Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus

Diabetes mellitus is a group of metabolic diseases characterized by hyperglycemia resulting from defects in insulin secretion, insulin action, or both. The chronic hyperglycemia of diabetes is associated with long-term damage, dysfunction, and failure of various organs, especially the eyes, kidneys, nerves, heart, and blood vessels.

Several pathogenic processes are involved in the development of diabetes. These range from autoimmune destruction of the β-cells of the pancreas with consequent insulin deficiency to abnormalities that result in resistance to insulin action. The basis of the abnormalities in carbohydrate, fat, and protein metabolism in diabetes is deficient action of insulin on target tissues. Deficient insulin action results from inadequate insulin secretion and/or diminished tissue responses to insulin at one or more points in the complex pathways of hormone action. Impairment of insulin secretion and defects in insulin action frequently coexist in the same patient, and it is often unclear which abnormality, if either alone, is the primary cause of the hyperglycemia.

Symptoms of marked hyperglycemia include polyuria, polydipsia, weight loss, sometimes with polyphagia, and blurred vision. Impairment of growth and susceptibility to certain infections may also accompany chronic hyperglycemia. Acute, life-threatening consequences of diabetes are hyperglycemia with ketoacidosis or the nonketotic hyperosmolar syndrome.

Long-term complications of diabetes include retinopathy with potential loss of vision; nephropathy leading to renal failure; peripheral neuropathy with risk of foot ulcers, amputation, and Charcot joints; and autonomic neuropathy causing gastrointestinal, genitourinary, and cardiovascular symptoms and sexual dysfunction. Glycation of tissue proteins and other macromolecules and excess production of polyol compounds from glucose are among the mechanisms thought to produce tissue damage from chronic hyperglycemia. Patients with diabetes have an increased incidence of atherosclerotic cardiovascular, peripheral vascular, and cerebrovascular disease. Hypertension, abnormalities of lipoprotein metabolism, and periodontal disease are often found in people with diabetes. The emotional and social impact of diabetes and the demands of therapy may cause significant psychosocial dysfunction in patients and their families.

The vast majority of cases of diabetes fall into two broad etiopathogenetic categories (discussed in greater detail below). In one category (type 1 diabetes), the cause is an absolute deficiency of insulin secretion. Individuals at increased risk of developing this type of diabetes can often be identified by serological evidence of an autoimmune pathologic process occurring in the pancreatic islets and by genetic markers. In the other, much more prevalent category (type 2 diabetes), the cause is a combination of resistance to insulin action and an inadequate compensatory insulin secretory response. In the latter category, a degree of hyperglycemia sufficient to cause pathologic and functional changes in various target tissues, but without clinical symptoms, may be present for a long period of time before diabetes is detected. During this asymptomatic period, it is possible to demonstrate an abnormality in carbohydrate metabolism by measurement of plasma glucose in the fasting state or after a challenge with an oral glucose load.

A major requirement for epidemiological and clinical research and for the clinical management of diabetes is an appropriate system of classification that provides a framework within which to identify and differentiate its various forms and stages. While there have been a number of sets of nomenclature and diagnostic criteria proposed for diabetes, no generally accepted systematic categorization existed until the NDDG classification system was published in 1979 (1). The World Health Organization (WHO) Expert Committee on Diabetes in 1980 and, later, the WHO Study Group on Diabetes Mellitus endorsed the substantive recommendations of the NDDG (2). These groups recognized two major forms of diabetes, which they termed insulin-dependent diabetes mellitus (IDDM, type 1 diabetes) and non-insulin-dependent diabetes mellitus (NIDDM, type 2 diabetes), but their classification system went on to include evidence that diabetes mellitus was an etiologically and clinically heterogeneous group of disorders that share hyperglycemia in common. The overwhelming evidence in favor of this heterogeneity included the following:

1.  1. There are several distinct disorders, most of them rare, in which glucose intolerance is a feature.

2.  2. There are large differences in the prevalence of the major forms of diabetes among various racial or ethnic groups worldwide.

3.  3. Patients with glucose intolerance present with great phenotypic variation; take, for example, the differences between thin, ketosis-prone, insulin-dependent diabetes and obese, nonketotic, insulin-resistant diabetes.

4.  4. Evidence from genetic, immunological, and clinical studies shows that in western countries, the forms of diabetes that have their onset primarily in youth are distinct from those that have their onset mainly in adulthood.

5.  5. A type of non-insulin-requiring diabetes in young people, inherited in an autosomal dominant fashion, is clearly different from the classic acuteonset diabetes that typically occurs in children.

6.  6. In tropical countries, several clinical presentations occur, including diabetes associated with fibrocalcific pancreatitis.

These and other lines of evidence were used to divide diabetes mellitus into five distinct types (IDDM, NIDDM, gestational diabetes mellitus [GDM], malnutritionrelated diabetes, and other types). The different clinical presentations and genetic and environmental etiologic factors of the five types permitted discrimination among them. All five types were characterized by either fasting hyperglycemia or elevated levels of plasma glucose during an oral glucose tolerance test (OGTT). In addition, the 1979 classification included the category of impaired glucose tolerance (IGT), in which plasma glucose levels during an OGTT were above normal but below those defined as diabetes.

