Chapter 396: Diabetes Mellitus: Diagnosis, Classification, and Pathophysiology

Diabetes mellitus (DM) refers to a group of common metabolic disorders that share the phenotype of hyperglycemia. Several distinct types of DM are caused by a complex interaction of genetics and environmental factors. Depending on the etiology of the DM, factors contributing to hyperglycemia include reduced insulin secretion, decreased glucose utilization, and increased glucose production. The metabolic dysregulation associated with DM causes secondary pathophysiologic changes in multiple organ systems that impose a tremendous burden on the individual with diabetes and on the health care system. In the United States, DM is the leading cause of end-stage renal disease (ESRD), nontraumatic lower extremity amputations, and adult blindness. It also predisposes to cardiovascular diseases. With an increasing incidence worldwide, DM is likely to continue to be a leading cause of morbidity and mortality in the future.

DM is classified on the basis of the pathogenic process leading to hyperglycemia, as opposed to earlier criteria such as age of onset or type of therapy (Fig. 396-1). There are two broad categories of DM, designated as either type 1 or type 2 DM (Table 396-1). However, there is increasing recognition of other forms of diabetes in which the molecular pathogenesis is better understood and may be associated with a single gene defect. These alternative forms may share features of type 1 and/or type 2 DM. Type 1 DM develops as a result of autoimmunity against the insulin-producing beta cells, resulting in complete or near-total insulin deficiency. Type 2 DM is a heterogeneous group of disorders characterized by variable degrees of insulin resistance, impaired insulin secretion, and increased hepatic glucose production. Distinct genetic and metabolic defects in insulin action and/or secretion give rise to the common phenotype of hyperglycemia in type 2 DM and have important therapeutic implications now that pharmacologic agents are available to target specific metabolic derangements. Both type 1 and type 2 diabetes are preceded by a period of progressive worsening of glucose homeostasis, followed by the development of hyperglycemia that exceeds the threshold for clinical diagnosis. In terms of type 2 diabetes, this phase is referred to as prediabetes and are more specifically classified as impaired fasting glucose (IFG) or impaired glucose tolerance (IGT) (Fig. 396-1). Recently, three distinct stages of type 1 DM have been defined based on the development of autoantibodies against pancreatic beta cell antigens or the development of worsening dysglycemia (discussed below).

Two features of the current classification of DM merit emphasis from previous classifications. First, the terms insulin-dependent diabetes mellitus (IDDM) or non-insulin-dependent diabetes mellitus (NIDDM) are no longer used. Because many individuals with type 2 DM eventually require insulin treatment for control of glycemia, the use of the term NIDDM generated considerable confusion. A second difference is that age or treatment modality is not a criterion. Although type 1 DM most commonly develops before the age of 30, autoimmunity against beta cells can develop at any age. It is estimated that between 5 and 10% of individuals who develop DM after age 30 have type 1 DM. Although type 2 DM more typically develops with increasing age, it is now being diagnosed more frequently in children and young adults, particularly in obese adolescents.

OTHER TYPES OF DM

Other etiologies of DM include specific genetic defects in insulin secretion or action, metabolic abnormalities that impair insulin secretion, mitochondrial abnormalities, and a host of conditions that impair glucose tolerance (Table 396-1). Maturity-onset diabetes of the young (MODY) and monogenic diabetes are subtypes of DM characterized by autosomal dominant inheritance, early onset of hyperglycemia (usually <25 years; sometimes in neonatal period), and impaired insulin secretion (discussed below). Mutations in the insulin receptor cause a group of rare disorders characterized by severe insulin resistance.

DM may also develop as a result of cystic fibrosis or chronic pancreatitis, in which the islets become damaged from a primary pathologic process originating in the pancreatic exocrine tissue. Hormones that antagonize insulin action can lead to DM. Thus, DM is often a feature of endocrinopathies such as acromegaly and Cushing’s disease. Viral infections have been implicated in pancreatic islet destruction but are an extremely rare cause of DM. A form of acute onset of type 1 diabetes, termed fulminant diabetes, has been noted in Japan and may be related to viral infection of the islets.

GESTATIONAL DM

Glucose intolerance developing during the second or third trimester of pregnancy is classified as gestational diabetes mellitus (GDM). Insulin resistance is related to the metabolic changes of pregnancy, during which the increased insulin demands may lead to IGT or diabetes. The American Diabetes Association (ADA) recommends that diabetes diagnosed within the first trimester be classified as preexisting pregestational diabetes rather than GDM. In 2015, the International Diabetes Federation (IDF) estimated that one in seven pregnancies worldwide was affected by either GDM or preexisting DM. Most women with GDM revert to normal glucose tolerance postpartum but have a substantial risk (35–60%) of developing DM in the next 10–20 years. In addition, children born to a mother with GDM also have an increased risk of developing metabolic syndrome and type 2 DM later in life. Currently, the ADA recommends that women with a history of GDM undergo lifelong screening for the development of diabetes or prediabetes at least every 3 years.

