors could be titrated to the maximum dose. Orthostatic hypotension or syncope occurred in 2.2% and 1.4%, and 0.6% and 0.8% in the bromocriptine and placebo groups, respectively. No predisposing factors were identified, but caution should be exercised in patients with low, normal blood pressure. Somnolence was reported in 4.3% of patients on bromocriptine, compared with 1.3% in the placebo, and response to the drug should be ascertained prior to operating machinery or combining with other sedating medications. Psychiatric disorders including hallucinations and pathologic gambling have been reported with other forms of bromocriptine, but were not seen in phase III trials.

Dosing and Administration Bromocriptine is dosed with 0.8-mg tablets administered within 2 hours of waking from sleep daily with food. From 0.8 mg daily, the dose may be increased weekly based on response by 0.8-mg tablet increments, to a maximum of 4.8 mg daily (0.8 mg × 6 tablets, although it is unclear if another commercial dose could be made available). The minimal effective dose is 1.6 mg daily. It is recommended to be taken with food as this may decrease nausea/vomiting.

Potential Future Medications

Many medications for the treatment of diabetes are currently in late-phase development. No guarantee of FDA approval is given for any agent in development.


Insulin degludec is a long-acting basal insulin that appears to be truly peakless. Hypoglycemia in clinical trials has been slightly less to date versus insulin glargine with similar glycemic control.

Incretin Class Medications

Once-daily lixisenatide and several weekly GLP-1 receptor agonist medications are being developed. Closest to market is albiglutide, but also in development is semaglutide.

Selective Sodium-Dependent Glucose Cotransporter-2 Inhibitors (SGLT-2 Inhibitors)

SGLT-2 inhibitors work in the kidney to block the reabsorption of some glucose. Normally, all glucose is reabsorbed back into the systemic circulation from the kidney at normal glucose levels: about 10% through the SGLT-1 receptor, and 90% through the SGLT-2 receptor. Early data have shown approximately 50 to 80 g of glucose per day may be allowed to pass into the urine with SGLT-2 inhibition. This lowers systemic glucose and allows weight loss. Glucose levels may be lowered in both type 1 and type 2 DM by this mechanism of action. Safety data have shown a slightly higher rate of genitourinary yeast infections. SGLT-1 is involved with glucose absorption in the gut, and inhibition of SGLT-1 has been historically thought to cause GI toxicity, but this is unclear and dual inhibitors may be marketed. Canagliflozin (Invokana) is currently the farthest in development, but several are being developed. Dapagliflozin has been rejected by the FDA several times due to concerns about cancer.

Pivotal Trials

Diabetes Control and Complications Trial

Images Much of the last century in diabetes care was dominated by the debate over whether glycemic control actually was causative in complications of DM. Animal studies and some human studies suggested that the worse the glycemia, the greater the risk of complications. But “the glucose hypothesis” was not ultimately accepted as proven until the publication of the DCCT in 1993.23 In this study, 1,441 patients with type 1 DM were divided into two groups: those without complications (726 subjects, primary prevention) and those with early microvascular complications (715 subjects, secondary prevention). These two groups were then again divided into two groups: one randomized to receive conventional therapy (one or two shots of insulin daily and infrequent SMBG with no attempt to change therapy based on home BG readings) and the other to receive intensive therapy (>3 injections of insulin daily or insulin pump, with frequent SMBG and alteration of insulin therapy based on SMBG results, plus frequent contact with a health professional). After 6.5 years, mean follow-up with a difference in HbA1c between the two groups being ≈2% (≈0.02; 22 mmol/mol Hb) (≈9% vs. ≈7% [≈0.09 vs. ≈0.07; ≈75 mmol/mol Hb vs. ≈53 mmol/mol Hb), retinopathy was decreased by 76% in the primary prevention cohort, with retinopathy progression reduced 54% in the secondary prevention group. Neuropathy was decreased by 60% in both groups combined. Microalbuminuria was decreased 39%, while macroproteinuria was reduced 54% with intensive therapy. Hypoglycemia was more common and weight gain greater with intensive therapy. A nonstatistically significant reduction in coronary events was seen in the intensively treated group as compared with the conventional group. The DCCT revolutionized therapy of DM, demanding that stricter glycemic control be the goal.

United Kingdom Prospective Diabetes Study

Images The UKPDS was a landmark study for the care of patients with type 2 DM, confirming the importance of glycemic control for reducing the risk of microvascular complications.24 More than 5,000 patients with newly diagnosed type 2 DM were entered into the study. Patients were followed for an average of 10 years. The major portion of the study assessed “conventional therapy” (no drug therapy unless the patient was symptomatic or had FPG >270 mg/dL [>15 mmol/L]), versus intensive therapy starting with either sulfonylureas or insulin, aimed at keeping the FPG <108 mg/dL (6 mmol/L). A subset of obese patients was studied using metformin as the primary therapeutic agent.

Significant findings from the study include the following:

   1. Microvascular complications (predominantly the need for laser photocoagulation on retinal lesions) are reduced by 25% when median HbA1c is 7% (0.07; 53 mmol/mol Hb) as compared with 7.9% (0.079; 63 mmol/mol Hb).24

   2. A continuous relationship exists between glycemia and microvascular complications, with a 35% reduction in risk for each 1% decrement in HbA1c (0.01; 11 mmol/mol Hb). No glycemic threshold for microvascular disease exists.42

   3. Glycemic control has minimal effect on macrovascular disease risk. Excess macrovascular risk appears to be related to conventional risk factors such as dyslipidemia and hypertension.43

   4. Sulfonylureas and insulin therapy do not increase macrovascular disease risk.24

   5. Metformin reduces macrovascular risk in obese patients.36

   6. Vigorous blood pressure control reduces microvascular and macrovascular events.43 There was no evidence for a threshold systolic blood pressure above 130 mm Hg for protection against complications. β-Blockers and angiotensin-converting enzyme (ACE) inhibitors appear to be equally efficacious.44

Long-Term Follow-up of DCCT (EDIC) and UKPDS

At the conclusion of the DCCT and UKPDS trials, willing subjects continued to be followed over time to ascertain microvascular and macrovascular outcomes. In the follow-up of the DCCT, called the Epidemiology of Diabetes Interventions and Complications (EDIC), several important points have been discovered. First, HbA1c levels between conventional and intensive groups converged to an HbA1c of approximately 8% (0.08; 64 mmol/mol Hb). Despite similar HbA1c values, continued microvascular and new macrovascular benefit was seen. Microvascular benefits were maintained for 10 to 15 years, and at 17 years of follow-up, a 57% (P = 0.02) reduction in death, first occurrence of nonfatal MI, and stroke was seen between the conventional and intensive groups.45,46

In the follow-up of the UKPDS, HbA1c (~8% [~0.08; ~64 mmol/mol Hb]) converged and values were not significantly different for the majority of the follow-up. After a mean follow-up of 8.5 years, a 24% (P = 0.001) reduction in microvascular complications (during the UKPDS, 25% reduction was reported) and a significant reduction in MI (15%; P = 0.014) and all-cause mortality (13%; P = 0.007) in the intensively treated group versus the conventional group were reported.47 Early glycemic control may have long-standing benefits to patients over several decades, despite later glycemic control deterioration. This concept is being called metabolic memory or the legacy effect. These studies lay the framework for why early intensive glycemic control is important not only for short-term but also for long-term prevention of complications.


Images The Action to Control Cardiovascular Risk in Diabetes (ACCORD),48 Action in Diabetes and Vascular Disease (ADVANCE),49 and Veterans Affairs Diabetes Trial (VADT)50 were three trials that reported on the effects of glycemic control and macrovascular disease risk.

ACCORD randomized 10,251 high CVD risk subjects (CVD event or significant risk) with type 2 DM to intensive glycemic control (goal HbA1c <6% [<0.06; <42 mmol/mol Hb]) or standard glycemic control (HbA1c 7% to 7.9% [0.07 to 0.079; 53 to 63 mmol/mol Hb]). Multiple oral agents and/or insulin were allowed to achieve glycemic goals. Baseline HbA1c level was 8.1% (0.081; 65 mmol/mol Hb), and the intensive glycemic group achieved a HbA1c of 6.4% (0.064; 46 mmol/mol Hb), whereas standard glycemic control achieved a HbA1c of 7.5% (0.075; 58 mmol/mol Hb) when the study stopped after a mean follow-up period of 3.5 years. The study was stopped after an interim analysis reported an increased rate of mortality in the intensive arm (1.41%/y vs. 1.14%/y; HR, 1.22; 95% CI, 1.01 to 1.46). Interestingly, the primary end point (myocardial, stroke, or cardiovascular death) was trending down due to a lower risk of nonfatal MI in the intensive therapy group. In addition, on subset analysis, individuals with a baseline HbA1c <8% (<0.08; <64 mmol/mol Hb) or no previous CVD had a significant reduction in the primary outcome. Increased mortality, though substantially increased in the intensive group, could not be associated with a specific medication, hypoglycemia (higher in intensive group), lipid levels, or weight gain. The dichotomous results have been hard to explain, although the ACCORD investigators reported that in the intensive group, it was subjects who could not attain intensive glycemic control goals who were at higher risk, not subjects who did achieve the goal. A 20% higher risk of death for each 1% (0.01; 11 mmol/mol Hb) above an HbA1c of 6% (0.06; 42 mmol/mol Hb) was reported.

ADVANCE randomized 11,140 subjects to intensive (≤6.5% HbA1c [≤0.065; ≤48 mmol/mol Hb]) or standard therapy (investigator-driven goals). Extended-release gliclazide, a sulfonylurea available outside of the United States, was used as first-line therapy in the intensive group versus no gliclazide in the standard therapy group, although multiple other agents were needed in both groups. A baseline HbA1c was only 7.5% (0.075; 58 mmol/mol Hb), and at the end of therapy, the intensive group versus standard group HbA1c was ~6.5% versus ~7.2% (~0.065 vs. ~0.072; 48 mmol/mol Hb vs. 55 mmol/mol Hb). ADVANCE reported a significant reduction in renal events, including new or worsening nephropathy (HR, 0.79; 95% CI, 0.66 to 0.93), but no difference in major macrovascular events (HR, 0.94; 95% CI, 0.84 to 1.06) with intensive versus standard therapy.

VADT randomized 1,791 subjects to intensive glycemic control (HbA1c goal <6% [<0.06; <42 mmol/mol Hb], and action required of >6.5% [>0.065; >48 mmol/mol Hb]) versus nonintensive glycemic control (investigator determined). At entry, the HbA1c was the highest of the three trials (9.4% [0.094; 79 mmol/mol Hb]). Multiple medications including insulin were used to achieve glycemic control. The intensive group achieved an HbA1c of 6.9% versus 8.5% (0.069 vs. 0.085; 52 mmol/mol Hb vs. 69 mmol/mol Hb) in the investigator-determined group. The primary end point of nonfatal MI, nonfatal stroke, CVD death, hospitalization for heart failure, and revascularization was not significantly different (HR, 0.88; 95% CI, 0.74 to 1.05) and mortality was unchanged.

These three trials should be viewed as confirmatory that short term (3 to 5 years) of intensive glycemic control does not positively affect the risk of macrovascular risk in type 2 DM. ACCORD reported that a subset of subjects who could not achieve intensive glycemic control may be at higher risk of death, but identifying these patients and implementing this recommendation into clinical practice may prove to be challenging. As previously mentioned in Long-Term Followup of DCCT (EDIC) and UKPDS above, reduction of macrovascular events from improved glycemic control may take over a decade to come to fruition.


Images Knowledge of the patient’s quantitative and qualitative meal patterns, activity levels, pharmacokinetics of insulin preparations and other injectables, and pharmacology of oral and antidiabetic agents for type 2 DM is essential to individualize the treatment plan an optimize BG control while minimizing risks for hypoglycemia and other adverse effects of pharmacologic therapies.

Type 1 Diabetes Mellitus

All patients with type 1 DM require insulin. However, how that insulin is delivered to the patient is a matter of considerable practice difference among patients and clinicians.