The NDDG/WHO classification highlighted the heterogeneity of the diabetic syndrome. Such heterogeneity has had important implications not only for treatment of patients with diabetes but also for biomedical research. This previous classification indicated that the disorders grouped together under the term diabetes differ markedly in pathogenesis, natural history, response to therapy, and prevention. In addition, different genetic and environmental factors can result in forms of diabetes that appear phenotypically similar but may have different etiologies.

The classification published in 1979 was based on knowledge of diabetes at that time and represented some compromises among different points of view. It was based on a combination of clinical manifestations or treatment requirements (e.g., insulin-dependent, non-insulin-dependent) and pathogenesis (e.g., malnutrition-related, “other types,” gestational). It was anticipated, however, that as knowledge of diabetes continued to develop, the classification would need revision. When the classification was prepared, a definitive etiology had not been established for any of the diabetes subclasses, except for some of the “other types.” Few susceptibility genes for diabetes had been discovered, and an understanding of the immunological basis for most type 1 diabetes was just beginning.

The current Expert Committee has carefully considered the data and rationale for what was accepted in 1979, along with research findings of the last 18 years, and we are now proposing changes to the NDDG/WHO classification scheme (Table 1). The main features of these changes are as follows:

1.  1. The terms insulin-dependent diabetes mellitus and non-insulin-dependent diabetes mellitus and their acronyms, IDDM and NIDDM, are eliminated. These terms have been confusing and have frequently resulted in classifying the patient based on treatment rather than etiology.

2.  2. The terms type 1 and type 2 diabetes are retained, with arabic numerals being used rather than roman numerals. We recommend adoption of arabic numerals in part because the roman numeral II can easily be confused by the public as the number 11. The class, or form, named type 1 diabetes encompasses the vast majority of cases that are primarily due to pancreatic islet β-cell destruction and that are prone to ketoacidosis. This form includes those cases currently ascribable to an autoimmune process and those for which an etiology is unknown. It does not include those forms of β-cell destruction or failure for which non-autoimmune-specific causes can be assigned (e.g., cystic fibrosis). While most type 1 diabetes is characterized by the presence of islet cell, GAD, IA-2, IA-2β, or insulin autoantibodies that identify the autoimmune process that leads to β-cell destruction, in some subjects, no evidence of autoimmunity is present; these cases are classified as type 1 idiopathic.

3.  3. The class, or form, named type 2 diabetes includes the most prevalent form of diabetes, which results from insulin resistance with an insulin secretory defect.

4.  4. A recent international meeting reviewed the evidence for and characteristics of malnutrition-related diabetes (3). While it appears that malnutrition may influence the expression of other types of diabetes, the evidence that diabetes can be directly caused by protein deficiency is not convincing. Therefore, the class termed malnutrition-related diabetes mellitus has been eliminated. Fibrocalculous pancreatopathy (formerly a subtype of malnutrition-related diabetes) has been reclassified as a disease of the exocrine pancreas.

5.  5. The stage termed impaired glucose tolerance (IGT) has been retained. The analogous intermediate stage of fasting glucose is named impaired fasting glucose (IFG).

6.  6. The class termed gestational diabetes mellitus (GDM) is retained as defined by the WHO and NDDG, respectively. Selective rather than universal screening for glucose intolerance in pregnancy is now recommended.

7.  7. The degree of hyperglycemia (if any) may change over time, depending on the extent of the underlying disease process (Fig. 1). A disease process may be present but may not have progressed far enough to cause hyperglycemia. The same disease process can cause IFG and/or IGT without fulfilling the criteria for the diagnosis of diabetes. In some individuals with diabetes, adequate glycemic control can be achieved with weight reduction, exercise, and/or oral glucose-lowering agents. These individuals therefore do not require insulin. Other individuals, who have some residual insulin secretion but require exogenous insulin for adequate glycemic control, can survive without it. Individuals with extensive β-cell destruction and therefore no residual insulin secretion require insulin for survival. The severity of the metabolic abnormality can progress, regress, or stay the same. Thus, the degree of hyperglycemia reflects the severity of the underlying metabolic process and its treatment more than the nature of the process itself.

8.  8. Assigning a type of diabetes to an individual often depends on the circumstances present at the time of diagnosis, and many diabetic individuals do not easily fit into a single class. For example, a person with GDM may continue to be hyperglycemic after delivery and may be determined to have, in fact, type 1 diabetes. Alternatively, a person who acquires diabetes because of large doses of exogenous steroids may become normoglycemic once the glucocorticoids are discontinued, but then may develop diabetes many years later after recurrent episodes of pancreatitis. Another example would be a person treated with thiazides who develops diabetes years later. Because thiazides in themselves seldom cause severe hyperglycemia, such individuals probably have type 2 diabetes that is exacerbated by the drug. Thus, for the clinician and patient, it is less important to label the particular type of diabetes than it is to understand the pathogenesis of the hyperglycemia and to treat it effectively.