The worldwide prevalence of DM has risen dramatically over the past two decades, from an estimated 30 million cases in 1985 to 415 million in 2017 (Fig. 396-2). Based on current trends, the IDF projects that 642 million individuals will have diabetes by the year 2040 (see http://www.idf.org/). Although the prevalence of both type 1 and type 2 DM is increasing worldwide, the prevalence of type 2 DM is rising much more rapidly, presumably because of increasing obesity, reduced activity levels as countries become more industrialized, and the aging of the population. The incidence of type 1 diabetes has been increasing at a rate of 3–5% per year worldwide. The cause for this increase is not well understood, but type 1 DM is increasingly being diagnosed at younger ages. In 2015, the prevalence of diabetes in individuals aged 20–79 ranged from 7.2–11.4%. The countries with the greatest number of individuals with diabetes in 2015 are China (109.6 million), India (73 million), the United States (30.3 million), Brazil (14 million), and the Russian Federation (9 million). In the most recent estimate for the United States (2017), the Centers for Disease Control and Prevention (CDC) estimated that 9.4% of the population had diabetes, and as many as 34% of U.S. adults had prediabetes. Approximately 25% of the individuals with diabetes in the United States were undiagnosed; globally, it is estimated that as many of 50% of individuals with diabetes may be undiagnosed. The prevalence of DM increases with age. In 2015, the prevalence of DM in the United Sates was estimated to be 0.25% in individuals age <20 years, 4.1% in persons aged 20–44 years, and 16.2% in persons 45–64 years old. In individuals aged >65 years, the prevalence of DM was 25.9%. Similar age-related trends have been observed worldwide. The prevalence of diabetes is similar among men and women, but diabetes-related mortality rates are higher in men compared to women.

There is considerable geographic variation in the incidence of both type 1 and type 2 DM. Currently, Scandinavia has the highest incidence of type 1 DM; the lowest incidence is in the Pacific Rim where it is twenty- to thirtyfold lower. Northern Europe and the United States have an intermediate rate. Much of the increased risk of type 1 DM is believed to reflect the frequency of high-risk human leukocyte antigen (HLA) alleles among ethnic groups in different geographic locations.

However, now populations less enriched with these classic high-risk HLA alleles are experiencing more rapid increases in type 1 DM incidence, suggesting an influence of environmental factors.

The prevalence of type 2 DM and its harbinger, IGT, is highest in certain Pacific islands and the Middle East and intermediate in countries such as India and the United States. This variability is likely due to genetic, behavioral, and environmental factors. DM prevalence also varies among different ethnic populations within a given country, with indigenous populations usually having a greater incidence of diabetes than the general population of the country. For example, the CDC estimated that the age-adjusted prevalence of DM in the United States (age >20 years; 2010–2012) was 8% in non-Hispanic whites, 9% in Asian Americans, 13% in Hispanics, 13% in non-Hispanic blacks, and 16% in American-Indian and Alaskan native populations. The onset of type 2 DM occurs, on average, at an earlier age in ethnic groups other than non-Hispanic whites. In Asia, the prevalence of diabetes is increasing rapidly, and the diabetes phenotype appears to be somewhat different from that in the United States and Europe, with an onset at a lower body mass index (BMI) and younger age, greater visceral adiposity, and reduced insulin secretory capacity.

Diabetes is a major cause of mortality. In recent years, diabetes has been listed as the seventh leading cause of death in the United States, but several studies indicate that diabetes-related deaths are likely underreported. Data from the IDF suggests that diabetes was responsible for almost 5 million deaths worldwide, accounting for 14.5% of global all-cause mortality in adults aged 20–79 years of age. Diabetes also has important economic implications. In 2015, it was estimated that $673 billion or 12% of health care expenditures worldwide were spent on diabetes (range 5–20%). Up to 75% of individuals with diabetes live in low or middle-income countries.

Glucose tolerance is classified into three broad categories: normal glucose homeostasis, impaired glucose homeostasis, or DM. Glucose tolerance can be assessed using the fasting plasma glucose (FPG), the response to oral glucose challenge, or the hemoglobin A_1c (HbA_1c). An FPG <5.6 mmol/L (100 mg/dL), a plasma glucose <7.9 mmol/L (140 mg/dL) following an oral glucose challenge, and an HbA_1c <5.7% are considered to define normal glucose tolerance. The International Expert Committee with members appointed by the ADA, the European Association for the Study of Diabetes, and the IDF have issued diagnostic criteria for DM (Table 396-2) based on the following premises: (1) the FPG, the response to an oral glucose challenge (oral glucose tolerance test [OGTT]), and HbA_1c differ among individuals, and (2) DM is defined as the level of glycemia at which diabetes-specific complications occur rather than deviation from a population-based mean. For example, the prevalence of retinopathy in Native Americans (Pima Indian population) begins to increase at an FPG >6.4 mmol/L (116 mg/dL) (Fig. 396-3).

Abnormal glucose homeostasis (Fig. 396-1) is defined as (1) FPG = 5.6–6.9 mmol/L (100–125 mg/dL), which is defined as impaired fasting glucose (IFG); the World Health Organization uses 6.1–6.9 mmol/L (110–125 mg/dL) for IFG; (2) plasma glucose levels between 7.8 and 11 mmol/L (140 and 199 mg/dL) following an oral glucose challenge, which is termed impaired glucose tolerance (IGT); or (3) HbA_1c of 5.7–6.4%. An HbA_1c of 5.7–6.4%, IFG, and IGT do not identify the same individuals, but individuals in all three groups are at greater risk of progressing to type 2 DM, have an increased risk of cardiovascular disease, and should be counseled about ways to decrease these risks (see below). Some use the terms prediabetes, increased risk of diabetes, or intermediate hyperglycemia (World Health Organization) and slightly different metrics for this category.