Historically, after the discovery of insulin by Banting and Best in 1921, frequent injections of regular insulin (initially the only insulin available) were given. Modifications of insulin led to longer-acting insulin suspensions and the use by many patients of one or two injections of longer-acting insulin each day. Because self-monitored BG and HbA1c testing were not available at that time, patients and practitioners had no idea how well their patients’ BG concentrations were controlled, other than a vague sense from an indirect method, measurement of glucose in the urine. While the renal threshold for glucose is relatively predictable in young healthy subjects, it is highly variable in older patients and patients with renal disease. The advent of SMBG and HbA1c testing in the 1980s revolutionized the care of diabetes, enabling patients and practitioners to directly access BG for assessment, and enabling the patient to make instantaneous changes in the insulin regimen if need be. Modern diabetes management would be impossible without these two tools.

Contemporary management of type 1 DM attempts to match carbohydrate intake with glucose-lowering processes, most commonly insulin, as well as with exercise. The goal is to allow the patient to live as normal a life as possible. Understanding the principles of glucose input and glucose egress from the blood allows the practitioner and the patient great latitude in the management of type 1 DM.

Normal secretion of insulin can be divided into a relatively constant background level of insulin (“basal”) during the fasting and postabsorptive period, with prandial spikes of insulin after eating (“bolus”) (Fig. 57-6).51 Insulin sensitivity and insulin secretion are not constant throughout the day, however, which renders the concept of stable basal insulin requirements to be inaccurate. However, in most clinical situations, attempting to emulate normal secretion of insulin is a useful paradigm for understanding and applying insulin treatment for the management of type 1 DM. The other basic principle to consider is that the timing of insulin onset, peak, and duration of effect must match meal patterns and exercise schedules to achieve near-normal BG values throughout the day.


FIGURE 57-6 Relationship between insulin and glucose over the course of a day and how various insulin and amylinomimetic regimens could be given. (A, aspart; CSII, continuous subcutaneous insulin infusion; D, detemir; G, glargine; GLU, glulisine; L, lispro; P, pramlintide; N, NPH; R, regular.)

Historically, the complexity of insulin regimens was related to the number of injections of insulin administered per day. This was a reasonable classification; however, a single injection of any insulin preparation daily will in no way mimic normal physiology, and therefore is unacceptable. Similarly, two injections of any insulin daily will fail to replicate normal patterns of insulin release.

Injection regimens that begin to approximate physiologic insulin release start with “split-mixed” injections consisting of a morning dose of an intermediate-acting insulin such as NPH and a “bolus” rapid-acting insulin or regular insulin before breakfast, and again before the evening meal. The presumption is made that the morning intermediate-acting insulin gives basal insulin for the day and covers the midday meal, the morning bolus insulin covers breakfast, the evening intermediate-acting insulin gives basal insulin for the rest of the day, and the evening bolus insulin covers the evening meal. If patients are very compulsive about consistency of timing of their injections and meals and intake of carbohydrate, such a strategy may be acceptable. However, the vast majority of patients are not sufficiently predictable in their schedule and food intake to allow “tight” glucose control with such an approach.

A modification that can be made to the above regimen is the movement of the evening NPH to bedtime (now three total injections per day) because the fasting glucose in the morning is too high or there is hypoglycemia in the early hours of sleep. This approach improves glycemic control and may reduce hypoglycemia sufficiently for those patients who decline or are unable to follow more intense regimens. However, most patients with type 1 DM need a more intense approach that also allows greater flexibility in their lifestyle.

Images The basal–bolus concept attempts to replicate normal insulin physiology with a combination of intermediate- or long-acting insulin to provide the basal component, and rapid-acting insulin to provide the bolus or premeal component. Various long-acting insulins have been used to provide the basal insulin component, including once- or twice-daily NPH, detemir, or glargine. Insulin glargine and insulin detemir are the most feasible basal insulins for most patients with type 1 DM.

The bolus or prandial insulin component can be regular insulin, insulin lispro, insulin aspart, or insulin glulisine injected before meals. The rapid onset of action and short time course of rapid-acting insulin analogs more closely replicate normal physiology than does regular insulin. The patient varies the amount of before meal rapid-acting insulin injected, depending on the preprandial BG level, the anticipated activity (upcoming exercise may reduce insulin requirement), and anticipated carbohydrate intake. Many patients start with a prescribed dose of insulin before meals that they vary by use of an “adjusted scale insulin” or “correction factor” to normalize a high premeal plasma glucose reading. Patients on more advanced regimens later may adjust the amount of mealtime insulin based on anticipated carbohydrate intake.

A “correction factor” can be calculated as a starting point to estimate the approximate plasma glucose–lowering effect of 1 unit of short-acting insulin in mg/dL. For regular insulin, one may use a factor of 1,500 (a corresponding factor for calculation of glucose in SI units would require multiplying by 0.0555) divided by the total daily insulin dose in number of units that the patient currently uses. For rapid-acting insulin analogs, a factor of 1,700 is more often used when calculating the correction factor. For example, if a patient is currently taking 40 units of basal insulin and 12 units of rapid-acting insulin at each of three meals, the total daily insulin dose equals 76 units. Using this calculation 1,700 divided by 76 equals 22; thus, each unit of rapid-acting insulin analog will lower the plasma glucose approximately 22 mg/dL (1.2 mmol/L). Review of follow-up BG data permits better individualization of the correction factor.

Carbohydrate counting is a very effective tool for determining the amount of rapid-acting insulin that should be injected preprandially in people with type 1 DM. Instead of using a prescribed or preset dose of rapid-acting insulin before meals, patients can self-adjust their premeal dose based on the estimated grams of carbohydrates that will be consumed. Although general algorithms for carbohydrate counting give rough guidelines, each patient will have to adjust the preprandial insulin dosage based on his or her own individual response to different food items.

One method of calculating how much carbohydrate (grams) 1 unit of rapid-acting insulin will cover is to use 500 divided by the total daily dose of insulin in number of units. Therefore, using the example above with a total daily insulin dose of 76 units, we would use 500 divided by 76, which estimates that 1 unit of rapid-acting insulin will cover approximately 7 g of carbohydrate. Review of follow-up BG data before and 2 hours after meals will enable more precise determination of an individual’s insulin-to-carbohydrate ratio.

In type 1 DM, approximately 50% of total daily insulin replacement should be basal insulin, and the other 50% will be bolus insulin, divided into doses before meals. If the patient’s ratio is not close to this recommendation, a reassessment of the regimen should be implemented. Empirically, patients may be begun on ≈0.6 unit/kg/day with basal insulin 50% of total dose and prandial insulin 20% of total dose prebreakfast, 15% prelunch, and 15% presupper. Type 1 DM patients generally require between 0.5 and 1 unit/kg/day. The need for significantly higher amounts of insulin suggests the presence of insulin resistance or, less often, of insulin antibodies.

Intensive basal–bolus multi-injection insulin therapy is recommended for all adult patients with type 1 DM at the time of diagnosis to reinforce the importance of glycemic control from the outset rather than change strategies over time because of lack of control. Occasional patients with an extended honeymoon period may need less intense therapy initially, but should later be converted to basal–bolus therapy at the onset of glycemic lability.

For those patients who insist on only two injections daily, intermediate-acting insulin and a rapid-acting insulin or regular insulin (starting at 0.6 unit/kg with two thirds in the morning, two thirds of the morning dose as intermediate-acting insulin, and one half of evening dose as intermediate-acting insulin) is an option; however, most often this approach will not be allowed as an aggressive glycemic control option due to increased risk of hypoglycemia.

Insulin pump therapy (continuous subcutaneous insulin infusion [CSII], generally using a rapid-acting analog insulin) is the most sophisticated form of insulin delivery. In a motivated patient, CSII may be more efficacious in achieving excellent glycemic control than multiple-dose insulin injections. Extensive discussion of this mode of therapy is beyond the scope of this text. Nevertheless, the basic principles for implementation are the same.

One advantage of pump therapy is that the basal insulin dose may be varied, related to changes in basal insulin requirements throughout the day. In selected patients, this feature allows better glycemic control with CSII. However, insulin pumps require even greater attention to detail and frequency of SMBG than does a basal–bolus regimen with four injections daily.52 In appropriately selected patients willing to pay sufficient attention to detail of SMBG and insulin administration, CSII can be a very effective form of therapy. CSII is only a tool for diabetes control, however. Thus, if the patient is not well controlled and/or unwilling to actively control the diabetes on injections, it is unlikely that the patient will have superior control on a pump. CSII placement and adjustment should be made by an experienced clinician, and only after a discussion with the patient about the reality of CSII, addressing expectations, and proper training on the pump.

Regardless of the insulin regimen chosen, gross adjustments in the total insulin dose are made based on HbA1c measurements and symptoms such as polyuria, polydipsia, and weight gain or loss. Finer insulin adjustments are determined on the basis of the results of frequent BG monitoring, documentation of mealtime carbohydrate intake, physical activity, and other factors that affect glycemic control.

All patients should have extensive education in the recognition and treatment of hypoglycemia. Many patients experiencing hypoglycemia are tempted to overtreat episodes of hypoglycemia resulting in rebound hyperglycemia afterwards. To minimize this, patients are advised to follow the “rule of 15.” If hypoglycemia is identified (BG less than 70 mg/dL [3.9 mmol/L]), the patient is instructed to consume 15 g of simple carbohydrate (8 oz [~250 mL] orange juice or four glucose tablets) and then retest his or her BG 15 minutes later. If BG is still less than 70 mg/dL (3.9 mmol/L), the patient may repeat the rule of 15 until his or her BG has normalized.

At each visit, patients with type 1 DM should be questioned about hypoglycemia. The frequency of hypoglycemia, particularly hypoglycemia requiring assistance of another person, a visit to an emergent or urgent care facility, or hospitalization, should be recorded.

In type 1 DM, it is common for patients to develop hypoglycemia unawareness. Hypoglycemic unawareness may result from progression of disease with autonomic neuropathy. The loss of warning signs of hypoglycemia is a relative contraindication to intensive therapy. More commonly, type 1 DM patients have loss of warning signs because of a presumed lower set point for release of counterregulatory hormones as a result of frequent episodes of hypoglycemia (“hypoglycemia begets hypoglycemia”). In such situations, more normal hypoglycemia awareness may be restored by reduction or redistribution of the insulin dose to eliminate significant and/or frequent hypoglycemic episodes.

In children and pubescent adolescents, glycemic goals may need to be tempered with the risks of hypoglycemia. Table 57-7 lists glycemic goals for different age groups of type 1 DM patients. Therefore, it is not unreasonable to use less intense management until the patient is postpubertal, if age-specific goals can be maintained.5

Occasional patients develop antibodies to injected insulin, but the significance of the antibodies is usually minimal. Human insulin therapy has not totally eliminated insulin allergies. In most patients local reactions will dissipate over time. If mild reactions at the site of injection occur, reassess the insulin injected. Many times the patient is injecting cold insulin, which may cause compensatory local vasodilation around the injection site in response to the injected cold liquid. Anecdotally, a different type or source of insulin could be tried. If the allergic reaction does not improve or is systemic, insulin desensitization can be carried out. Protocols for desensitization are available from major insulin manufacturers.

While more common in the animal insulin era, lipohypertrophy is still seen in some patients with long-standing type 1 DM. Some patients give their insulin injections in the same site repeatedly to minimize discomfort; over time this can result in lipohypertrophy. Lipohypertrophy can sometimes be visualized on physical examination and also can be identified by palpation of injection sites. Because insulin absorption from an area of lipohypertrophy is unpredictable, it is mandatory to avoid insulin injections into these areas.

There are several common errors in the management of patients with type 1 DM that can cause erratic glucose fluctuations:

   1. Failure to take into account action of insulin: The timing of meals and/or physical activity must be planned around the peaks of insulin action accordingly.

   2. Choice of insulin injection sites: There is variability of insulin absorption from site to site such that random selection of insulin injection sites may cause wide glucose swings. The most consistent absorption of insulin is from the abdominal wall. Patients are encouraged to take all their injections in the abdomen. If the patient is unable or unwilling to follow this advice, then systematic site rotation is the next preferable option. The patient should always give the insulin injection in the same region of the body the same time of the day each day. For instance, the arms are always used every morning. Needless to say, the patient should not inject in a limb and then go out and exercise that limb, which could cause increased blood flow and insulin absorption.