This form of diabetes, previously encompassed by the terms insulin-dependent diabetes, type 1 diabetes, or juvenile-onset diabetes, results from a cellular-mediated autoimmune destruction of the β-cells of the pancreas (4). Markers of the immune destruction of the β-cell include islet cell autoantibodies (ICAs), autoantibodies to insulin (IAAs), autoantibodies to glutamic acid decarboxylase (GAD₆₅), and autoantibodies to the tyrosine phosphatases IA-2 and IA-2β (5,6,7,8,9,10,11,12,13). One and usually more of these autoantibodies are present in 85–90% of individuals when fasting hyperglycemia is initially detected. Also, the disease has strong HLA associations, with linkage to the DQA and B genes, and it is influenced by the DRB genes (14,15). These HLA-DR/DQ alleles can be either predisposing or protective.

In this form of diabetes, the rate of β-cell destruction is quite variable, being rapid in some individuals (mainly infants and children) and slow in others (mainly adults [16]). Some patients, particularly children and adolescents, may present with ketoacidosis as the first manifestation of the disease. Others have modest fasting hyperglycemia that can rapidly change to severe hyperglycemia and/or ketoacidosis in the presence of infection or other stress. Still others, particularly adults, may retain residual β-cell function sufficient to prevent ketoacidosis for many years. Many such individuals with this form of type 1 diabetes eventually become dependent on insulin for survival and are at risk for ketoacidosis. At this latter stage of the disease, there is little or no insulin secretion, as manifested by low or undetectable levels of plasma C-peptide. Immune-mediated diabetes commonly occurs in childhood and adolescence, but it can occur at any age, even in the 8th and 9th decades of life.

Autoimmune destruction of β-cells has multiple genetic predispositions and is also related to environmental factors that are still poorly defined. Although patients are rarely obese when they present with this type of diabetes, the presence of obesity is not incompatible with the diagnosis. These patients are also prone to other autoimmune disorders such as Graves’ disease, Hashimoto’s thyroiditis, Addison’s disease, vitiligo, and pernicious anemia.

Some forms of type 1 diabetes have no known etiologies. Some of these patients have permanent insulinopenia and are prone to ketoacidosis, but have no evidence of autoimmunity. Although only a minority of patients with type 1 diabetes fall into this category, of those who do, most are of African or Asian origin. Individuals with this form of diabetes suffer from episodic ketoacidosis and exhibit varying degrees of insulin deficiency between episodes. This form of diabetes is strongly inherited, lacks immunological evidence for β-cell autoimmunity, and is not HLA associated. An absolute requirement for insulin replacement therapy in affected patients may come and go (17).

This form of diabetes, previously referred to as non-insulin-dependent diabetes, type 2 diabetes, or adult-onset diabetes, is a term used for individuals who have insulin resistance and usually have relative (rather than absolute) insulin deficiency (18,19,20,21). At least initially, and often throughout their lifetime, these individuals do not need insulin treatment to survive. There are probably many different causes of this form of diabetes, and it is likely that the proportion of patients in this category will decrease in the future as identification of specific pathogenic processes and genetic defects permits better differentiation among them and a more definitive subclassification. Although the specific etiologies of this form of diabetes are not known, autoimmune destruction of β-cells does not occur, and patients do not have any of the other causes of diabetes listed above or below.

Most patients with this form of diabetes are obese, and obesity itself causes some degree of insulin resistance (22,23). Patients who are not obese by traditional weight criteria may have an increased percentage of body fat distributed predominantly in the abdominal region (24). Ketoacidosis seldom occurs spontaneously in this type of diabetes; when seen, it usually arises in association with the stress of another illness such as infection (25,26,27). This form of diabetes frequently goes undiagnosed for many years because the hyperglycemia develops gradually and at earlier stages is often not severe enough for the patient to notice any of the classic symptoms of diabetes (28,29,30). Nevertheless, such patients are at increased risk of developing macrovascular and microvascular complications (30,31,32,33,34). Whereas patients with this form of diabetes may have insulin levels that appear normal or elevated, the higher blood glucose levels in these diabetic patients would be expected to result in even higher insulin values had their β-cell function been normal (35). Thus, insulin secretion is defective in these patients and insufficient to compensate for the insulin resistance. Insulin resistance may improve with weight reduction and/or pharmacological treatment of hyperglycemia but is seldom restored to normal (36,37,38,39,40). The risk of developing this form of diabetes increases with age, obesity, and lack of physical activity (29,41). It occurs more frequently in women with prior GDM and in individuals with hypertension or dyslipidemia, and its frequency varies in different racial/ethnic subgroups (29,30,41). It is often associated with a strong genetic predisposition, more so than is the autoimmune form of type 1 diabetes (42,43). However, the genetics of this form of diabetes are complex and not clearly defined.