These values for the FPG, the glucose following an oral glucose challenge, and HbA_1c are continuous rather than discrete variables. A FPG ≥7.0 mmol/L (126 mg/dL), a glucose ≥11.1 mmol/L (200 mg/dL) 2 h after an oral glucose challenge, or an HbA_1c ≥6.5% meets the criteria for the diagnosis of DM (Table 396-2). A random plasma glucose concentration ≥11.1 mmol/L (200 mg/dL) accompanied by classic symptoms of DM (polyuria, polydipsia, weight loss) is also sufficient for the diagnosis of DM (Table 396-2). The current criteria for the diagnosis of DM emphasize the HbA_1c or the FPG as the most reliable and convenient tests for identifying DM in asymptomatic individuals. However, some individuals may meet criteria for one test but not the other. Also, it is important to note that race and ethnicity may impact the reliability of HbA_1c levels. For example, African Americans have a higher HbA_1c value compared to non-Hispanic whites with a similar level of glycemia. An OGTT, although a valid means for diagnosing DM, is not often used in routine clinical care with the exception of pregnancy care and screening for gestational diabetes.

The diagnosis of DM has profound implications for an individual from both a medical and a financial standpoint. Thus, abnormalities on screening tests for diabetes should be repeated before making a definitive diagnosis of DM, unless acute metabolic derangements or a markedly elevated plasma glucose are present (Table 396-2). These criteria also allow for the diagnosis of DM to be withdrawn in situations when the glucose intolerance reverts to normal.

SCREENING

Widespread use of the FPG or the HbA_1c as a screening test for type 2 DM is recommended because (1) a large number of individuals who meet the current criteria for DM are asymptomatic and unaware that they have the disorder, (2) epidemiologic studies suggest that type 2 DM may be present for up to a decade before diagnosis, (3) some individuals with type 2 DM have one or more diabetes-specific complications at the time of their diagnosis, (4) treatment of type 2 DM may favorably alter the natural history of DM, (5) diagnosis of prediabetes should spur efforts for diabetes prevention. The ADA recommends screening all individuals aged >45 years every 3 years and screening individuals at an earlier age if they are overweight (BMI >25 kg/m² or ethnically relevant definition for overweight) and have one additional risk factor for diabetes (Table 396-3). Although a number of immunologic markers for type 1 DM are becoming available (discussed below), their routine use outside a clinical trial is discouraged, pending the identification of clinically beneficial interventions for individuals at high risk for developing type 1 DM.

OVERALL REGULATION OF GLUCOSE HOMEOSTASIS

Glucose homeostasis reflects a balance between energy intake from ingested food, hepatic glucose production (gluconeogenesis), and peripheral tissue glucose uptake and utilization. Insulin is the most important regulator of this metabolic equilibrium, but neural input, metabolic signals, and other hormones (e.g., glucagon) result in integrated control of glucose supply and utilization (Fig. 396-4). The organs that regulate glucose and lipids communicate by neural and humoral mechanisms with fat and muscle producing adipokines, myokines, and metabolites that influence liver function. In the fasting state, low insulin levels, together with modest increases in glucagon, increase glucose production by promoting hepatic gluconeogenesis and glycogen breakdown (glycogenolysis) and reducing glucose uptake in insulin-sensitive tissues (skeletal muscle and fat), thereby promoting mobilization of stored precursors such as amino acids and free fatty acids (lipolysis). Glucagon, secreted by pancreatic alpha cells normally only when blood glucose or insulin levels are low or during exercise, is increased in DM and stimulates glycogenolysis and gluconeogenesis by the liver and to a small degree by the renal medulla (Chap. 399). Postprandially, the glucose load elicits a rise in insulin and fall in glucagon, leading to a reversal of these processes. Insulin, an anabolic hormone, promotes the storage of carbohydrate and fat and protein synthesis. The major portion of postprandial glucose is used by skeletal muscle, an effect of insulin-stimulated glucose uptake. Other tissues, most notably the brain, use glucose in an insulin-independent fashion. Factors secreted by skeletal myocytes, adipocytes (leptin, resistin, adiponectin, etc.), and bone also influence glucose homeostasis.

INSULIN BIOSYNTHESIS

Insulin, produced by the beta cells of the pancreatic islets, is initially synthesized as a single-chain 86-amino-acid precursor polypeptide, preproinsulin. Subsequent proteolytic processing removes the amino-terminal signal peptide, giving rise to proinsulin. Proinsulin is structurally related to insulin-like growth factors I and II, which bind weakly to the insulin receptor. Cleavage of an internal 31-residue fragment from proinsulin generates C-peptide with the A (21 amino acids) and B (30 amino acids) chains of insulin being connected by disulfide bonds. The mature insulin molecule and C-peptide are stored together and co-secreted from secretory granules in the beta cells. Because C-peptide is cleared more slowly than insulin, it is a useful marker of insulin secretion and allows discrimination of endogenous and exogenous sources of insulin in the evaluation of hypoglycemia (Chaps. 399 and 80). Elevated levels of serum proinsulin have been observed in both type 1 and 2 DM and are thought to be indicative of beta cell dysfunction. Pancreatic beta cells co-secrete islet amyloid polypeptide (IAPP) or amylin, a 37-amino-acid peptide, along with insulin. The role of IAPP in normal physiology is incompletely defined, but it is the major component of the amyloid fibrils found in the islets of patients with type 2 diabetes, and an analogue is sometimes used in treating type 1 and type 2 DM. Human insulin is produced by recombinant DNA technology; structural modifications at one or more amino acid residues modify insulin’s physical and pharmacologic characteristics (Chap. 397).