   3. Overinsulinization: The answer to all high BG is not necessarily more insulin. Hyperglycemia could be due to too little insulin or it could be due to “rebounding” from a previous low glucose and treating it with excessive amounts of carbohydrate. Fastidious BG testing, particularly during the night (or selected use of CGM), can assist in sorting this out. Many clinicians do not adequately differentiate type 1 DM from type 2 DM when choosing doses of insulin. Patients with type 1 DM are insulin deficient but have normal insulin sensitivity. Patients with type 2 DM have varying degrees of insulin resistance. Therefore, a small change in the dose of insulin for a patient with type 1 DM can have a dramatic effect on glucose concentrations, whereas in patients with type 2 DM and insulin resistance a change in dose many times that amount of insulin has little effect on glucose concentrations. Large changes in insulin dose in patients with type 1 DM are not usually indicated unless the patient’s BG control is very poor. Widely erratic BGs and/or weight gain may be due to too high a dose of insulin.

   4. Injection technique and BG monitoring: When in doubt, always reevaluate the patient’s technique for insulin dosing, insulin injection, and BG testing. Sometimes simple errors result in unpredictable glycemic control.

Type 1 DM patients who continue to have erratic postprandial control despite implementation of the above strategies may be appropriate for treatment with pramlintide (Symlin). Pramlintide is not recommended to be mixed with insulin; therefore, the patient will need to take an additional injection at each meal. With initiation of pramlintide the doses of prandial insulin (rapid-acting analog or regular insulin) should be reduced by 30% to 50%, to prevent hypoglycemia. Pramlintide should be titrated based on GI adverse effects and postprandial glycemic goals. Injecting pramlintide prior to the meal and the rapid-acting insulin at the time of or after the meal may better match the appearance of the food with the postprandial increase in glucose due to delayed gastric emptying. The patient must be cognizant of the risk of hypoglycemia, GI side effects, and how to reduce both.

Islet cell and whole pancreas transplantation is occasionally used in patients who require immunosuppressive therapy for other reasons, such as renal transplants.53 Many patients are able to stop insulin and/or only require insulin secretion support therapy with sulfonylureas or GLP-1 agonists. However, within 2 years as many 80% or more will need to reinitiate some form of insulin therapy.

Type 2 Diabetes Mellitus

Images Pharmacotherapy for type 2 DM has changed dramatically in the last few years with the addition of several new drug classes and recommendations to achieve more stringent glycemic control. Symptomatic patients may initially require treatment with insulin or combination oral therapy to reduce glucose toxicity (which may reduce β-cell insulin secretion and worsen insulin resistance). Patients with HbA1c ≈7% (≈0.07; ≈53 mmol/mol Hb) or less are usually treated with therapeutic lifestyle measures and an agent that will not cause hypoglycemia. Those with HbA1c >7% but <8.5% (>0.07 but <0.085; >53 but <69 mmol/mol Hb) could be initially treated with single oral agents, or combination therapy. Patients with higher initial HbA1c may benefit from initial therapy with two oral Images agents, or insulin. This section addresses management of hyperglycemia; however, this needs to be balanced within a multifactorial risk reduction framework of blood pressure reduction, dyslipidemia and antiplatelet therapy, and smoking cessation. All therapeutic decisions should consider the needs and preferences of the patient, if feasible. Individualization of therapy is necessary for success.

Depending on patient motivation and adherence to therapeutic lifestyle changes, most patients with HbA1c greater than 9% to 10% (0.09 to 0.10; 75 to 86 mmol/mol Hb) will likely require therapy with two or more oral agents to reach glycemic goals. Treatment of type 2 DM often necessitates use of multiple therapeutic agents (combination therapy), to obtain glycemic goals.

The best oral therapy regimen for patients with type 2 DM is widely debated. Based on the results of the UKPDS and safety record, obese patients (>120% ideal body weight) without contraindications should be started on metformin titrated to ≈2,000 mg/day.5,54 Near-normal-weight patients may be better treated with insulin secretagogues, although metformin will work in this population. Metformin is the only oral antihyperglycemic agent to ever report a reduction in total mortality. Despite this, long-term durability of HbA1c reduction, due to the inability to stop progressive β-cell failure, is suboptimal with metformin, and patients over several years will often need additional therapy. An insulin secretagogue, such as a sulfonylurea, is often added second, although it has clearly been shown that sulfonylureas do not produce durable HbA1c reductions in the majority of patients. Better choices to sustain HbA1c reductions would be a TZD or GLP-1 agonist, but each has limitations as well. Goal-oriented therapy is what we currently strive for, meaning the intervention should be in relation to the distance from the glycemic goal. When initial therapy is no longer keeping the patient at goal, if the HbA1c is close to goal, one additional agent may be appropriate. If >1% to 1.5% (>0.01 to 0.015; >11 to 16 mmol/mol Hb) above goal, it is unlikely any one oral agent will result in reaching the glycemic goal, and multiple oral agents or insulin therapy may be appropriate. TZDs may be substituted in situations in which a patient is intolerant of, or has a contraindication to, metformin, understanding that TZDs should be used with caution in heart failure. Figure 57-7 is a consensus algorithm by the ADA and the European Association for the Study of Diabetes.54 No algorithm can substitute for good clinical judgment, and algorithms for glycemic control start with the premise that the clinician will identify medication contraindications, adverse reactions, and comorbidities that may be advantageous or harmful if the medication was taken.


FIGURE 57-7 Position Statement on the Treatment of Type 2 Diabetes Mellitus: American Diabetes Association and European Association for the Study of Diabetes. (Adapted from reference 54.)

We should also treat type 2 DM by matching therapy to the suspected underlying problem. Consider some simple questions to guide therapy: (a) How long has the patient had diabetes? The longer a patient has had diabetes, the more insulinopenic he or she likely is and the more likely that insulin therapy will be needed. (b) Fasting, postprandial, or both plasma glucose readings poorly controlled? Some drugs address postprandial glucose excursions better, whereas some address FPG better. (c) How far do we have to go to goal and what is the goal? Each oral agent has limits on HbA1c reduction, and the reduction is baseline HbA1c dependent. (d) Adverse effect profile? Contraindications, hypoglycemia potential, and tolerability are based on the current status of the patient. (e) Comorbidities? CVD, dementia, life expectancy, depression, osteoporosis, and other conditions where select medications may be poor choices and additionally those comorbidities may drive our HbA1c goal. Based on the ADVANCE, ACCORD, and VADT trials, a HbA1c goal may now be above 7% (0.07; 53 mmol/mol Hb) if certain comorbidities are present. See Figure 57-8 for HbA1c individualization based on comorbidities from the Texas Diabetes Council. Drugs such as metformin, TZDs, sulfonylureas, repaglinide, liraglutide, extended-release exenatide, intermediate-acting insulins given at bedtime, and basal insulins control FPG more effectively. Exenatide, DPP-4 inhibitors, α-glucosidase inhibitors, nateglinide, and regular and rapid-acting insulin better control postprandial glucose excursions. We can also guide therapy based on the risk of hypoglycemia. Metformin, TZDs, liraglutide, exenatide, DPP-4 inhibitors, and α-glucosidase inhibitors have a low risk of hypoglycemia. Combining these agents will allow aggressive targeting of near-normal HbA1c levels while minimizing hypoglycemia and weight gain. Combining these agents early in the diagnosis of type 2 DM is logical to potentially realize the microvascular and macrovascular reduction seen in the long-term follow-up of UKPDS.


FIGURE 57-8 A1C goals. See for current algorithms. (Reprinted with permission from the Texas Diabetes Council.)

Preserving β-cell function, thus arresting the progressive nature of type 2 DM, could be paradigm changing, but to date medications have only shown to slow, not arrest, progression. In the UKPDS, insulin, metformin, or sulfonylureas did not halt β-cell failure. TZDs (out to 5 years with rosiglitazone) and GLP-1 agonists (open-label exenatide has shown durable HbA1c reduction to 3 years and liraglutide to 2 years) may potentially slow β-cell failure. Pathophysiologically treating type 2 DM for potential β-cell preservation is possible, but unproven. It appears unlikely any one drug class will arrest β-cell failure, necessitating combination therapy. TZD and GLP-1 agonist combination is logical as TZDs reduce apoptosis of β cells and GLP-1 agonists augment pancreatic function through insulin secretion in a glucose-dependent manner and reduction of inappropriate glucagon, but long-term data are lacking. β-Cell function is heavily damaged by the time type 2 DM is diagnosed, and it is possible that β-cell failure is inevitable by type 2 DM diagnosis. HbA1c reduction is dependent on baseline values, with higher reductions seen with higher values, but, again, therapy should be goal oriented. Triple therapy is often with metformin, a sulfonylurea, and a TZD or DPP-4 inhibitor, but a logical alternative is to use metformin, a TZD, and a GLP-1 agonist, which can lower glucose levels and increase satiety, reducing the weight gain potential of a TZD, and still has a low risk of hypoglycemia. A DPP-4 inhibitor may be an alternative, although without weight loss potential, if an injectable product is not preferred. If the HbA1c is >8.5% to 9% (>0.085 to 0.09; >69 to 75 mmol/mol Hb) on multiple therapies, insulin therapy should be considered. Sulfonylureas are often stopped when insulin is added and insulin sensitizers continued. This may be beneficial to decrease hypoglycemia, but continuing the sulfonylurea is permissible until multiple daily injections are started, at which time it should definitely be discontinued. Combination therapy with a TZD and insulin should be closely monitored for excessive weight gain and edema.

Virtually all patients with type 2 DM ultimately become relatively insulinopenic and will require insulin therapy. Insulin therapy for type 2 DM has changed dramatically in the last few years. Specifically, patients are often “transitioned” to insulin by using a bedtime injection of an intermediate- or long-acting insulin, and using oral agents primarily for control during the day. This strategy leads to less hyperinsulinemia during the day and is associated with less weight gain and has equal efficacy and a lower risk of hypoglycemia for up to 3 years when compared with starting prandial insulin or split-mix twice-daily insulin.55 Because most patients are insulin resistant, insulin sensitizers are commonly used with insulin therapy. Patients with type 2 DM are usually well buffered against hypoglycemia. Patients should be monitored for hypoglycemia by asking about nocturnal sweating, nightmares (both indicative of nocturnal hypoglycemia), palpitations, tremulousness, and neuroglycopenic symptoms, as well as SMBG. When bedtime insulin plus daytime oral medications fail to achieve glycemic goals, a conventional multiple daily dose insulin regimen while continuing the insulin sensitizers is often tried. This may mean adding an injection of bolus insulin with the largest meal of the day for a total of two injections. If this is unsuccessful, a bolus injection can be given with the second largest meal of the day, for a total of three injections. After this, the standard basal–bolus model is followed. Alternatively, patients may be switched to split-mix insulin such as 70/30 mix insulin, Humalog Mix 75/25, or Novolog Mix 70/30. These are often given twice daily before the first and third meals (see Type 1 Diabetes Mellitus under Therapeutics above for longer explanation), but if inadequate control is seen, a third dose of mix insulin may be given with the third meal of the day. This allows for better prandial coverage, but can also increase the risk of hypoglycemia.56 Use of GLP-1 agonists or pramlintide for prandial coverage can be considered. GLP-1 agonists, based on chosen drug, can be dosed weekly, daily, or twice daily, whereas pramlintide can be given before each meal. Concerns and problems with insulin administration as addressed in Type 1 Diabetes Mellitus under Therapeutics above generally relate to the therapy of type 2 DM. However, patients with type 2 DM rarely have hypoglycemia unawareness. Also, the variability of insulin resistance means that insulin doses may range from 0.7 to 2.5 units/kg or more. Figure 57-9 is an algorithm for insulin therapy options in type 2 diabetes developed by the Texas Diabetes Council. This algorithm gives most options for insulin therapy, but the choice of regimen should be individualized based on the discussion with the patient.



FIGURE 57-9 Insulin algorithm for type 2 DM in children and adults. See for current algorithms. (Reprinted with permission from the Texas Diabetes Council.)

The availability of short-acting insulin secretagogues, rapid-acting insulin analogs, exenatide, DPP-4 inhibitors, and α-glucosidase inhibitors, all of which target postprandial glycemia, has reminded practitioners that glycemic control is a function of fasting, preprandial, and postprandial glycemic excursions. Many clinicians and patients neglect monitoring postprandial glucose. However, postprandial glycemic excursions proportionally contribute more than the FPG to the HbA1c percentage when the HbA1c nears goals, and thus will need to be targeted for optimal glycemic control in many patients. It remains controversial whether targeting after-meal glucose excursions will have more of an effect on complications risk than more conventional strategies.