Several forms of diabetes are associated with monogenetic defects in β-cell function. These forms of diabetes are frequently characterized by onset of hyperglycemia at an early age (generally before age 25 years). They are referred to as maturity-onset diabetes of the young (MODY) and are characterized by impaired insulin secretion with minimal or no defects in insulin action (44,45,46). They are inherited in an autosomal dominant pattern. Abnormalities at three genetic loci on different chromosomes have been identified to date. The most common form is associated with mutations on chromosome 12 in a hepatic transcription factor referred to as hepatocyte nuclear factor (HNF)-1α (47,48). A second form is associated with mutations in the glucokinase gene on chromosome 7p and results in a defective glucokinase molecule (49,50). Glucokinase converts glucose to glucose-6-phosphate, the metabolism of which, in turn, stimulates insulin secretion by the β-cell. Thus, glucokinase serves as the “glucose sensor” for the β-cell. Because of defects in the glucokinase gene, increased plasma levels of glucose are necessary to elicit normal levels of insulin secretion. A third form is associated with a mutation in the HNF-4α gene on chromosome 20q (51,52). HNF-4α is a transcription factor involved in the regulation of the expression of HNF-1α. The specific genetic defects in a substantial number of other individuals who have a similar clinical presentation are currently unknown.

Point mutations in mitochondrial DNA have been found to be associated with diabetes mellitus and deafness (53,54,55). The most common mutation occurs at position 3243 in the tRNA leucine gene, leading to an A-to-G transition. An identical lesion occurs in the MELAS syndrome (mitochondrial myopathy, encephalopathy, lactic acidosis, and stroke-like syndrome); however, diabetes is not part of this syndrome, suggesting different phenotypic expressions of this genetic lesion (56).

Genetic abnormalities that result in the inability to convert proinsulin to insulin have been identified in a few families, and such traits are inherited in an autosomal dominant pattern (57,58). The resultant glucose intolerance is mild. Similarly, the production of mutant insulin molecules with resultant impaired receptor binding has also been identified in a few families and is associated with an autosomal inheritance and only mildly impaired or even normal glucose metabolism (59,60,61).

There are unusual causes of diabetes that result from genetically determined abnormalities of insulin action. The metabolic abnormalities associated with mutations of the insulin receptor may range from hyperinsulinemia and modest hyperglycemia to severe diabetes (62,63). Some individuals with these mutations may have acanthosis nigricans. Women may be virilized and have enlarged, cystic ovaries (64,65). In the past, this syndrome was termed type A insulin resistance (62). Leprechaunism and the Rabson-Mendenhall syndrome are two pediatric syndromes that have mutations in the insulin receptor gene with subsequent alterations in insulin receptor function and extreme insulin resistance (63). The former has characteristic facial features and is usually fatal in infancy, while the latter is associated with abnormalities of teeth and nails and pineal gland hyperplasia.

Alterations in the structure and function of the insulin receptor cannot be demonstrated in patients with insulin-resistant lipoatrophic diabetes. Therefore, it is assumed that the lesion(s) must reside in the postreceptor signal transduction pathways.

Any process that diffusely injures the pancreas can cause diabetes. Acquired processes include pancreatitis, trauma, infection, pancreatectomy, and pancreatic carcinoma (66,67,68). With the exception of cancer, damage to the pancreas must be extensive for diabetes to occur. However, adenocarcinomas that involve only a small portion of the pancreas have been associated with diabetes. This implies a mechanism other than simple reduction in β-cell mass. If extensive enough, cystic fibrosis and hemochromatosis will also damage β-cells and impair insulin secretion (69,70). Fibrocalculous pancreatopathy may be accompanied by abdominal pain radiating to the back and pancreatic calcifications on X ray (71). Pancreatic fibrosis and calcium stones in the exocrine ducts have been found at autopsy.

Several hormones (e.g., growth hormone, cortisol, glucagon, epinephrine) antagonize insulin action. Excess amounts of these hormones (e.g., acromegaly, Cushing’s syndrome, glucagonoma, pheochromocytoma) can cause diabetes (72,73,74,75). This generally occurs in individuals with preexisting defects in insulin secretion, and hyperglycemia typically resolves when the hormone excess is removed.

Somatostatinoma- and aldosteronoma-induced hypokalemia can cause diabetes, at least in part, by inhibiting insulin secretion (75,76). Hyperglycemia generally resolves after successful removal of the tumor.

Many drugs can impair insulin secretion. These drugs may not cause diabetes by themselves, but they may precipitate diabetes in individuals with insulin resistance (77,78). In such cases, the classification is unclear because the sequence or relative importance of β-cell dysfunction and insulin resistance is unknown. Certain toxins such as Vacor (a rat poison) and intravenous pentamidine can permanently destroy pancreatic β-cells (79,80,81,82). Such drug reactions fortunately are rare. There are also many drugs and hormones that can impair insulin action. Examples include nicotinic acid and glucocorticoids (77,78). Patients receiving α-interferon have been reported to develop diabetes associated with islet cell antibodies and, in certain instances, severe insulin deficiency (83,84). The list shown in Table 1 is not all-inclusive, but reflects the more commonly recognized drug-, hormone-, or toxin-induced forms of diabetes.

Certain viruses have been associated with β-cell destruction. Diabetes occurs in patients with congenital rubella (85), although most of these patients have HLA and immune markers characteristic of type 1 diabetes. In addition, coxsackievirus B, cytomegalovirus, adenovirus, and mumps have been implicated in inducing certain cases of the disease (86,87,88).