INSULIN SECRETION

Glucose is the key regulator of insulin secretion by the pancreatic beta cell, although amino acids, ketones, various nutrients, gastrointestinal peptides, and neurotransmitters also influence insulin secretion. Glucose levels >3.9 mmol/L (70 mg/dL) stimulate insulin synthesis, primarily by enhancing protein translation and processing. Glucose stimulation of insulin secretion begins with its transport into the beta cell by a facilitative glucose transporter (Fig. 396-5). Glucose phosphorylation by glucokinase is the rate-limiting step that controls glucose-regulated insulin secretion. Further metabolism of glucose-6-phosphate via glycolysis generates ATP, which inhibits the activity of an ATP-sensitive K⁺ channel. This channel consists of two separate proteins: one is the binding site for certain oral hypoglycemics (e.g., sulfonylureas, meglitinides); the other is an inwardly rectifying K⁺ channel protein (Kir6.2). Inhibition of this K⁺ channel induces beta cell membrane depolarization, which opens voltage-dependent calcium channels (leading to an influx of calcium) and stimulates insulin secretion. Insulin secretory profiles reveal a pulsatile pattern of hormone release, with small secretory bursts occurring about every 10 min, superimposed upon greater amplitude oscillations of about 80–150 min. A number of metabolic pathways internal to the beta cell as well as external hormonal cues amplify glucose-stimulated insulin secretion. Glucagon-like peptide-1 (GLP-1) and glucose-dependent insulinotropic peptide (GIP) are incretin hormones that bind specific receptors on the beta cell to stimulate insulin secretion through cyclic AMP production, but have this effect only when the blood glucose is above the fasting level. Incretin hormones also suppress glucagon production and secretion. Incretin analogues or pharmacologic agents that prolong the activity of endogenous GLP-1 are used therapeutically in type 2 DM. Classically, GLP-1 release was thought to occur solely from neuroendocrine L-cells of the gastrointestinal tract following food ingestion. However, recent pre-clinical studies suggest that intraislet production of GLP-1 from alpha cells may play a role in the regulation of insulin secretion.

INSULIN ACTION

Once insulin is secreted into the portal venous system, ~50% is removed and degraded by the liver. Unextracted insulin enters the systemic circulation where it binds to receptors in target sites. Insulin binding to its receptor stimulates intrinsic tyrosine kinase activity, leading to receptor autophosphorylation and the recruitment of intracellular signaling molecules, such as insulin receptor substrates (IRS). IRS and other adaptor proteins initiate a complex cascade of phosphorylation and dephosphorylation reactions, resulting in the widespread metabolic and mitogenic effects of insulin. As an example, activation of the phosphatidylinositol-3′-kinase (PI-3-kinase) pathway stimulates translocation of a facilitative glucose transporter (e.g., GLUT4) to the cell surface, an event that is crucial for glucose uptake by skeletal muscle and fat. Activation of other insulin receptor signaling pathways induces glycogen synthesis, protein synthesis, lipogenesis, and regulation of various genes in insulin-responsive cells.

TYPE 1 DM

Type 1 DM is the result of interactions of genetic, environmental, and immunologic factors that ultimately lead to immune-mediated destruction of the pancreatic beta cells and insulin deficiency. Type 1 DM can develop at any age, but most commonly develops before 20 years of age. Most, but not all, individuals with type 1 DM have evidence of islet-directed autoimmunity. However, some individuals who have the clinical phenotype of type 1 DM lack immunologic markers indicative of an autoimmune process involving the beta cells and the genetic markers of type 1 DM. These individuals are thought to develop insulin deficiency by unknown, nonimmune mechanisms and may be ketosis prone; many are African American or Asian in heritage. The temporal decline of beta cell function and mass preceding the development of type 1 DM is shown schematically in Fig. 396-6. In susceptible individuals, the autoimmune process is thought to be triggered by an infectious or environmental stimulus. In the majority of patients, autoantibodies against beta cell antigens appear after this triggering event, followed by progressive loss of insulin secretion. The rate of decline in beta cell function varies widely among individuals, with some patients progressing rapidly to clinical diabetes and others evolving to diabetes more slowly and over a period of several years. Features of diabetes do not become evident until a threshold loss of insulin secretion and beta cell mass occurs. Autopsy studies suggest the degree of loss of beta cell mass is variable at the time of disease presentation but may be as high as 70–80%. At this point, residual, functional beta cells exist but are insufficient in number and quality to maintain glucose tolerance. The events that trigger the transition from glucose intolerance to frank diabetes are often associated with increased insulin requirements, as might occur during infections or at puberty. After the initial clinical presentation of type 1 DM, a “honeymoon” phase may ensue during which time glycemic control is achieved with modest doses of insulin or, rarely, insulin is not needed. However, this fleeting phase of endogenous insulin production from residual beta cells disappears and the individual becomes insulin deficient. Many individuals with long-standing type 1 DM produce a small amount of insulin (as reflected by C-peptide production), and some individuals with 50 years of type 1 DM have insulin-positive cells in the pancreas at autopsy.

GENETIC CONSIDERATIONS

Susceptibility to type 1 DM involves multiple genes. The concordance of type 1 DM in identical twins ranges between 40 and 60%, indicating that additional modifying factors are likely involved in determining whether diabetes develops. The major susceptibility gene for type 1 DM is located in the HLA region on chromosome 6. Polymorphisms in the HLA complex account for 40–50% of the genetic risk of developing type 1 DM. This region contains genes that encode the class II major histocompatibility complex (MHC) molecules, which present antigen to helper T cells and thus are involved in initiating the immune response (Chap. 343). The ability of class II MHC molecules to present antigen is dependent on the amino acid composition of their antigen-binding sites. Amino acid substitutions may influence the specificity of the immune response by altering the binding affinity of different antigens for class II molecules.

Most individuals with type 1 DM have the HLA DR3 and/or DR4 haplotype. Refinements in genotyping of HLA loci have shown that the haplotypes DQA1*0301, DQB1*0302, and DQB1*0201 are most strongly associated with type 1 DM. These haplotypes are present in 40% of children with type 1 DM as compared to 2% of the normal U.S. population. However, most individuals with predisposing haplotypes do not develop diabetes.