Special Populations

Children and Adolescents with Type 2 DM

Type 2 DM is increasing in adolescence.1,6 Obesity and physical inactivity seem to be particular culprits in the pathogenesis of this disease. Given the many years that the patient will have to live with diabetes, and recent evidence that the time frame after diagnosis for microvascular complications may mimic that of older adults, extraordinary efforts should be expended on lifestyle modification measures in an attempt to normalize glucose levels. Failing that strategy, the only labeled oral agent for use in children (10 to 16 years of age) is metformin, although sulfonylureas are also commonly used in therapy. TZDs and DPP-4 inhibitors have not been adequately studied in children, but studies to ascertain safety and efficacy are currently under way. GLP-1 agonist therapy, as it potentially helps the child to lose weight, is attractive, but the long-term effects of this therapeutic modality are unknown. Insulin therapy continues to be the standard therapy after metformin and a sulfonylurea. In adolescent females, the possibility of future pregnancy should be considered in the prescription of any drug regimen. Screening and recommendations for treatment of hypertension, dyslipidemia, nephropathy, retinopathy, hypothyroidism, and celiac disease are available.6

Elderly Patients with DM

Elderly patients with newly diagnosed DM (almost always type 2 DM) present a different therapeutic challenge. Consideration of the risks of hypoglycemia, the extent of comorbidities including severe microvascular disease, CVD, dexterity, self-care and social situations, falls risk, mental status, and the probable life span should help determine glycemic goals. If extensive comorbidities, hypoglycemic unawareness, unstable CVD, dementia, high falls risk, or similar diagnosis is made, the clinician may adjust the glycemic goal. Avoidance of hypoglycemia, especially severe hypoglycemia, as well as elevated glucose levels that may exacerbate the comorbidities is necessary (Fig. 57-8). It should also be remembered that elderly may have an altered presentation of hypoglycemia, as they lose adrenergic symptoms due to loss of autonomic nerve function as they age. This may raise the rise of neuroglycopenic symptoms shortly after identification of hypoglycemia. If the patient is newly diagnosed and does not have the above problems, a goal HbA1c <7% (<0.07; <53 mmol/mol Hb) is justified. If the person has significant comorbidities as mentioned above, then a goal <8% (<0.08; <64 mmol/mol Hb) may be reasonable, and if the person has “end-stage” illness, glycemic control should limit symptomatic (polyuria/polydipsia) or mental status issues. If oral agents will work, DPP-4 inhibitors, shorter-acting insulin secretagogues, low-dose sulfonylureas (preferably not long-acting ones), or α-glucosidase inhibitors may be used. The risk for lactic acidosis, which increases with older age and the age-related decline in renal function, makes metformin therapy difficult, but lower doses may be used if not contraindicated. In a patient in whom weight gain or loss may not be unwelcome, TZDs or GLP-1 agonists, respectively, may be considered, but falls risk and fracture risk must be considered with TZDs. DPP-4 inhibitors or α-glucosidase inhibitors are oral medications that may be advantageous due to a low risk of hypoglycemia. Simple insulin regimens such as an injection of basal insulin daily may be appropriate for glycemic control in elderly patients, especially if tight glycemic control is not the goal. The Texas Diabetes Council publishes an algorithm; see

Gestational DM and Pregnancy with Preexisting Diabetes

GDM is diagnosed as previously described. The adverse outcomes associated with GDM include birth defects, increased rates of miscarriage, necessity of cesarean section delivery, neonatal hypoglycemia, preeclampsia/eclampsia, preterm delivery, shoulder dystocia/birth injury, and hyperbilirubinemia. Dietary therapy to minimize wide fluctuations in BG is of paramount importance.5,8 Intensive educational efforts are usually necessary. Pregnant women without DM maintain plasma glucose concentrations between 50 and 130 mg/dL (2.8 and 7.2 mmol/L). Normoglycemia is the goal, and failure to maintain this despite dietary interventions often will necessitate medication use. Goals during therapy are minimally a preprandial goal of ≤90 mg/dL (≤5 mmol/L), and either 1-hour postprandial plasma glucose levels ≤120 mg/dL (≤6.7 mmol/L) or 2-hour postprandial plasma glucose levels ≤110 mg/dL (≤6.1 mmol/L), and avoidance of ketones as much as possible. In patients who have preexisting type 1 or type 2 DM and become pregnant, premeal, bedtime, and overnight glucose should be 60 to 90 mg/dL (3.3 to 5 mmol/L), with a peak postprandial of 100 to 120 mg/dL (5.6 to 6.7 mmol/L). HbA1c during pregnancy should be less than 6% (<0.06; <42 mmol/mol Hb), but frequent SMBG is the method of choice for monitoring glycemic control. Titration of insulin and switching to more complicated regimens is guided by SMBG results. Use of basal insulins other than NPH is still debated, but with the ease of use of detemir or glargine insulin, their use in GDM is increasing. In addition, pump therapy for the duration of the pregnancy is often instituted, as it can obtain excellent glycemic control and is quickly adjustable. Both metformin and glyburide have been studied as alternatives to insulin therapy. Glyburide was not detected in the cord serum of any infant in one study, whereas metformin crosses the placenta. Further study is needed prior to routinely recommending them in GDM. Patients with GDM should be evaluated 6 weeks after delivery to ensure that normal glucose tolerance has returned. Because these patients’ lifetime risk for the development of type 2 DM is >50%, periodic assessment after that is warranted.

Clinical Controversy…

Treatment of Type 2 Dm in Older Adults

The U.S. population continues to age. The ACCORD,48 ADVANCE,49 and VADT50 had older individuals who were in their 60s at enrollment. As stated earlier in the chapter, all improved glycemic control but did not reduce the risk of CVD over 3 to 5 years, although more people died in the ACCORD, resulting in termination of the study. ADVANCE reported improvement in nephropathy outcomes, and this did not differ by age, but otherwise these neutral studies did not report improved microvascular outcomes. In addition, one Japanese study had subjects with a mean age of 72 years, and a 6-year follow-up, but changes in glycemia were minimal, thus showing no benefit.57 This is unfortunate, as up to one in three people in this age category may have type 2 DM. Diabetes in older adults is complicated by clinical and functional heterogeneity. Patients may be relatively healthy, free-living adults all the way to the other end of the spectrum with assistive living, multiple comorbidities, and cognitive issues. Based on this, what is the optimal medication therapy for this group of individuals?

Critical evaluation of most medications in populations over 65 years of age is severely lacking. Many clinicians have decided that insulin, especially basal insulin, is a reasonable choice in this age group, and that minimal orals (maybe metformin if not contraindicated) are reasonable. Yet, in the new ADA guidelines and clinical practice recommendations, a patient-centered approach is stated. It is unlikely that most patients would choose basal insulin as their initial intervention if asked. Also, the cost for basal insulin is not minimal, and one must ask if it is truly more cost-effective than many other interventions. Severe hypoglycemia must be avoided in this population, as it has been associated with a higher risk of death for more than 1 year after the incident. Any of the agents can avoid severe hypoglycemia if used properly, but the risk factors for hypoglycemia are: use of insulin or insulin secretagogues, duration of diabetes, antecedent hypoglycemia, erratic meals, exercise, and renal insufficiency. In addition, self-care, visual acuity, and dexterity issues may be of concern in patients.

Medications that do not cause hypoglycemia may be advantageous. Metformin, if not contraindicated, continues to be an excellent first choice. As metformin may be used in Stage III CKD, with a reduced dose, this may be a reasonable choice. Second-choice agents such as DPP-4 inhibitors, if close to the chosen HbA1c goal, or a GLP-1 receptor agonist, if farther from the chosen goal, may be warranted. Each has its own issues, as both may be cost prohibitive and GLP-1 receptor agonists may be inappropriate for patients with GI issues or gastroparesis, or normal-weight to underweight patients. α-Glucosidase inhibitors, if close to goal and constipation is an issue, may be helped by these agents, although GI tolerability is problematic.

As the care for people with diabetes improves, it is imperative that issues concerning older adults continue to be addressed. It is important for our society that optimal therapy in older adults be properly addressed, as this population will continue to grow, and currently there is no consensus. Several organizations have recommendations in regards to older adults,58,59 but recommendations on optimal pharmacotherapy are not included.

Clinical Controversy…

Oral Agents in Pregnancy

The use of oral antidiabetic agents for the management of gestational diabetes or type 2 DM during pregnancy is controversial. For those patients who fail to maintain optimal glycemic control during pregnancy with diet and lifestyle modification, traditionally the next step has been to proceed with insulin therapy. More recently, however, some clinicians have begun using oral agents including sulfonylureas and/or metformin in patients with GDM or type 2 DM during pregnancy.

A randomized, open-label, controlled trial evaluated the efficacy of glyburide compared with insulin initiated after 11 weeks’ gestation.60 The control of BG compared with insulin therapy was similar, with less hypoglycemia in the glyburide group. There was not any evidence of significant difference in complications, including cord serum insulin concentrations, incidence of macrosomia (birth weight ≥4,000 g), cesarean delivery, or neonatal hypoglycemia between regimens. Glyburide was not detected in the cord serum of any infant. However, this study limited enrollment to beyond 11 weeks’ gestation; therefore, no conclusions can be made regarding teratogenicity from using glyburide in the first trimester.

A more recent retrospective cohort study of 10,682 women with GDM who required medical therapy, however, revealed that babies born to women with GDM who were managed on glyburide were more likely to be macrosomic and to be admitted to the intensive care unit compared with those treated with insulin therapy.61

Metformin has also been used in the management of GDM and type 2 DM in pregnancy, and also in polycystic ovarian syndrome to prevent miscarriage. Early studies dating back to the 1980s did not show any differences in perinatal mortality, maternal hypoglycemia, lactic acidosis, or congenital anomalies.62,63

However, a more recent cohort study investigating the effects of metformin, sulfonylureas, and insulin in pregnant women with diabetes found a significantly higher rate of preeclampsia in women treated with metformin compared with women treated with sulfonylurea or insulin (32% metformin vs. 7% sulfonylureas vs. 10% insulin). The perinatal mortality was also significantly increased in women treated with metformin in the third trimester compared with women not treated with metformin (11.6% vs. 1.3%).64

In contrast, another study of 751 women with GDM randomly assigned subjects at 20 to 33 weeks of gestation to open treatment with metformin (with supplemental insulin if required) or insulin. This study did not find any increased rate of preeclampsia or other perinatal complications compared with insulin.65

Subsequently there have also been meta-analyses that also revealed no differences in maternal or neonatal outcomes with the use of glyburide or metformin compared with the use of insulin in women with GDM.66,67

The current guidelines of the ADA continue to suggest insulin therapy as the preferred treatment for managing women with gestational diabetes or type 2 DM in pregnancy who fail to achieve optimal control with diet/lifestyle modification alone.68 Moreover, neither metformin nor glyburide has formal FDA approval for the management of diabetes in pregnancy.

The use of oral antidiabetic agents in pregnancy is becoming more common, but nevertheless remains controversial. Clinicians who prescribe oral agents to manage their patients with diabetes during pregnancy must consider the potential benefits (avoidance of injections, decreased cost, and patient preference) against the potential risk (unknown safety, inconsistent effect on the pregnancy outcomes, and potential liability due to using non–FDA-approved therapy).

Special Situations

Sick Days

Acute self-limited illness rarely presents a major problem for patients with type 2 DM, but can be a significant challenge for insulinopenic type 1 DM patients.69 While caloric intake generally declines, insulin sensitivity also decreases, meaning that it may take greater amounts of insulin to control BG concentrations. Patients need to be adept at frequent SMBG, checking urine ketones, use of short-acting insulin, and understanding that sugar intake in this situation is not detrimental but may be necessary to balance the glucose levels when extra insulin is needed during illness. Plan to maintain a meal plan containing 120 to 150 g of carbohydrates per day. We encourage patients to continue their usual insulin regimen and to use supplemental rapid-acting insulin based on SMBG results, with additional insulin given if ketonuria develops. Ketone testing should be in type 1 DM patients prone to ketonuria, if two consecutive plasma glucose readings are above 250 mg/dL (13.9 mmol/L), or if vomiting occurs, as it is a possible sign of ketosis. Sugar and electrolyte solutions, can be used to maintain hydration, to provide needed electrolytes if there are significant GI or urinary losses, and to provide sugar to keep the patient from developing hypoglycemia because of the extra insulin that is usually needed. If patients with type 1 DM are consistently hyperglycemic, we suggest they abstain from sugary drinks and increase intake of sugar-free liquids. In contrast, patients with type 2 DM may need to switch to sugar-free drinks if BG levels are continually elevated. Most patients can be taught how to sufficiently manage sick days and avoid hospitalization.