In this category, there are two known conditions, and others are likely to occur. The stiff-man syndrome is an autoimmune disorder of the central nervous system characterized by stiffness of the axial muscles with painful spasms (89). Patients usually have high titers of the GAD autoantibodies and approximately one-third will develop diabetes.

Anti-insulin receptor antibodies can cause diabetes by binding to the insulin receptor, thereby blocking the binding of insulin to its receptor in target tissues (63). However, in some cases, these antibodies can act as an insulin agonist after binding to the receptor and can thereby cause hypoglycemia. Anti-insulin receptor antibodies are occasionally found in patients with systemic lupus erythematosus and other autoimmune diseases (63). As in other states of extreme insulin resistance, patients with anti-insulin receptor antibodies often have acanthosis nigricans. In the past, this syndrome was termed type B insulin resistance.

Many genetic syndromes are accompanied by an increased incidence of diabetes mellitus (90). These include the chromosomal abnormalities of Down’s syndrome, Kline-felter’s syndrome, and Turner’s syndrome. Wolfram’s syndrome is an autosomal recessive disorder characterized by insulin-deficient diabetes and the absence of β-cells at autopsy (91). Additional manifestations include diabetes insipidus, hypogonadism, optic atrophy, and neural deafness. Other syndromes are listed in Table 1.

GDM is defined as any degree of glucose intolerance with onset or first recognition during pregnancy. The definition applies regardless of whether insulin or only diet modification is used for treatment or whether the condition persists after pregnancy. It does not exclude the possibility that unrecognized glucose intolerance may have antedated or begun concomitantly with the pregnancy (92). Six weeks or more after pregnancy ends, the woman should be reclassified, as described below (see diagnostic criteria for diabetes mellitus), into one of the following categories: 1) diabetes, 2) IFG, 3) IGT, or 4) normoglycemia. In the majority of cases of GDM, glucose regulation will return to normal after delivery.

GDM complicates ∼4% of all pregnancies in the U.S., resulting in ∼135,000 cases annually (93). The prevalence may range from 1 to 14% of pregnancies, depending on the population studied (93). GDM represents nearly 90% of all pregnancies complicated by diabetes (94). Clinical recognition of GDM is important because therapy, including medical nutrition therapy, insulin when necessary, and antepartum fetal surveillance, can reduce the well-described GDM-associated perinatal morbidity and mortality (95). Maternal complications related to GDM also include an increased rate of cesarean delivery and chronic hypertension (95,96,97). Although many patients diagnosed with GDM will not develop diabetes later in life, others will be diagnosed many years postpartum as having type 1 diabetes, type 2 diabetes, IFG, or IGT (98,99,100,101,102,103).

Deterioration of glucose tolerance occurs normally during pregnancy, particularly in the 3rd trimester. The criteria for abnormal glucose tolerance in pregnancy, which are widely used in the U.S., were proposed by O’Sullivan and Mahan (98) in 1964 and were based on data obtained from OGTTs performed on 752 pregnant women. Abnormal glucose tolerance was defined as two or more blood glucose values out of four that were greater than or equal to two standard deviations above the mean. These values were set based on the prediction of diabetes developing later in life.

In 1979, the NDDG revised the O’Sullivan and Mahan criteria, converting the whole blood values to plasma values (1). These criteria were adopted by the American Diabetes Association and the American College of Obstetricians and Gynecologists (ACOG) (104), but are at variance with WHO criteria.

Carpenter and Coustan (105) suggested that the NDDG conversion of the O’Sullivan and Mahan values from the original Somogyi-Nelson determinations may have resulted in values that are too high. They proposed cutoff values for plasma glucose that appear to represent more accurately the original O’Sullivan and Mahan determinations. In three studies, these criteria identified more patients with GDM whose infants had perinatal morbidity (106,107,108). Additional studies have been completed to define abnormal 75-g OGTT values in different populations (109,110,111). This method has provided values for plasma glucose concentrations that are similar to the Carpenter/Coustan extrapolations of the 100-g OGTT.

Recommendations from the American Diabetes Association’s Fourth International Workshop-Conference on Gestational Diabetes Mellitus held in March 1997 support the use of the Carpenter/Coustan diagnostic criteria as well as the alternative use of a diagnostic 75-g 2-h OGTT (111a). These criteria are summarized below.

Previous recommendations have been that screening for GDM be performed in all pregnancies. However, there are certain factors that place women at lower risk for the development of glucose intolerance during pregnancy, and it is likely not cost-effective to screen such patients. This low-risk group comprises women who are <25 years of age and of normal body weight, have no family history (i.e., first-degree relative) of diabetes, have no history of abnormal glucose metabolism or poor obstetric outcome, and are not members of an ethnic/racial group with a high prevalence of diabetes (e.g., Hispanic American, Native American, Asian American, African-American, Pacific Islander) (112,113,114). Pregnant women who fulfill all of these criteria need not be screened for GDM.