In addition to MHC class II associations, genome association studies have identified at least 20 additional genetic loci that contribute susceptibility to type 1 DM (i.e., polymorphisms in the promoter region of the insulin gene, the CTLA-4 gene, interleukin 2 receptor, and PTPN22, etc.). Among recent cohorts of individuals with new-onset type 1 diabetes, there is a decreased representation of the highest risk HLA alleles and increasing penetrance of disease in those genotypes classically associated with lower risk. Genes that confer protection against the development of the disease also exist. The haplotype DQA1*0102, DQB1*0602 is extremely rare in individuals with type 1 DM (<1%) and appears to provide protection from type 1 DM.

Although the risk of developing type 1 DM is increased tenfold in relatives of individuals with the disease, the risk is relatively low: 3–4% if the parent has type 1 DM and 5–15% in a sibling (depending on which HLA haplotypes are shared). Hence, most individuals with type 1 DM (75%) do not have a first-degree relative with this disorder.

Pathophysiology

Although other islet cell types (alpha cells [glucagon-producing], delta cells [somatostatin-producing], or PP cells [pancreatic polypeptide-producing]) are functionally and embryologically similar to beta cells, they are spared from the autoimmune destruction. However, altered patterns of hormone secretion from these other cell types in type 1 DM likely contributes to metabolic instability. Alpha cell dysfunction as reflected by fasting hyperglucagonemia, hyperglucagonemia in the post-prandial state, and an impaired glucagon response to hypoglycemia. Pathologically, the pancreatic islets have modest infiltration of lymphocytes (a process termed insulitis). After beta cells are destroyed, it is thought that the inflammatory process abates and the islets become atrophic. Studies of the autoimmune process in humans and in animal models of type 1 DM (NOD mouse and BB rat) have identified the following abnormalities in the humoral and cellular arms of the immune system: (1) islet cell autoantibodies; (2) activated lymphocytes in the islets, peripancreatic lymph nodes, and systemic circulation; (3) T lymphocytes that proliferate when stimulated with islet proteins; and (4) release of cytokines within the insulitis. Beta cells seem to be particularly susceptible to the toxic effect of some cytokines (tumor necrosis factor α [TNF-α], interferon γ, and interleukin 1 [IL-1]). The precise mechanisms of beta cell death are not known but may involve formation of nitric oxide metabolites, apoptosis, and direct CD8+ T cell cytotoxicity. The islet destruction is mediated by T lymphocytes rather than islet autoantibodies, as these antibodies do not generally react with the cell surface of islet cells and are not capable of transferring DM to animals. Efforts to suppress the autoimmune process at the time of diagnosis of diabetes have largely been ineffective or only temporarily effective in slowing beta cell destruction. Thus, increased emphasis has now been placed on interventions earlier in the disease course (i.e., during Stage 1 and 2 disease; Fig. 396-6).

Pancreatic islet molecules targeted by the autoimmune process include proinsulin, insulin, glutamic acid decarboxylase (GAD; the biosynthetic enzyme for the neurotransmitter GABA), ICA-512/IA-2 (homology with tyrosine phosphatases), and a beta cell–specific zinc transporter (ZnT-8). Most of the autoantigens are not beta cell–specific, which raises the question of how the beta cells are selectively destroyed. Current theories favor initiation of an autoimmune process directed at one beta cell molecule, which then spreads to other islet molecules as the immune process destroys beta cells and creates a series of secondary autoantigens. Stress pathways and processes arising within the beta cell may exacerbate autoimmunity through the development of modified proteins or “neoantigens” that serve as additional immune targets.

Immunologic Markers

Islet cell autoantibodies (ICAs) are a composite of several different antibodies directed at pancreatic islet molecules such as GAD, insulin, IA-2/ICA-512, and ZnT-8, and serve as a marker of the autoimmune process of type 1 DM. Assays for autoantibodies to GAD-65 and insulin are commercially available. Testing for ICAs can be useful in classifying the type of DM as type 1 and in identifying nondiabetic individuals at risk for developing type 1 DM. ICAs are present in the majority of individuals (>85%) diagnosed with new-onset type 1 DM, in a significant minority of individuals with newly diagnosed type 2 DM (5–10%), and occasionally in individuals with GDM (<5%). ICAs are present in 3–4% of first-degree relatives of individuals with type 1 DM. In combination with impaired insulin secretion after IV glucose tolerance testing, they predict a >50% risk of developing type 1 DM within 5 years. Increasing numbers of autoantibodies are associated with an increased risk of diabetes development. In children with multiple autoantibodies, ~70% developed type 1 DM after 10 years of follow-up, with 80% developing diabetes after 15-year of follow-up. At present, the measurement of ICAs in nondiabetic individuals remains a research tool because no treatments have been demonstrated to prevent the occurrence or progression to type 1 DM.

Environmental Factors

Numerous environmental events have been proposed to trigger the autoimmune process in genetically susceptible individuals; however, none have been conclusively linked to diabetes. Identification of an environmental trigger has been difficult because the event may precede the onset of DM by several years (Fig. 396-6). Putative environmental triggers include viruses (coxsackie, rubella, enteroviruses most prominently), bovine milk proteins, nitrosourea compounds, vitamin D deficiency, and environmental toxins. There is increasing interest in the microbiome and type 1 diabetes (Chap. 459).

Prevention of Type 1 DM

A number of interventions have prevented diabetes in animal models, but relatively few interventions have been tested in humans in Stages 1 and 2 of type 1 DM. The Diabetes Prevention Trial–Type 1 concluded that administering insulin (IV or PO) to individuals at high risk for developing type 1 DM did not prevent type 1 DM. However, this is an area of active clinical investigation with several trials evaluating interventions that target different aspects of the immune response in early stage type 1 DM.