Diabetic Ketoacidosis and Hyperosmolar Hyperglycemic State

DKA and HHS are true diabetic emergencies.70 A comprehensive discussion of their treatment is beyond the scope of this chapter. In patients with known diabetes, DKA is usually precipitated by insulin omission in type 1 DM, and intercurrent illness, particularly infection, in both type 1 and type 2 DM. However, patients with type 1 or type 2 DM (the latter being usually nonwhites or Hispanics) may present with DKA at initial presentation. It is possible that some of the patients deemed to have type 2 DM actually have type 1 idiopathic DM. Patients with DKA may be alert, stuporous, or comatose at presentation. The hallmark diagnostic laboratory values for DKA include hyperglycemia, anion gap acidosis, and large ketonemia or ketonuria. Diagnostic criteria for HHS are similar with the exception of significantly higher plasma glucose, elevated effective serum osmolality, and little to no ketonuria or ketonemia when compared with DKA. HHS typically evolves over several days to weeks, whereas DKA evolves much quickly. Afflicted patients will have either fluid deficits of several liters or sodium and potassium deficits of several hundred milliequivalents. Restoration of intravascular volume acutely with normal saline, followed by hypotonic saline to replace free water, potassium supplements, and constant infusion of insulin restore the patient’s metabolic status relatively quickly. A flow sheet is often helpful in tracking the fluid and insulin therapies and laboratory parameters in these patients. Bicarbonate administration is generally not needed and may be harmful, especially in children. Treatment of the inciting medical condition is also vital. Hourly bedside monitoring of glucose and frequent monitoring (every 2 to 4 hours) of potassium is essential. Metabolic improvement is manifested by an increase in the serum bicarbonate or pH. Serum phosphorus usually starts high and plummets to lower-than-normal levels, although replacing phosphorus, while not unreasonable, is of questionable benefit in most patients. Fluid administration alone will reduce the glucose concentration, so a decrement in glucose values does not necessarily mean that the patient’s metabolic status is improving. Rare patients will require larger amounts of insulin than those usually given (5 to 10 units/h). We double the patient’s insulin dose if the serum bicarbonate has not improved after the first 4 hours of insulin therapy. Constant infusion of a fixed dose of insulin and the administration of IV glucose when the BG level decreases to <250 mg/dL (<13.9 mmol/L) are preferable to titration of the insulin infusion based on the glucose level. The latter strategy may delay clearance of the ketosis and prolong treatment. The insulin infusion should be continued until the urine ketones clear and the anion gap closes. Long-acting insulin should be given 1 to 3 hours prior to discontinuing the insulin infusion. Intramuscular regular insulin or subcutaneous insulin lispro or aspart given every 1 to 2 hours can be utilized rather than an insulin infusion in patients without hypoperfusion. Patients may develop hyperchloremic metabolic acidosis with treatment if they have been given large volumes of normal saline in the course of their treatment. Such a situation does not require any specific treatment.

HHS usually occurs in older patients with type 2 DM, at times undiagnosed, or in younger patients with prolonged hyperglycemia and dehydration or significant renal insufficiency. Large ketonemia is usually not seen, as residual insulin secretion suppresses lipolysis. Infection or another medical illness is the usual precipitant. Fluid deficits are usually greater and BG concentrations higher (at times >1,000 mg/dL [>55.5 mmol/L]) in these patients than in patients with DKA. BG levels should be lowered very gradually with hypotonic fluids and low-dose insulin infusions (1 to 2 units/h). Rapid correction of the glucose levels, a drop greater than 75 to 100 mg/dL/h (4.2 to 5.6 mmol/L/h), is not recommended, as it can result in cerebral edema. This is especially true for children with DKA. Mortality is high with the HHS.

Hospitalization for Intercurrent Medical Illness

Patients on oral agents may need transient therapy with insulin to achieve adequate glycemic control. In patients requiring insulin, patients should receive scheduled doses of insulin with additional short-acting insulin. “Sliding-scale” insulin is to be discouraged, as it is notorious for not controlling glucose and for sometimes resulting in therapeutic misadventures, with wide amplitudes of glycemic excursions.71 In-hospital mortality is increased in many hyperglycemic conditions. At least one study documented a reduction in mortality in type 2 diabetes patients with acute MIs72 who receive constant IV insulin during the acute phase of the event to maintain near-normal glucose concentrations. Similar mortality results have been documented in some intensive care unit settings using IV insulin and tight glucose control.72,73 The ADA and American Association of Clinical Endocrinologists released a joint consensus statement on inpatient glycemic control stating that glucose control measures should be implemented if the BG is ≥180 mg/dL (≥10 mmol/L), and maintained between 140 and 180 mg/dL (7.8 and 10 mmol/L).74 The Normoglycemia in Intensive Care Evaluation—Survival Using Glucose Algorithm Regulation trial, and several other negative trials, has resulted in this loosening of inpatient glycemic goals. Critically ill patients had higher 90-day mortality when a goal of 81 to 108 mg/dL (4.5 to 6 mmol/L) was targeted than when BG of ≤180 mg/dL (≤10 mmol/L) (achieved 144 mg/dL [8 mmol/L]) was targeted.75 For non–critically ill patients there are no established BG goals. Reasonable BG goals for these patients are <140 mg/dL (<7.8 mmol/L) premeal and <180 mg/dL (<10 mmol/L) random.5 Many protocols for IV insulin infusion are currently available, and implementation for an inpatient setting should use a well-established protocol. It is prudent to stop metformin in all patients who arrive in acute care settings until full elucidation of the reason for presentation can be ascertained, as contraindications to metformin are prevalent in hospitalized patients. Discharge planning is also important, as approximately one third of patients will have newly diagnosed diabetes and another one third will likely have prediabetes, as determined by obtaining an HbA1c on admission (best) or prior to discharge.

Perioperative Management

Surgical patients may experience worsening of glycemia for reasons similar to those listed above for intercurrent medical illness. Therapy should be individualized based on the type of DM, nature of the surgical procedure, previous therapy, and metabolic control prior to the procedure. Patients on oral agents may need transient therapy with insulin to control BG. In patients requiring insulin, scheduled doses of insulin or continuous insulin infusions are preferred. For patients who can eat soon after surgery, the time-honored approach of giving one half of the usual morning NPH insulin dose with dextrose 5% in water IV is acceptable, with resumption of scheduled insulin, perhaps at reduced doses, within the first day. Patients receiving basal/bolus insulin therapy can continue receiving their usual dose of long-acting insulin while holding the premeal bolus doses until the patient can tolerate meals. For patients requiring more prolonged periods without oral nutrition and for major surgery, such as coronary artery bypass grafting and major abdominal surgery, constant infusion of IV insulin is preferred. Use of IV insulin infusion has been shown to reduce deep sternal wound infections in patients after coronary artery bypass grafting, although there is no need to start the infusion during or before the procedure. Metformin should be discontinued temporarily after any major surgery until it is clear that the patient is hemodynamically stable and normal renal function is documented.

Reproductive-Age Women and Preconception Care for Women

An increasing prevalence of DM has been noted in reproductive-age women.5,76,77 Prepregnancy planning is absolutely mandatory, as organogenesis is largely completed within 8 weeks, so good glycemic control should be obtained prior to conception. Unfortunately, major congenital malformations due to poor glucose control remain the leading cause of mortality and serious morbidity in infants of mothers with type 1 or type 2 DM. For women with DM controlled by lifestyle measures alone, conversion to insulin as soon as the pregnancy is confirmed is appropriate. Patients previously treated with insulin may need intensification of their regimen to achieve therapeutic goals. Normal pregnancy is associated with a decrease in the BG concentration as it is diverted to the fetus.

Human Immunodeficiency Virus (HIV) Patients and Diabetes

Patients living with HIV are at higher risk for the development of type 2 DM.78 This risk may be related to HIV infection, concomitant infections such as hepatitis C, and concomitant medications often used to treat HIV or comorbidities. Pentamidine, used for P. carinii pneumonia, is directly β-cell toxic in some patients; hypoglycemia may be followed by hyperglycemia. Megestrol, used as an appetite stimulant, can have glucocorticoid-like effects in some patients. In addition, protease inhibitors, used to manage HIV, have been shown to potentially worsen insulin sensitivity, decrease the ability of the β cell to secrete insulin, and/or worsen lipotoxicity. Long-term stavudine may also increase the risk of diabetes. Redistribution of fat from medication or HIV infection, with resultant increases in visceral fat and decreases in subcutaneous fat, is not uncommon. Metformin continues to be the first-line therapy choice for HIV patients, as weight gain can be minimized, but additional cautions must be noted. Stavudine, zidovudine, and didanosine may cause lactemia, especially on long-term use, whereas abacavir, lamivudine, and tenofovir have less incidence of lactemia. It may be advisable to check lactate levels in appropriate subjects prior to metformin use. If lactate levels are greater than two times normal, alternative therapy should be considered. If excess visceral adiposity is noted, a TZD, which redistributes fat back to subcutaneous adipose tissue and causes visceral fat apoptosis, may be considered. Significant drug–drug interactions may also be present (refer to specific diabetes drugs in Chap. 103).

Special Topics

Prevention of Diabetes Mellitus

Images Efforts to prevent type 1 DM with niacinamide, injected insulin, or oral insulin therapy have been unsuccessful. Anti-CD3 and anti-CD20 monoclonal antibodies and a GAD vaccine have shown to delay, but not stop, β-cell destruction in type 1 DM. In addition, a 24–amino acid sequence derived from human heat shock protein 60 called DiaPep277® may slow loss of C-peptide secretion in type 1 DM. Future directions include potential combination therapy trials. The Diabetes Prevention Program79 confirmed that modest weight loss in association with exercise can have a dramatic impact on insulin sensitivity and the conversion from IGT to type 2 diabetes. In this study approximately 2,000 individuals with IGT were randomized to lifestyle changes (diet, exercise, and weight loss) versus usual care. The study, which was originally planned to be ongoing for 5 years, was stopped early after 2.8 years. The usual care group developed diabetes at the rate of 11% each year. The lifestyle arm developed diabetes at a rate of 5% per year, a 58% reduction in the risk of developing diabetes.79 Surprisingly, a modest amount of diet and exercise yielded impressive results. The exercise program in the lifestyle group was walking 30 minutes 5 days each week. The mean weight loss over the 2.8-year study period was only 8 lb (3.6 kg). In the Diabetes Prevention Program79 discussed above, approximately 1,000 of the study patients were randomized to metformin therapy. Metformin therapy reduced the risk of developing type 2 DM by 31% compared with usual care and resulted in a 4-lb (1.8-kg) weight loss. Interestingly, young and overweight individuals on metformin had a greater reduction in the risk of developing diabetes than normal-weight and older study patients.79

All diabetes medications studied for the prevention of diabetes, when discontinued, do not appear to have residual positive effects on β-cell function. Thus, patients must continue the medication for continued “prevention,” although the question arises if this is merely early treatment. Troglitazone, a TZD removed from the market, was able to prevent the development of diabetes in women with a history of gestational diabetes. Total preservation of β-cell function was demonstrated over a 5-year period in women who had near-normal β-cell function at baseline and who initially responded to the drug.80 The preservation of β-cell function was observed for at least 8 months after the drug had been discontinued. The DREAM trial evaluated rosiglitazone and/or ramipril treatment for the delay or prevention of type 2 DM in impaired glucose-tolerant subjects.81,82 Rosiglitazone 8 mg daily, over approximately 3 years, reduced the incidence of type 2 diabetes by 60%. In addition, a 37% nonsignificant increase in cardiovascular events was reported. Ramipril 15 mg daily did not significantly prevent the conversion to diabetes. It is possible that longer exposure could have made a difference, but the study was stopped prematurely. In contrast, valsartan, an angiotensin receptor blocker (ARB), administered for 5 years was recently reported to reduce the risk of progression from IGT to type 2 DM by 14%. The ACT Now trial used pioglitazone 45 mg daily in an IGT population and found a 72% reduction in the risk of development of diabetes over 2.4 years.83 It should be noted that no pharmacologic agents are currently FDA approved or recommended for prevention of type 2 diabetes, although the ADA recommends metformin in conjunction with lifestyle changes if the patient is younger, obese, has a family history of diabetes, dyslipidemia, hypertension, or a HbA1c >6% (>0.06; >42 mmol/mol Hb).5 The next step is discussions with the FDA to decide how and if medications to prevent diabetes can be approved for this indication.