Risk assessment for GDM should be undertaken at the first prenatal visit. Women with clinical characteristics consistent with a high risk of GDM (marked obesity, personal history of GDM, glycosuria, or a strong family history of diabetes) should undergo glucose testing (see below) as soon as feasible. If they are found not to have GDM at that initial screening, they should be retested between 24 and 28 weeks of gestation. Women of average risk should have testing undertaken at 24–28 weeks of gestation.

A fasting plasma glucose level >126 mg/dl (7.0 mmol/l) or a casual plasma glucose >200 mg/dl (11.1 mmol/l) meets the threshold for the diagnosis of diabetes, if confirmed on a subsequent day, and precludes the need for any glucose challenge. In the absence of this degree of hyperglycemia, evaluation for GDM in women with average or high-risk characteristics should follow one of two approaches:

One-step approach:

Perform a diagnostic OGTT without prior plasma or serum glucose screening. The one-step approach may be cost-effective in high-risk patients or populations (e.g., some Native-American groups).

Two-step approach:

Perform an initial screening by measuring the plasma or serum glucose concentration 1 h after a 50-g oral glucose load (glucose challenge test [GCT]) and perform a diagnostic OGTT on that subset of women exceeding the glucose threshold value on the GCT. When the two-step approach is employed, a glucose threshold value >140 mg/dl (7.8 mmol/l) identifies approximately 80% of women with GDM, and the yield is further increased to 90% by using a cutoff of >130 mg/dl (7.2 mmol/l).

With either approach, the diagnosis of GDM is based on an OGTT. Diagnostic criteria for the 100-g OGTT are derived from the original work of O’Sullivan and Mahan, modified by Carpenter and Coustan, and are shown in the top of Table 2. Alternatively, the diagnosis can be made using a 75-g glucose load and the glucose threshold values listed for fasting, 1 h, and 2 h (Table 2, bottom); however, this test is not as well validated as the 100-g OGTT.

The terms IGT and IFG refer to a metabolic stage intermediate between normal glucose homeostasis and diabetes. This stage includes individuals who have IGT and individuals with fasting glucose levels ≥110 mg/dl (6.1 mmol/l) but <126 mg/dl (7.0 mmol/l) (IFG). The term IFG was coined by Charles et al. (115) to refer to a fasting plasma glucose (FPG) level ≥110 mg/dl (6.1 mmol/l) but <140 mg/dl (7.8 mmol/l). We are using a similar definition, but with the upper end lowered to correspond to the new diagnostic criteria for diabetes. A fasting glucose concentration of 109 mg/dl (6.1 mmol/l) has been chosen as the upper limit of “normal.” Although it is recognized that this choice is somewhat arbitrary, it is near the level above which acute phase insulin secretion is lost in response to intravenous administration of glucose (116) and is associated with a progressively greater risk of developing micro- and macrovascular complications (117,118,119,120,121).

Note that many individuals with IGT are euglycemic in their daily lives (122) and may have normal or near normal glycated hemoglobin levels (123). Individuals with IGT often manifest hyperglycemia only when challenged with the oral glucose load used in the standardized OGTT.

In the absence of pregnancy, IFG and IGT are not clinical entities in their own right but rather risk factors for future diabetes and cardiovascular disease (117). They can be observed as intermediate stages in any of the disease processes listed in Table 1. IFG and IGT are associated with the insulin resistance syndrome (also known as syndrome X or the metabolic syndrome), which consists of insulin resistance, compensatory hyperinsulinemia to maintain glucose homeostasis, obesity (especially abdominal or visceral obesity), dyslipidemia of the high-triglyceride and/or low-HDL type, and hypertension (124). Insulin resistance is directly involved in the pathogenesis of type 2 diabetes. IFG and IGT appear as risk factors for this type of diabetes at least in part because of their correlation with insulin resistance. In contrast, the explanation for why IFG and IGT are also risk factors for cardiovascular disease is less clear. The insulin resistance syndrome includes well-recognized cardiovascular risk factors such as low HDL levels and hypertension. In addition, it includes hypertriglyceridemia, which is highly correlated with small dense LDL and increased plasminogen activator inhibitor-1 (PAI-1) levels. The former is thought to have enhanced atherogenicity, perhaps as a result of its greater vulnerability to oxidation than normal LDL. PAI-1 is a cardiovascular risk factor probably because it inhibits fibrinoloysis. Thus, the insulin resistance syndrome contains many features that increase cardiovascular risk. IFG and IGT may not in themselves be directly involved in the pathogenesis of cardiovascular disease, but rather may serve as statistical risk factors by association because they correlate with those elements of the insulin resistance syndrome that are cardiovascular risk factors.

The diagnostic criteria for diabetes mellitus have been modified from those previously recommended by the NDDG (1) or WHO (2). The revised criteria for the diagnosis of diabetes are shown in Table 3. Three ways to diagnose diabetes are possible, and each must be confirmed, on a subsequent day, by any one of the three methods given in Table 3. For example, one instance of symptoms with casual plasma glucose ≥200 mg/dl (11.1 mmol/l), confirmed on a subsequent day by 1) FPG ≥126 mg/dl (7.0 mmol/l), 2) an OGTT with the 2-h postload value ≥200 mg/dl (11.1 mmol/l), or 3) symptoms with a casual plasma glucose ≥200 mg/dl (11.1 mmol/l), warrants the diagnosis of diabetes.