TYPE 2 DM

Insulin resistance and abnormal insulin secretion are central to the development of type 2 DM. Although the primary defect is controversial, most studies support the view that insulin resistance precedes an insulin secretory defect but that diabetes develops only when insulin secretion becomes inadequate. Type 2 DM likely encompasses a range of disorders with the common phenotype of hyperglycemia. Most of our current understanding (and the discussion below) of the pathophysiology and genetics is based on studies of individuals of European descent. It is becoming increasing apparent that DM in other ethnic groups (Asian, African, and Latin American) has a somewhat different, but yet undefined, pathophysiology. In general, Latinos have greater insulin resistance and East Asians and South Asians have more beta cell dysfunction, but both defects are present in both populations. East and South Asians appear to develop type 2 DM at a younger age and a lower BMI. In some groups, DM that is ketosis prone (often in obese individuals) or ketosis-resistant (often lean) is sometimes seen.

GENETIC CONSIDERATIONS

Type 2 DM has a strong genetic component. The concordance of type 2 DM in identical twins is between 70 and 90%. Individuals with a parent with type 2 DM have an increased risk of diabetes; if both parents have type 2 DM, the risk approaches 40%. Insulin resistance, as demonstrated by reduced glucose utilization in skeletal muscle, is present in many nondiabetic, first-degree relatives of individuals with type 2 DM. The disease is polygenic and multifactorial, because in addition to genetic susceptibility, environmental factors (such as obesity, poor nutrition, and physical inactivity) modulate the phenotype. The in utero environment also contributes, and either increased or reduced birth weight increases the risk of type 2 DM in adult life. Children of pregnancies complicated by gestational hyperglycemia also exhibit an increased risk of type 2 DM. The genes that predispose to type 2 DM are incompletely identified, but genome-wide association studies have identified a large number of genes that convey a relatively small risk for type 2 DM (>70 genes, each with a relative risk of 1.06–1.5). Most prominent is a variant of the transcription factor 7–like 2 gene that has been associated with type 2 DM in several populations and with IGT. Genetic polymorphisms associated with type 2 DM have also been found in the genes encoding the peroxisome proliferator–activated receptor γ, inward rectifying potassium channel, zinc transporter, IRS, and calpain 10. The mechanisms by which these genetic loci increase the susceptibility to type 2 DM are not clear, but most are predicted to alter islet function or development or insulin secretion. Although the genetic susceptibility to type 2 DM is under active investigation (it is estimated that <10% of genetic risk is determined by loci identified thus far), it is currently not possible to use a combination of known genetic loci to predict type 2 DM.

Pathophysiology

Type 2 DM is characterized by impaired insulin secretion, insulin resistance, excessive hepatic glucose production, abnormal fat metabolism, and systemic low-grade inflammation. Obesity, particularly visceral or central (as evidenced by the hip-waist ratio), is very common in type 2 DM (≥80% of patients are obese). In the early stages of the disorder, glucose tolerance remains near-normal, despite insulin resistance, because the pancreatic beta cells compensate by increasing insulin output (Fig. 396-7). As insulin resistance and compensatory hyperinsulinemia progress, the pancreatic islets in certain individuals are unable to sustain the hyperinsulinemic state. IGT, characterized by elevations in postprandial glucose, then develops. A further decline in insulin secretion and an increase in hepatic glucose production lead to overt diabetes with fasting hyperglycemia. Ultimately, beta cell failure ensues. Possibly because of inadequate insulin suppression, glucagon is relatively overproduced and secreted, further augmenting hepatic glucose production. Although both insulin resistance and impaired insulin secretion contribute to the pathogenesis of type 2 DM, the relative contribution of each varies from individual to individual.

Metabolic Abnormalities

ABNORMAL MUSCLE AND FAT METABOLISM

Insulin resistance, the decreased ability of insulin to act effectively on target tissues (especially muscle, liver, and fat), is a prominent feature of type 2 DM and results from a combination of genetic susceptibility and obesity. Insulin resistance is relative, however, because supranormal levels of circulating insulin will normalize the plasma glucose. Insulin dose-response curves exhibit a rightward shift, indicating reduced sensitivity, and a reduced maximal response, indicating an overall decrease in maximum glucose utilization (30–60% lower than in normal individuals). Insulin resistance impairs glucose utilization by insulin-sensitive tissues and increases hepatic glucose output; both effects contribute to the hyperglycemia. Increased hepatic glucose output predominantly accounts for increased FPG levels, whereas decreased peripheral glucose utilization results in postprandial hyperglycemia. In skeletal muscle, there is a greater impairment in nonoxidative glucose usage (glycogen formation) than in oxidative glucose metabolism through glycolysis. Glucose metabolism in insulin-independent tissues is not altered in type 2 DM.

The precise molecular mechanism leading to insulin resistance in type 2 DM has not been elucidated. Insulin receptor levels and tyrosine kinase activity in skeletal muscle are reduced, but these alterations are most likely secondary to hyperinsulinemia and are not a primary defect. Therefore, “postreceptor” defects in insulin-regulated phosphorylation/dephosphorylation appear to play the predominant role in insulin resistance. Abnormalities include the accumulation of lipid intermediates within skeletal myocytes, which may impair mitochondrial oxidative phosphorylation and reduce insulin-stimulated mitochondrial ATP production. Impaired fatty acid oxidation and lipid accumulation within skeletal myocytes also may generate reactive oxygen species such as lipid peroxides. Of note, not all insulin signal transduction pathways are resistant to the effects of insulin (e.g., those controlling cell growth and differentiation using the mitogenic-activated protein kinase pathway). Consequently, hyperinsulinemia may increase the insulin action through these pathways, potentially accelerating diabetes-related conditions such as atherosclerosis.