Patient Education

Images It is not satisfactory to give patients with DM brief instructions with a few pamphlets and expect them to manage their disease adequately.84 Diabetes education is a lifetime exercise. Successful treatment of DM involves lifestyle changes for the patient (e.g., medical nutrition therapy, physical activity, SMBG and possibly of urine for ketones, recognition of hyperglycemia and hypoglycemia, and taking prescribed medications). The American Association of Diabetes Educators (AADE) has developed the AADE7 self-care behaviors of healthy eating, being active, monitoring, taking medication, problem solving, reducing risk, and healthy coping, which is a good starting framework for patient discussions. The patient must be involved in the decision-making process and must learn as much about the disease and associated complications as possible. Emphasis should be placed on the evidence that indicates that complications can be prevented or minimized with glycemic control and management of risk factors for CVD. Recognition of the need for proper patient education to empower them into self-care has generated programs for certification in diabetes education for pharmacists. Certified diabetes educators (CDEs) must document their patient education hours and sit for a certification examination that assesses the knowledge, tasks, and skills of an educator in order to become certified. An increasing number of nurses, pharmacists, dietitians, and physicians are becoming CDEs to document to the public that they meet a minimum standard for diabetes education, and to fulfill quality initiatives in meeting guidelines for education recognition. Being a CDE does not guarantee reimbursement of services, and CDEs who are not dietitians will often need to become part of a recognized program to obtain reimbursement. Currently the AADE and ADA have accreditation programs.

Treatment of Concomitant Conditions and Complications


Patients with established retinopathy should see an ophthalmologist or optometrist trained in diabetic eye disease. A dilated eye examination is required to fully evaluate diabetic eye disease. Early background retinopathy may reverse with improved glycemic control and optimized blood pressure control. More advanced retinopathy will not fully regress with improved glycemia, but caution should be taken on the expediency of glycemia lowering, as aggressive reductions in glycemia may acutely worsen retinopathy. The pathophysiology of retinopathy is better understood to involve inappropriate growth factor increase and microcirculation ischemia. Bevacizumab, used off-label, and ranibizumab, recently FDA approved, are antivascular endothelial growth factor monoclonal antibodies given by intravitreal injection. Although approved for macular edema, use may also apply to other neovascular ocular conditions. A protein kinase C inhibitor has been studied, but results have been modest. Laser photocoagulation has markedly improved sight preservation in diabetic patients. People with diabetes also have a higher rate of cataracts and possibly open-angle glaucoma.


Neuropathy in diabetes can generally be placed into three categories: peripheral neuropathy, autonomic neuropathy, and focal neuropathy. Distal symmetrical peripheral neuropathy is the most common complication seen in type 2 DM patients in outpatient clinics.85 Paresthesias, perceived hot or cold, numbness, or pain may be the predominant symptom. The feet are involved far more often than the hands as it affects longer nerves first and progresses proximally. Improved glycemic control is the primary treatment and may alleviate some of the symptoms. If neuropathy is painful, symptomatic pain treatment is indicated, although it will not change the course of the neuropathy nor has one medication been shown to be superior to another. Treatment may be with low-dose tricyclic antidepressants, anticonvulsants (gabapentin, pregabalin, rarely carbamazepine), duloxetine, venlafaxine, topical capsaicin, and various pain medications, including tramadol and nonsteroidal antiinflammatory drugs. Duloxetine and pregabalin have FDA approval for this indication. The numb variant of peripheral neuropathy is not treated with medication, but may lead to pressure areas on the foot and subsequent ulcer in a subset of patients. Clinical manifestations of diabetic autonomic neuropathy include resting tachycardia, exercise intolerance, orthostatic hypotension, constipation, gastroparesis, erectile dysfunction, sudomotor dysfunction (anhidrosis, heat intolerance, gustatory sweating, and/or dry skin), impaired neurovascular function, and hypoglycemic autonomic failure. Gastroparesis can be a severe and debilitating complication of DM. Improved glycemic control, discontinuation of medications that slow gastric motility, and the use of metoclopramide (preferably for only a few weeks at a time) or low-dose erythromycin may be helpful. Gastric pacemakers as therapeutic hardware are rarely used, though available. Cisapride, removed from the market several years ago, is still available for compassionate use and domperidone, available outside of the United States, may be useful. Orthostatic hypotension, after stopping antihypertensives and liberalizing dietary sodium intake, may require pharmacologic management with mineralocorticoids or adrenergic agonist agents. In severe cases, supine hypertension is extreme, mandating that the patient sleep in a sitting or semirecumbent position. Patients with cardiac autonomic neuropathy are at a higher risk for silent MI and sudden cardiac death. The hallmark of diabetic diarrhea is its nocturnal occurrence. Diabetic diarrhea frequently responds to a 10- to 14-day course of an antibiotic such as doxycycline or metronidazole. In more unresponsive cases, octreotide may be useful. Erectile dysfunction is common in diabetes, and initial treatment should include a trial of one of the phosphodiesterase type 5 inhibitors prior to referral. People with diabetes often require the highest doses of these medications to have an adequate response. Sudomotor dysfunction, as earlier defined, results in loss of sweating and resultant dry, cracked skin. Use of hydrating creams and ointments is needed. Hypoglycemic unawareness requires the patient to avoid hypoglycemia, as the body will slowly increase the glycemic level at which it will signal the autonomic signals, although damage may severely lessen the response. Focal neuropathies are uncommon, but occur more often in older, poorly controlled diabetes patients. Diabetic amyotrophy, which is characterized by a proximal thigh muscle pain and weakness, is one of the most debilitating. In addition, cranial nerve III, IV, and VI neuropathies, as well as Bell’s palsy, may occur. The presentation can be quite dramatic, but the course is usually self-limited, and partial or full recovery happens in a few weeks to months. Carpal tunnel syndrome, caused by radial nerve entrapment, is also more common in people with diabetes,

Microalbuminuria and Nephropathy

DM, particularly type 2 DM, is the biggest contributor statistically to the development of end-stage renal disease in the United States.1,5 The ADA recommends a screening urinary analysis for albumin at diagnosis in persons with type 2 DM. Precise onset of type 2 DM can rarely be ascertained, and patients will often present at diagnosis with microvascular complications. In type 1 DM, microalbuminuria rarely occurs with short duration of disease or before puberty. Screening individuals with type 1 DM should begin with puberty and after 5 years’ disease duration. There are three methods for assessing microalbuminuria: (a) measurement of the urine albumin-to-creatinine ratio in a random spot collection (preferably the first morning void); (b) 24-hour timed collection; and (c) timed (e.g., 4- or 10-hour overnight) collection. Microalbuminuria on a spot urine specimen is defined as an albumin-to-creatinine ratio of 30 to 300 mg/g (3.4 to 34 mg/mmol creatinine). On timed collections, microalbuminuria is defined as 30 to 300 mg/24 hours or an albumin excretion rate of 20 to 200 mcg/min. Because of day-to-day variability, microalbuminuria should be confirmed on at least two of three samples over 3 to 6 months unless unequivocal. Additionally, when assessing urine protein or albumin, conditions that may cause transient elevations in urinary albumin excretion should be excluded. These conditions include intense exercise, recent urinary tract infections, hypertension, short-term hyperglycemia, heart failure, and acute febrile illness.5

In type 2 DM, the presence of microalbuminuria is a strong risk factor for macrovascular disease and is frequently present at the time of diagnosis. Microalbuminuria is a weaker predictor for future end-stage kidney disease in type 2 versus type 1 DM. Glucose and blood pressure control are most important for the prevention of nephropathy, and blood pressure control is the most important for retarding the progression of established nephropathy. ACE inhibitors and ARBs, considered first-line recommended treatment modalities, have shown efficacy in preventing the clinical progression of renal disease in patients with diabetes. Combined renin–angiotensin–aldosterone system blockage (with ACE inhibitors, ARBs, aldosterone receptor blockers, and/or direct renin inhibitors) cannot currently be recommended for routine practice in nephropathy. Diuretics frequently are necessary due to the volume-expanded state of the patient and are recommended second-line therapy. The ADA and the National Kidney Foundation blood pressure goal of <130/80 mm Hg can be difficult to achieve. Three or more antihypertensives are often needed to treat to goal blood pressures (see also Chap. 29).

Peripheral Vascular Disease and Foot Ulcers

Claudication and nonhealing foot ulcers are common in type 2 DM patients.86 Smoking cessation, correction of lipid abnormalities, and antiplatelet therapy are important strategies in treating claudicants. Cilostazol may be useful for reducing intermittent claudication symptoms in select patients. Revascularization is successful in selected patients, although small vessel disease that cannot be bypassed is common in diabetes. Local débridement and appropriate footwear and foot care are vitally important in the early treatment of foot lesions. In more advanced lesions multiple treatments including grafts, topical wound healing, and even hyperbaric treatments may be necessary. Diabetic foot care is an excellent example of the adage, “an ounce of prevention is worth a pound of cure.” Thus, a foot examination at each visit is recommended. A yearly Semmes-Weinstein 5.07/10-g force monofilament test for sensation can be used to identify high-risk patients who need further podiatric evaluation.

Coronary Heart Disease

Images The risk for coronary heart disease (CHD) is two to four times greater in diabetic patients than in nondiabetic individuals. CHD is the major source of mortality in patients with DM. Multiple risk factor intervention (lipids, hypertension, smoking cessation, and antiplatelet therapy)5 will reduce the burden of excess macrovascular events. The ADA recommends aspirin therapy in all secondary prevention situations, and if allergic to aspirin, consider clopidogrel. Recent evidence in primary prevention studies of antiplatelet therapy in type 2 DM has not shown benefit. The ADA currently recommends that if the 10-year risk of CVD is at least 10%, or the patient is at your judgment high risk, or in women at least 60 years old or men at least 50 years old, primary prevention antiplatelet therapy can be considered. Epidemiologic data suggest that CHD prevention guidelines for type 2 DM apply equally to patients with type 1 DM.5 β-Blocker therapy supplies an even greater protection from recurrent CHD events in diabetic patients than in nondiabetic subjects. Masking of hypoglycemic symptoms is a greater problem in type 1 DM patients than in patients with type 2 DM, although this risk can be adequately managed with proper glycemic control interventions (see also Chap. 6).

Lipids The Collaborative Atorvastatin Diabetes Study (CARDS) randomized diabetes subjects with no documented CVD to atorvastatin 10 mg daily (n = 1,428) or placebo (n = 1,410). The trial was stopped 2 years early (mean duration of follow-up was 3.9 years) after meeting the primary efficacy end point of major cardiovascular events, which were reduced by 37% (P = 0.001). All-cause death was reduced 27% (P = 0.059), and potentially could have had its significance influenced by the early stoppage of the trial.87 The Heart Protection Study randomized 5,963 patients aged >40 years with diabetes and total cholesterol >135 mg/dL (>3.49 mmol/L). A significant 22% reduction (95% CI, 13 to 30) in the event rate for major cardiovascular events was seen with simvastatin 40 mg/day. This was evident even at lower LDL levels (<116 mg/dL [<3 mmol/L]), and suggests that ~30% to 40% reduction in LDL levels regardless of starting LDL levels may be appropriate.88 The ADA recommends statin therapy, regardless of baseline lipid or LDL-C levels in patients with overt CVD or without documented CVD who are over the age of 40 and have CVD risk factors besides diabetes.5

The proper use of fibrates in diabetes continues to be controversial. The Fenofibrate Intervention and Event Lowering in Diabetes (FIELD) was conducted in 9,795 subjects (22% with previously documented CVD) with type 2 DM given fenofibrate 200 mg daily or placebo. A relative reduction of 11% (P = 0.16) was seen in any coronary event in conjunction with a slight increase in the risk of all-cause mortality (0.7%; P = 0.18). Reasons for this have been speculated on, including the increased use of statins in the placebo group, but continue to be controversial.89 On subset analysis, only subjects without CVD had a significant reduction in CVD events. The lipid arm of the ACCORD90 reported on 5,518 subjects randomized to fenofibrate or placebo given with low-dose simvastatin (~20 mg/day). Fenofibrate addition did not significantly lower cardiovascular events (0.92; 95% CI, 0.79 to 1.08). Niacin in combination with a statin recently failed to improved CVD outcomes as well.