For epidemiological studies, estimates of diabetes prevalence and incidence should be based on an FPG ≥126 mg/dl (7.0 mmol/l). This recommendation is made in the interest of standardization and also to facilitate field work, particularly where the OGTT may be difficult to perform and where the cost and demands on participants’ time may be excessive. This approach will lead to slightly lower estimates of prevalence than would be obtained from the combined use of the FPG and OGTT (Table 4).

The Expert Committee recognizes an intermediate group of subjects whose glucose levels, although not meeting criteria for diabetes, are nevertheless too high to be considered altogether normal. This group is defined as having FPG levels ≥110 mg/dl (6.1 mmol/l) but <126 mg/dl (7.0 mmol/l) or 2-h values in the OGTT of ≥140 mg/dl (7.8 mmol/l) but <200 mg/dl (11.1 mmol/l). Thus, the categories of FPG values are as follows:

• FPG <110 mg/dl (6.1 mmol/l) = normal fasting glucose;

• FPG ≥110 (6.1 mmol/l) and <126 mg/dl (7.0 mmol/l) = IFG;

• FPG ≥126 mg/dl (7.0 mmol/l) = provisional diagnosis of diabetes (the diagnosis must be confirmed, as described above).

The corresponding categories when the OGTT is used are the following:

• 2-h postload glucose (2-h PG) <140 mg/dl (7.8 mmol/l) = normal glucose tolerance;

• 2-h PG ≥140 (7.8 mmol/l) and <200 mg/dl (11.1 mmol/l) = IGT;

• 2-h PG ≥200 mg/dl (11.1 mmol/l) = provisional diagnosis of diabetes (the diagnosis must be confirmed, as described above).

Since the 2-h OGTT cutoff of 140 mg/dl (7.8 mmol/l) will identify more people as having impaired glucose homeostasis than will the fasting cutoff of 110 mg/d (6.1 mmol/l), it is essential that investigator always report which test was used.

The revised criteria are still based on measures of hyperglycemia. Whereas many different diagnostic schemes have been used all have been based on some measurement of blood or urine glucose, as reviewed by McCance et al. (125). The metabolic defects underlying hyperglycemia, such as islet cell autoimmunity or insulin resistance, should be referred to independently from the diagnosis of diabetes, i.e., in the classification of the disease. Determining the optimal diagnostic level of hyperglycemia depends on a balance between the medical, social, and economic costs of making a diagnosis in someone who is not truly at substantial risk of the adverse effects of diabetes and those of failing to diagnose someone who is (126). Unfortunately, not all these data are available, so we relied primarily on medical data.

Plasma glucose concentrations are distributed over a continuum, but there is an approximate threshold separating those subjects who are at substantially increased risk for some adverse outcomes caused by diabetes (e.g., microvascular complications from those who are not. Based in part on estimates of the thresholds for microvascular disease, the previous WHO criteria defined diabetes by FPG ≥140 mg/dl (7.8 mmol/l), 2-h PG ≥200 mg/dl (11.1 mmol/l) in the OGTT, or both. These criteria effectively defined diabetes by the 2-h PG alone because the fasting and 2-h cutpoint values are not equivalent. Almost all individuals with FPG ≥140 mg/dl (7.8 mmol/l) have 2-h PG ≥200 mg/dl (11.1 mmol/l) if given an OGTT, whereas only about one-fourth of those with 2-h PG ≥200 mg/dl (11.1 mmol/l) and without previously known diabetes have FPG ≥140 mg/dl (7.8 mmol/l) (127). Thus, the cutpoint of FPG ≥140 mg/dl (7.8 mmol/l) defined a greater degree of hyperglycemia than did the cutpoint of 2-h PG ≥200 mg/dl (11.1 mmol/l). It is the consensus of the Expert Committee that this discrepancy is unwarranted and that the cutpoint values for both tests should reflect a similar degree of hyperglycemia and risk of adverse outcomes.

Under the previous WHO and the NDDG criteria, the diagnosis of diabetes is largely a function of which test is performed. Many individuals who would have 2-h PG ≥200 mg/dl (11.1 mmol/l) in an OGTT are not tested with an OGTT because they lack symptoms or because they have an FPG <140 mg/dl (7.8 mmol/l). Thus, if it is desired that all people with diabetes be diagnosed and the previous criteria are followed, OGTTs must be performed periodically in everyone. However, in ordinary practice, not only is the OGTT performed infrequently, but it is usually not used even to confirm suspected cases (128). In summary, the diagnostic criteria are now revised to 1) avoid the discrepancy between the FPG and 2-h PG cutpoint values and 2) facilitate and encourage the use of a simpler and equally accurate test—fasting plasma glucose—for diagnosing diabetes.

The cutpoint for the 2-h PG has been justified largely because at approximately that point the prevalence of the microvascular complications considered specific for diabetes (i.e., retinopathy and nephropathy) increases dramatically. This property of the 2-h PG has been compared with the FPG in population studies of the Pima Indians in the U.S., among Egyptians, and in the Third National Health and Nutrition Examination Survey (NHANES III) in the U.S. In other studies, the relationships between glycemia and macrovascular disease have also been examined.