The obesity accompanying type 2 DM, particularly in a central or visceral location, is thought to be part of the pathogenic process (Chap. 394). In addition to these white fat depots, humans now are recognized to have brown fat, which has much greater thermogenic capacity. Efforts are under way to increase the activity or quantity of brown fat. The increased adipocyte mass leads to increased levels of circulating free fatty acids and other fat cell products. For example, adipocytes secrete a number of biologic products (nonesterified free fatty acids, retinol-binding protein 4, leptin, TNF-α, resistin, IL-6, and adiponectin). In addition to regulating body weight, appetite, and energy expenditure, adipokines also modulate insulin sensitivity. The increased production of free fatty acids and some adipokines may cause insulin resistance in skeletal muscle and liver. The venous drainage of the visceral adipose beds is the portal circulation and this likely contributes to hepatic dysfunction. Free fatty acids also impair glucose utilization in skeletal muscle, promote glucose production by the liver, and impair beta cell function. In contrast, the production by adipocytes of adiponectin, an insulin-sensitizing peptide, is reduced in obesity, and this may contribute to hepatic insulin resistance. Adipocyte products and adipokines also produce an inflammatory state and may explain why markers of inflammation such as IL-6 and C-reactive protein are often elevated in type 2 DM. In addition, inflammatory cells have been found infiltrating adipose tissue.

IMPAIRED INSULIN SECRETION

Insulin secretion and sensitivity are interrelated (Fig. 396-7). In type 2 DM, insulin secretion initially increases in response to insulin resistance to maintain normal glucose tolerance. Initially, the insulin secretory defect is mild and selectively involves glucose-stimulated insulin secretion, including a greatly reduced first secretory phase. The response to other nonglucose secretagogues, such as arginine, is preserved, but overall beta cell function is reduced by as much as 50% at the onset of type 2 DM. Abnormalities in proinsulin processing are reflected by increased secretion of proinsulin in type 2 DM. Eventually, the insulin secretory defect is progressive.

The reason(s) for the decline in insulin secretory capacity in type 2 DM is unclear. The assumption is that a second genetic defect—superimposed upon insulin resistance—leads to defects in beta cell function, mass, and potentially cellular identity and differentiation status. Beta cell mass is decreased by ~50% in individuals with long-standing type 2 DM. Islet amyloid polypeptide or amylin, co-secreted by the beta cell, forms amyloid fibrillar deposits found in the islets of individuals with long-standing type 2 DM. Whether such islet amyloid deposits are a primary or secondary event is not known. The metabolic environment of diabetes also negatively impacts islet function. For example, chronic hyperglycemia paradoxically impairs islet function (“glucose toxicity”) and leads to a worsening of hyperglycemia. Improvement in glycemic control is often associated with improved islet function. In addition, elevated levels of free fatty acids (“lipotoxicity”), and systemic and local elevations in pro-inflammatory cytokines from increased numbers of islet-associated macrophages, may also worsen islet function. Reduced GLP-1 action may contribute to the reduced insulin secretion.

INCREASED HEPATIC GLUCOSE AND LIPID PRODUCTION

In type 2 DM, insulin resistance in the liver reflects the failure of hyperinsulinemia to suppress gluconeogenesis, which results in fasting hyperglycemia and decreased glycogen storage by the liver in the postprandial state. Increased hepatic glucose production occurs early in the course of diabetes, although likely after the onset of insulin secretory abnormalities and insulin resistance in skeletal muscle. As a result of insulin resistance in adipose tissue, lipolysis and free fatty acid flux from adipocytes are increased and efficiently cleared by liver leading to increased very-low-density lipoprotein [VLDL]-triglyceride) synthesis in hepatocytes and secretion from liver. This is also responsible for the dyslipidemia found in type 2 DM (elevated triglycerides, reduced high-density lipoprotein [HDL], and increased small dense low-density lipoprotein [LDL] particles). If this lipid is retained, steatosis in the liver may lead to nonalcoholic fatty liver disease and abnormal liver function tests.

Insulin Resistance Syndromes

The insulin resistance condition comprises a spectrum of disorders, with hyperglycemia representing one of the most readily diagnosed features. The metabolic syndrome, the insulin resistance syndrome, and syndrome X are terms used to describe a constellation of metabolic derangements that includes insulin resistance, hypertension, dyslipidemia (decreased HDL and elevated triglycerides), central or visceral obesity, type 2 DM or IGT/IFG, and accelerated cardiovascular disease. This syndrome is discussed in Chap. 401.

A number of relatively rare forms of severe insulin resistance include features of type 2 DM or IGT (Table 396-1). Mutations in the insulin receptor that interfere with binding or signal transduction are a rare cause of insulin resistance. Acanthosis nigricans and signs of hyperandrogenism (hirsutism, acne, and oligomenorrhea in women) are also common physical features. Two distinct syndromes of severe insulin resistance have been described in adults: (1) type A, which affects young women and is characterized by severe hyperinsulinemia, obesity, and features of hyperandrogenism; and (2) type B, which affects middle-aged women and is characterized by severe hyperinsulinemia, features of hyperandrogenism, and autoimmune disorders. Individuals with the type A insulin resistance syndrome have an undefined defect in the insulin-signaling pathway; individuals with the type B insulin resistance syndrome have autoantibodies directed at the insulin receptor. These receptor autoantibodies may block insulin binding or may stimulate the insulin receptor, leading to intermittent hypoglycemia.