The NCEP-ATP III15 guidelines classify the presence of DM as a CHD risk equivalent, and therefore recommend that LDL-C be lowered to <100 mg/dL (<2.59 mmol/L). An optional LDL goal in high-risk DM patients, such as those who already have CHD, has been updated to <70 mg/dL (<1.81 mmol/L)91 (Table 57-12). The primary goal of treatment is to obtain the LDL-C goal. After the LDL-C goal is reached (usually with a statin), via NCEP-ATP, triglycerides are possibly considered for pharmacologic management, assuming unresponsiveness to glycemic control efforts, weight management, and exercise. In such situations, a non–HDL-C goal is established (a surrogate for all apolipoprotein B–containing particles). The non–HDL-C goal for patients with DM is <130 mg/dL (<3.36 mmol/L). Niacin or a fibrate can be added to reach that goal if triglycerides are 201 to 499 mg/dL (2.27 to 5.64 mmol/L), although there is little evidence this will lower CVD. Patients with marked hypertriglyceridemia (≥500 mg/dL [≥5.65 mmol/L]) are at risk for pancreatitis. Efforts to reduce triglycerides with glycemic control, elimination of other secondary causes (including medications), and drug therapy (fibrates, statins, and potentially niacin) are effective treatment strategies. Readers are also referred to the Chapters 3 and 11 for further information.

TABLE 57-12 Classification of Lipid and Lipoprotein Levels in Adults 5,91,92



The role of hypertension in increasing microvascular and macrovascular risk in patients with DM has been confirmed in the UKPDS.44 The ADA has loosened its goals for blood pressure (<140/80 mm Hg) in patients with DM.5 The ACCORD blood pressure arm studied type 2 DM patients, with a goal of achieving a systolic blood pressure of either <120 mm Hg (achieved 119 mm Hg) or <140 mm Hg (133 mm Hg achieved).92 The lower pressure group did not have lower CVD or renal outcomes, but did have a lower risk of stroke. A goal of <130 mm Hg can still be considered in patients at high risk of a stroke or if renal disease is present. ACE inhibitors and ARBs are generally recommended for initial therapy, as they have shown to be cardioprotective, and likely have special renal protection. Many patients require multiple agents, on average three agents, to obtain goals, so diuretics, calcium channel blockers, and β-blockers frequently are useful as second and third agents. African Americans need special consideration. They receive renoprotection from ACE inhibitors or ARBs, but as a population may lower blood pressure slightly less with these agents. It is recommended that combination therapy with a diuretic or calcium channel blocker be considered as first-line therapy. After initial therapy, which agent to add next is still controversial. Blood pressure goals are generally more difficult to achieve than glycemic goals or lipid goals in most diabetic patients. Readers are referred to Chapter 3 for more information.


Whole pancreas and islet cell transplantation are options in patients with type 1 DM; those with end-stage renal disease also receive kidney transplantation. Lifelong immunosuppression is required.

Personalized Pharmacotherapy

Individualization of therapy in DM is based on several factors. There is no optimal regimen in diabetes, and it is mostly based on reaching glycemic goals. In type 1 DM, as insulin must be used, it is to tailor the insulin regimen to the lifestyle of the patient. This almost always involves basal–bolus therapy, individualized based on SMBG readings. In addition, if the patient does not mind being attached to a pump, therapy can be further tailored to the patient. It may in unusual circumstances involve simplification of the regimen to premix insulins, as basal–bolus therapy may not fit into their lifestyle. Elevated glucagon levels in some patients, as the hormone amylin is low or absent, may require glucagon suppression therapy. FDA-approved therapy includes addition of pramlintide, but there is evidence for use of nonapproved medications such as GLP-1 receptor antagonists to act as an alternative to pramlintide, which requires many extra injections per day. In addition, metformin has been used in some type 1 DM patients who are adherent (and thus at low risk of DKA) but have suboptimal control of their FPG readings. Metformin use should not be routine, but in examples such as adolescents who miss injections after frank discussions with patients and parents.

No one regimen in type 2 DM is considered to be optimal for all patients. Common individualization points include mechanism of action, contraindications, side effects, and potential adverse events including hypoglycemia, efficacy (including fasting vs. postprandial control), long-term safety, ease of use, and cost. In addition, “nonglycemic” effects such as weight changes, lipid effects, cardiovascular outcomes, and even perceived β-cell preservation/effects may influence final choices.


A comprehensive pharmaceutical care plan for the patient with DM will integrate considerations of goals to optimize BG control and protocols to screen for, prevent, or manage microvascular and macrovascular complications. In terms of standards of care for persons with DM, one can review the document published by the ADA that outlines initial and ongoing assessments for patients with DM.5 The major performance measure by the National Committee for Quality Assurance (NCQA), such as Health Plan Employer Data and Information Set (HEDIS), should assess the ability to meet current standards of care and recognize the minimal treatment goals for glycemia, lipids, and hypertension, and provide targets for monitoring and adjusting pharmacotherapy as discussed in various sections above. Publicly reported quality measures continue to move closer to current guidelines, but lack the ability to differentiate reasons why a patient is not controlled. Glycemic control (tested minimally yearly; HbA1c <8% [<0.08; <64 mmol/mol Hb] is good control and HbA1c >9% [>0.09; >75 mmol/mol Hb] is poor control), lipid (percentage of patients with LDL <100 mg/dL [<2.59 mmol/L]), and hypertension (percentage of patients with blood pressure <130/80 mm Hg, but also with blood pressure <140/90 mm Hg) are NCQA-based measures. Glycemic control is paramount in managing type 1 or type 2 DM, but as readily identified from the above discussion, it requires frequent assessment and adjustment in diet, exercise, and pharmacologic therapies. The ADA also has clinical practice recommendations that are widely cited and followed.5 Minimally, HbA1c should be measured twice a year in patients meeting treatment goals on a stable therapeutic regimen. Quarterly assessments are recommended for those whose therapy has changed or who are not meeting glycemic goals. Fasting lipid profiles should be obtained as part of an initial assessment and thereafter at each follow-up visit if not at goal, annually if stable and at goal, or every 2 years if the lipid profile suggests low risk. Documenting regular frequency of foot examinations (each visit), urine albumin assessment (annually), dilated ophthalmologic examinations (yearly or more frequently with identified abnormalities), and office visits for follow-up are also important. Assessment for pneumococcal vaccine administration (and one-time revaccination recommended in individuals at least 65 years old), annual administration of influenza vaccine, new recommendations that all people with diabetes receive the hepatitis B vaccine series, and routine assessment for and management of other cardiovascular risks (e.g., smoking and antiplatelet therapy) are components of preventive medicine strategies. The multiplicity of assessments for each patient visit is likely to be better facilitated utilizing an integrative computer program and electronic medical record, standardized progress note forms, or flow sheets, which assist the clinician in identifying whether the patient has met standards of care in the frequency of monitoring and achievement of defined targets of therapy. Adherence continues to be of issue, as with many chronic diseases, and use of frequent education and potential simplification of regimens, if possible through combination medications, may be warranted (Table 57-13). In addition, many patients do not take medications because of tolerance, side effects, and perceived risk versus benefit from their clinician or from family, friends, or the Internet (Table 57-14).

TABLE 57-13 Combination Products Available in Type 2 Diabetes Mellitusa


TABLE 57-14 Drug Monitoring for Diabetes Mellitus Medications








    1. Centers for Disease Control and Prevention. National Diabetes Fact Sheet: National Estimates and General Information on Diabetes and Prediabetes in the United States, 2011. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, 2011.

    2. American Diabetes Association. Clinical practice recommendations. Diagnosis and classification of diabetes mellitus. Diabetes Care 2013;36(Suppl 1): S67–S74.

    3. Naik R, Lernmark A, Palmer J. Pathophysiology and genetics of type 1 (insulin-dependent) diabetes. In: Porte D Jr, Sherwin RS, Baron A, Ellenberg M, Rifkin H, eds. Ellenberg & Rifkin’s Diabetes Mellitus, 6th ed. New York, NY: McGraw-Hill, 2003:301–330.

    4. Bennett P, Rewers M, Knowler W. Epidemiology of diabetes mellitus. In: Porte D Jr, Sherwin RS, Baron A, Ellenberg M, Rifkin H, eds. Ellenberg & Rifkin’s Diabetes Mellitus, 6th ed. New York, NY: McGraw-Hill, 2003:277–300.

    5. American Diabetes Association. Standards for medical care in diabetes—2013. Diabetes Care 2013;36(Suppl 1):S11–S66.

    6. American Diabetes Association. Standards of care. Children and adolescents in: Clinical practice recommendations. Diabetes Care 2013;36(Suppl 1):S40–S43.

    7. DeFronzo RA. From the triumvirate to the ominous octet: A new paradigm for the treatment of type 2 diabetes mellitus. Diabetes 2009;58:773–795.

    8. International Association of Diabetes and Pregnancy Study Groups Consensus Panel, Metzger BE, Gabbe SG, et al. International Association of Diabetes and Pregnancy Study Groups recommendations on the diagnosis and classification of hyperglycemia in pregnancy. Diabetes Care 2010;33: 676–682.

    9. Zipitis CS, Akobeng AK. Vitamin D supplementation in early childhood and risk of type 1 diabetes: A systematic review and meta-analysis. Arch Dis Child 2008;93: 512–517.

   10. Winter W, Harris N, Schatz M. Immunological markers in the diagnosis and prediction of autoimmune type 1a diabetes. Clin Diabetes 2002;20:183–191.

   11. Cherrington AD. Banting Lecture 1997: Control of glucose uptake and release by the liver in vivo. Diabetes 1999;48:1198–1214.

   12. McGarry JD. Banting Lecture 2001: Dysregulation of fatty acid metabolism in the etiology of type 2 diabetes. Diabetes 2002;51:7–18.

   13. DeFronzo RA. Pathogenesis of type 2 diabetes mellitus: Metabolic and molecular implications for identifying diabetes genes. Diabetes 1997;5:117–269.

   14. DeFronzo RA. Pathogenesis of type 2 diabetes mellitus. Med Clin North Am 2004;88:787–835.

   15. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive summary of the third report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA 2001;285(19):2486–2497.

   16. Alberti KGMM, Eckel RH, Grundy SM, et al. Harmonizing the metabolic syndrome. A joint interim statement of the International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity. Circulation 2009;120:1640–1645.

   17. Ervin RB. Prevalence of Metabolic Syndrome among Adults 20 Years of Age and Over, by Sex, Age, Race and Ethnicity, and Body Mass Index: United States, 2003–2006. National Health Statistics Reports, No. 13. Hyattsville, MD: National Center for Health Statistics, 2009.

   18. Alexander CM, Landsman PB, Teutsch SM, et al. NCEP-defined metabolic syndrome, diabetes, and prevalence of coronary heart disease among NHANES III participants age 50 years and older. Diabetes 2003;52(5):1210–1214.

   19. Ford ES, Giles WH, Dietz WH. Prevalence of the metabolic syndrome among US adults: Findings from the third National Health and Nutrition Examination Survey. JAMA 2002;287(3):356–359.

   20. Kahn R, Buse J, Ferrannini E, Stern M, American Diabetes Association, European Association for the Study of Diabetes. The metabolic syndrome: Time for a critical appraisal: Joint statement from the American Diabetes Association and the European Association for the Study of Diabetes. Diabetes Care 2005;28(9):2289–2304.

   21. Grundy SM, Cleeman JI, Daniels SR, et al. Diagnosis and management of the metabolic syndrome: An American Heart Association/National Heart, Lung, and Blood Institute scientific statement. Circulation 2005;112(17): 2735–2752.

   22. Gaede P, Vedel P, Larsen N, et al. Multifactorial intervention and cardiovascular disease in patients with type 2 diabetes. N Engl J Med 2003;348:383–393.

   23. Diabetes Control and Complications Trial 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.

   24. UK Prospective Diabetes Study Group. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet 1998; 352:837–853.

   25. Nathan DM, Kuenen J, Borg R, et al. Translating the A1C assay into estimated average glucose values. Diabetes Care 2008;31:1473–1478.

   26. Malanda UL, Welschen LMC, Riphagen II, et al. Self-monitoring of blood glucose in patients with type 2 diabetes mellitus who are not using insulin. Cochrane Database Syst Rev 2012;(1):CD005060.

   27. American Diabetes Association. Nutrition recommendations and interventions for diabetes. Diabetes Care 2008; 31(Suppl 1):S61–S78.

   28. American Diabetes Association. Diabetes mellitus and exercise. Diabetes Care 2010;33(Suppl 1):S26–S27.

   29. Malmberg K, for the DIGAMI Study Group. Prospective randomised study of intensive insulin treatment on long term survival after acute myocardial infarction in patients with diabetes mellitus. BMJ 1997;314:1512–1515.

   30. Budnitz DS, Lovegrove MC, Shehab N, Richards CL. Emergency hospitalizations for adverse drug events in older Americans. N Engl J Med 2011;365:2002–2012.

   31. The ORIGIN Trial Investigators. Basal insulin and cardiovascular and other outcomes in dysglycemia. N Engl J Med 2012;367:319–328.

   32. Blevins T, Pullman J, Malloy J, et al. DURATION-5. Exenatide once weekly resulted in greater improvements in glycemic control compared to exenatide twice daily in patients with type 2 diabetes. J Clin Endocrinol Metab 2011;96:1301–1310.

   33. Drucker DJ, Buse JB, Taylor K, et al. Exenatide once weekly versus twice daily for the treatment of type 2 diabetes: A randomised, open-label, non-inferiority study. Lancet 2008;372:1240–1250.

   34. Schor S. The University Group Diabetes Program. A statistician looks at the mortality results. JAMA 1971; 217:1673–1675.

   35. Triplitt C. Drug interactions of medications commonly used in diabetes. Diabetes Spectrum 2006;19:202–211.

   36. 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.

   37. Dormandy JA, Charbonnel B, Eckland DJA, et al. Secondary prevention of vascular events in patients with type 2 diabetes in the PROactive study prospective pioglitazone clinical trial in macrovascular events: A randomized controlled trial. Lancet 2005;366:1279–1289.

   38. Nissen SE, Wolski K. The effect of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes. N Engl J Med 2007;356:1–15.

   39. Home PD, Pocock SJ, Beck-Nielsen H, et al. Rosiglitazone evaluated for cardiovascular outcomes in oral agent combination therapy for type 2 diabetes (RECORD): A multicentre, randomised, open-label trial. Lancet 2009; 373:2125–2135.

   40. Chiasson JL, Josse RG, Gomis R, et al. Acarbose for prevention of type 2 diabetes mellitus: The STOP-NIDDM randomized trial. Lancet 2002;359:2072–2077.

   41. Chiasson JL, Josse RG, Gomis R, et al. Acarbose treatment and the risk of cardiovascular disease and hypertension in patients with impaired glucose tolerance: The STOP-NIDDM trial. JAMA 2003;290:486–494.

   42. Stratton IM, Adler AI, Neil HA. Association of glycaemia with macrovascular and microvascular complications of type 2 diabetes (UKPDS 35): Prospective observational study. BMJ 2000;321:405–412.

   43. Adler AI, Stratton IM, Neil HA. Association of systolic blood pressure with macrovascular and microvascular complications of type 2 diabetes (UKPDS 36): Prospective observational study. BMJ 2000;321:412–419.

   44. UK Prospective Diabetes Study Group. Efficacy of atenolol and captopril in reducing risk of macrovascular and microvascular complications in type 2 diabetes: UKPDS 39. BMJ 1998;317:713–720.

   45. Writing Team for the Diabetes Control and Complications/Epidemiology of Diabetes Interventions and Complications Research Group. Sustained effect of intensive treatment of type 1 diabetes mellitus on development and progression of diabetic nephropathy: The Epidemiology of Diabetes Interventions and Complications (EDIC) study. JAMA 2003;290:2159–2167.

   46. DCCT/EDIC Study Research Group. Intensive diabetes treatment and cardiovascular disease in patients with type 1 diabetes. N Engl J Med 2005;353:2643–2653.

   47. Holman RR, Paul SK, Bethel MA, Mathews DR, Neil HAW. 10-Year follow-up of intensive glucose control in type 2 diabetes. N Engl J Med 2008;359:1577–1589.

   48. The Action to Control Cardiovascular Risk in Diabetes Study Group. Effects of intensive glucose lowering in type 2 diabetes. N Engl J Med 2008;358:2545–2559.

   49. The ADVANCE Collaborative Group. Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes. N Engl J Med 2008;358:2560–2572.

   50. Duckworth W, Abraira C, Mortiz T, et al. Glucose control and vascular complications in veterans with type 2 diabetes. N Engl J Med 2009;360:1–11.

   51. Strowig S, Raskin P. Intensive management of insulin-dependent diabetes mellitus. In: Porte D Jr, Sherwin RS, eds. Ellenberg & Rifkin’s Diabetes Mellitus, 5th ed. Stamford, CT: Appleton & Lange, 1997:709–733.

   52. Schade DS, Valentine V. To pump or not to pump. Diabetes Care 2002;25:2100–2102.

   53. Kort H, Koning EJ, Rabelink TJ, et al. Islet transplantation in type 1 diabetes. BMJ 2011;342:d247.

   54. Inzucchi SE, Bergenstal RM, Buse JB. Management of hyperglycemia in type 2 diabetes: A patient-centered approach: Position statement of the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD). Diabetes Care 2012;35: 1364–1379.

   55. Holman RR, Farmer AJ, Davies MJ, et al. Three-year efficacy of complex insulin regimens in type 2 diabetes. N Engl J Med 2009;361:1736–1747.

   56. Triplitt CL. How to initiate, titrate, and intensify insulin treatment in type 2 diabetes. US Pharm 2007;32:10–16.

   57. Araki A, Iimuro S, Sakurai T, et al. Japanese Elderly Diabetes Intervention Trial Study Group. Long-term multiple risk factor interventions in Japanese elderly diabetic patients: The Japanese Elderly Diabetes Intervention Trial: Study design, baseline characteristics and effects of intervention. Geriatr Gerontol Int 2012;12(Suppl 1):7–17.

   58. Kirkman MS, Briscoe VJ, Clark N, et al. Diabetes in older adults. Diabetes Care 2012;35:2650–2664.

   59. Sinclair AJ, Paolisso G, Castro M, et al., European Diabetes Working Party for Older People. European Diabetes Working Party for Older People 2011 clinical guidelines for type 2 diabetes mellitus. Executive summary. Diabetes Metab 2011; 37(Suppl 3):S27–S38.

   60. Langer O, Conway DL, Berkus MD, Xenakis EM. A comparison of glyburide and insulin in women with gestational diabetes mellitus. N Engl J Med 2000;343: 1134–1138.

   61. Cheng YW, Chung JH, Block-Kurbisch I, Inturrisi M, Caughey AB. Treatment of gestational diabetes mellitus: Glyburide compared to subcutaneous insulin therapy and associated perinatal outcomes. J Mater Fetal Neonatal Med 2012;25(4):379–384.

   62. Coetzee EJ, Jackson WP. Pregnancy in established non-insulin-dependent diabetics. A five-and-a-half year study at Groote Schuur Hospital. S Afr Med J 1980;58: 795–802.

   63. Coetzee EJ, Jackson WP. Oral hypoglycaemics in the first trimester and fetal outcome. S Afr Med J 1984;65: 635–637.

   64. Hellmuth E, Damm P, Mølsted-Pedersen L. Oral hypoglycaemic agents in 118 diabetic pregnancies. Diabet Med 2000;17:507–511.

   65. Rowan JA, Hague WM, Gao W, Battin MR, Moore MP; MiG Trial Investigators. Metformin versus insulin for the treatment of gestational diabetes. N Engl J Med 2008;358:2003–2015.

   66. Nicholson W, Bolen S, Witkop CT, et al. Benefits and risks of oral diabetes agents compared with insulin in women with gestational diabetes: A systematic review. Obstet Gynecol 2009;113:193.

   67. Dhulkotia JS, Ola B, Fraser R, Farrell T. Oral hypoglycemic agents vs insulin in management of gestational diabetes: A systematic review and metaanalysis. Am J Obstet Gynecol 2010;203:457.

   68. American Diabetes Association. Gestational diabetes mellitus. Diabetes Care 2004;27(1):s88–s90.

   69. Laffel L. Sick-day management in type 1 diabetes. Endocrinol Metab Clin North Am 2000;29:707–723.

   70. Kitabchi AE, Umpierrez GE, Miles JM, Fisher JN. Hyperglycemic crises in adult patients with diabetes. A consensus statement from the American Diabetes Association. Diabetes Care 2009;32:1335–1343.

   71. Hirsch IB. Sliding scale insulin—Time to stop sliding. JAMA 2009;301:213–214.

   72. Malmberg K, Norhammar A, Wedel H, Rydén L. Glycometabolic state at admission: Important risk marker of mortality in conventionally treated patients with diabetes mellitus and acute myocardial infarction: Long-term results from the Diabetes and Insulin–Glucose Infusion in Acute Myocardial Infarction (DIGAMI) study. Circulation 1999;99:2626–2632.

   73. van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin therapy in the critically ill patients. N Engl J Med 2001;345:1359–1367.

   74. Moghissi ES, Korykowski MT, DiNardo M, et al. American Association of Clinical Endocrinologists and American Diabetes Association consensus statement on inpatient glycemic control. Diabetes Care 2009;32:1119–1131.

   75. The NICE-SUGAR Study Investigators. Intensive versus conventional glycemic control in critically ill patients. N Engl J Med 2009;360:1283–1297.

   76. American Diabetes Association. Preconception care of women with diabetes. Diabetes Care 2004;27(Suppl 1): S76–S78.

   77. Mahmud M, Mazza D. Preconception care of women with diabetes: A review of current guideline recommendations. BMC Womens Health 2010;10:5.

   78. Spollett G. Hyperglycemia in HIV/AIDS. Diabetes Spectrum 2006;19:163–166.

   79. Diabetes Prevention Program Research Group. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med 2002;346:393–403.

   80. Buchanan TA, Xiang AH, Peters RK, et al. Preservation of pancreatic β-cell function and prevention of type 2 diabetes by pharmacological treatment of insulin resistance in high-risk Hispanic women. Diabetes 2002;51:2796–2803.

   81. The Dream Trial Investigators. The effect of ramipril on the incidence of diabetes. N Engl J Med 2006;355: 1551–1562.

   82. The DREAM Trial Investigators. The effect of rosiglitazone on the frequency of diabetes in patients with impaired glucose tolerance or impaired fasting glucose. A randomised controlled trial. Lancet 2006;368:1096–1105.

   83. DeFronzo R, Tripathy D, Schwenke D, et al. Pioglitazone for diabetes prevention in impaired glucose tolerance. N Engl J Med 2011;364:1104–1115.

   84. Haas L, Maryniuk M, Beck J, et al. National standards diabetes self-management education. Diabetes Care 2013;36(Suppl 1):S100–S108.

   85. Boulton AJ, Malik RA, Arezzo JC, Sosenko JM. Diabetic somatic neuropathies. Diabetes Care 2004;27:1458–1486.

   86. Boulton AJM, Armstrong DG, Albert SF, et al. Comprehensive foot exam and risk assessment. Diabetes Care 2008;31:1679–1685.

   87. Colhoun HM, Betteridge DJ, Durrington PN, et al. Primary prevention of cardiovascular disease with atorvastatin in type 2 diabetes in the Collaborative Atorvastatin Diabetes Study (CARDS): A multicentre, randomised placebo controlled trial. Lancet 2004;364:685–696.

   88. Heart Protection Study Collaborative Group. MRC/BHF Heart Protection Study of cholesterol-lowering with simvastatin in 5963 people with diabetes: A randomised placebo-controlled trial. Lancet 2003;361:2005–2016.

   89. The FIELD Study Investigators. Effects of long-term fenofibrate therapy on cardiovascular events in 9795 people with type 2 diabetes mellitus (the FIELD study): Randomised controlled trial. Lancet 2005;366: 1849–1861.

   90. The ACCORD Study Group. Effects of combination lipid therapy in type 2 diabetes. N Engl J Med 2010;362: 1563–1574 [erratum 362:1748].

   91. Grundy SM, Cleeman JI, Merz CN. Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III guidelines. Circulation 2004;110:227–239.

   92. The ACCORD Study Group. Effects of intensive blood-pressure control in type 2 diabetes mellitus. N Engl J Med 2010;362:1628–1630.