Polycystic ovary syndrome (PCOS) is a common disorder that affects premenopausal women and is characterized by chronic anovulation and hyperandrogenism (Chap. 385). Insulin resistance is seen in a significant subset of women with PCOS, and the disorder substantially increases the risk for type 2 DM, independent of the effects of obesity.

Lipodystrophies are group of heterogeneous disorders characterized by selective loss of adipose tissue, leading to severe insulin resistance and hypertriglyceridemia. Lipodystrophies can be inherited or acquired and associated with variable degrees of adipose tissue loss.

Prevention

Type 2 DM is preceded by a period of IGT or IFG, and a number of lifestyle modifications and pharmacologic agents prevent or delay the onset of DM. Individuals with prediabetes or increased risk of diabetes should be referred to a structured program to reduce body weight and increase physical activity as well as being screened for cardiovascular disease. The Diabetes Prevention Program (DPP) demonstrated that intensive changes in lifestyle (diet and exercise for 30 min/d five times/week) in individuals with IGT prevented or delayed the development of type 2 DM by 58% compared to placebo. This effect was seen in individuals regardless of age, sex, or ethnic group. In the same study, metformin prevented or delayed diabetes by 31% compared to placebo. The lifestyle intervention group lost 5–7% of their body weight during the 3 years of the study; the effects of the intervention persisted for at least 15 years. Studies in Finnish and Chinese populations noted similar efficacy of diet and exercise in preventing or delaying type 2 DM. A number of agents, including α-glucosidase inhibitors, metformin, thiazolidinediones, GLP-1 receptor pathway modifiers, and orlistat, prevent or delay type 2 DM but are not approved by the Food and Drug Administration for this purpose. Individuals with a strong family history of type 2 DM and individuals with IFG or IGT should be strongly encouraged to maintain a normal BMI and engage in regular physical activity. Pharmacologic therapy for individuals with prediabetes is currently controversial because its cost-effectiveness and safety profile are not known. The ADA suggests that metformin be considered in individuals with both IFG and IGT who are at very high risk for progression to diabetes (age <60 years, BMI ≥35 kg/m², and women with a history of GDM). Individuals with IFG, IGT, or an HbA_1c of 5.7–6.4% should be monitored annually to determine if diagnostic criteria for diabetes are present.

Several monogenic forms of DM have been identified. More than 10 different variants of MODY, caused by mutations in genes encoding islet-enriched transcription factors or glucokinase (Fig. 396-5; Table 396-1), are transmitted as autosomal dominant disorders. MODY 1, MODY 3, and MODY 5 are caused by mutations in hepatocyte nuclear transcription factor (HNF) 4α, HNF-1α, and HNF-1β, respectively. As their names imply, these transcription factors are expressed in the liver but also in other tissues, including the pancreatic islets and kidney. These factors most likely affect islet development or the expression of genes important in glucose-stimulated insulin secretion or the maintenance of beta cell mass. For example, individuals with an HNF-1α mutation (MODY 3) have a progressive decline in glycemic control but may respond to sulfonylureas. In fact, some of these patients were initially thought to have type 1 DM but were later shown to respond to a sulfonylurea, and insulin was discontinued. Individuals with a HNF-1β mutation have progressive impairment of insulin secretion and hepatic insulin resistance, and require insulin treatment with minimal response to sulfonylureas. These individuals often have other abnormalities such as renal cysts, mild pancreatic exocrine insufficiency, and abnormal liver function tests. Individuals with MODY 2, the result of mutations in the glucokinase gene, have mild-to-moderate, but stable hyperglycemia that does not respond to oral hypoglycemic agents. Glucokinase catalyzes the formation of glucose-6-phosphate from glucose, a reaction that is important for glucose sensing by the beta cells (Fig. 396-5) and for glucose utilization by the liver. As a result of glucokinase mutations, higher glucose levels are required to elicit insulin secretory responses, thus altering the set point for insulin secretion. MODY 4 is a rare variant caused by mutations in pancreatic and duodenal homeobox 1, a transcription factor that regulates pancreatic development and insulin gene transcription. Homozygous inactivating mutations cause pancreatic agenesis, whereas heterozygous mutations may result in DM. Studies of populations with type 2 DM suggest that mutations in MODY-associated genes are an uncommon (<5%) cause of type 2 DM.

Transient or permanent neonatal diabetes (onset <6 months of age) occurs. Permanent neonatal diabetes is a heterogeneous group of disorders caused by genetic mutations that impact beta cell function and/or pancreatic development (Fig. 396-5). Affected individuals typically require treatment with insulin and exhibit phenotypic overlap with type 1 DM. Activating mutations in the ATP-sensitive potassium channel subunits (Kir6.2 and ABCC8) impair glucose-stimulated insulin secretion. However, these individuals may respond to sulfonylureas and can be treated with these agents. Mutations in the transcription factor GATA6 are the most common cause of pancreatic agenesis. Homozygous glucokinase mutations cause a severe form of neonatal diabetes, while mutations in mitochondrial DNA are associated with diabetes and deafness. A number of mutations identified in the coding sequence of the insulin gene have been found to interfere with proinsulin folding, processing, and bioactivity and are designated as Mutant Ins-gene-induced Diabetes of Youth (MIDYs). Some of the neonatal diabetes syndromes are associated with a spectrum of neurologic dysfunction and a variety of extrapancreatic manifestations. Any individual who developed diabetes at 6 months of age or who has atypical features of type 1 or type 2 diabetes should be screened for forms of monogenic diabetes.

APPROACH TO THE PATIENT