The relationships of FPG and 2-h PG to the development of retinopathy were evaluated in Pima Indians over a wide range of plasma glucose cutpoints (Fig. 2A) (129). Both variables were similarly associated with retinopathy, indicating that by this criterion, each could work equally well for diagnosing diabetes. The authors concluded that both measures were equivalent in terms of the properties previously used to justify diagnostic criteria.

These findings were confirmed in a similar study in Egypt, in which the FPG and 2-h PG were each strongly and equally associated with retinopathy (Fig. 2B) (130). For both the FPG and the 2-h PG, the prevalence of retinopathy was markedly higher above the point of intersection of the two components of the bimodal frequency distribution (FPG = 129 mg/dl [7.2 mmol/l] and 2-h PG = 207 mg/dl [11.5 mmol/l]).

In the NHANES III, 2,821 individuals aged 40–74 years received an OGTT, a measurement of HbA_1c, and an assessment of retinopathy by fundus photography (K. Flegal, personal communication). Figure 2C shows that all three measures of glycemia (FPG, 2-h PG, and HbA_1c) are strongly associated with retinopathy, which is similar to the relationships found in the Pima Indians (129) and Egyptians (130), although the relationship was strongest for 2-h PG. As in other studies, the prevalence rose dramatically in the highest decile of each variable, corresponding to FPG ≥120 mg/dl (6.7 mmol/l), 2-h PG ≥195 mg/dl (10.8 mmol/l), and HbA_1c ≥6.2%. As in the Pima Indian (129) and Egyptian (130) studies, estimates of these “thresholds” for retinopathy are somewhat imprecise. More precision cannot easily be obtained by using narrower glycemic intervals (e.g., 20 instead of the 10 shown in Fig. 2) because of the limited numbers of cases of retinopathy in each sample (32 cases in the Pima study, 146 in the Egyptian study, and 111 in NHANES III). There are no absolute thresholds because some retinopathy occurred at all glucose levels, presumably because of measurement or disease variability and because of nondiabetic causes of retinopathy.

The associations between FPG and 2-h PG and macrovascular disease have been examined in adults without known diabetes (131). The 2-h PG was somewhat more closely associated with major coronary heart disease, but there was no significant difference in the association of the FPG or the 2-h PG with other indexes of macrovascular disease. Similarly, the relationship between glycemia and peripheral arterial disease was studied in 50- to 74-year-old Caucasians (132). The prevalence of arterial disease was strongly related to the FPG and 2-h PG. The associations appeared to be of the same strength for both variables.

In a recent analysis of the Paris Prospective Study, the incidence of fatal coronary heart disease was related to both FPG and 2-h PG determined at a baseline examination (118). Incidence rates were markedly increased at FPG ≥125 mg/dl (6.9 mmol/l) or 2-h PG ≥140 mg/dl (7.8 mmol/l). Similarly, the incidence of coronary artery disease and the all-cause mortality rates were predicted by the FPG in the Baltimore Longitudinal Study of Aging (R. Andres, C. Coleman, D. Elahi, J. Fleg, D.C. Muller, J.D. Sorkin, J.D. Tobin, personal communication). The incidence rates of both these outcomes increased markedly and almost linearly above FPG levels in the range of 110–120 mg/dl (6.1–6.7 mmol/l). In conclusion, both the FPG and 2-h PG provide important information regarding risk of both micro- and macrovascular disease, and the approximate thresholds for increased risk correspond with those for retinopathy and with the revised diagnostic criteria.

Reproducibility is another important property of a diagnostic test, a property for which the FPG appears to be preferable. When OGTTs were repeated in adults during a 2- to 6-week interval, the intra-individual coefficients of variation were 6.4% for the FPG and 16.7% for the 2-h PG (133).

It is important to review the rationale for retaining the diagnostic cutpoint of 200 mg/dl (11.1 mmol/l) for the 2-h PG. This cutpoint was originally adopted for three reasons (1,2). First, 200 mg/dl (11.1 mmol/l) has been found to approximate the cutpoint separating the two components of the bimodal distribution of 2-h PG. Second, in several studies, the prevalence of microvascular disease sharply increased above 2-h PG levels of ∼200 mg/dl (11.1 mmol/l). Third, an enormous body of clinical and epidemiological data has been collected based on the 2-h PG cutpoint of 200 mg/dl (11.1 mmol/l). Thus, this value has been retained for the diagnosis of diabetes because it would be very disruptive, and add little benefit, to alter the well-accepted 2-h PG diagnostic level of ≥200 mg/dl (11.1 mmol/l).

Changing the diagnostic cutpoint for the FPG to 126 mg/dl (7.0 mmol/l) is based on the belief that the cutpoints for the FPG and 2-h PG should diagnose similar conditions, given the equivalence of the FPG and the 2-h PG in their associations with vascular complications and their discrimination between two components of a bimodal frequency distribution (129,130). McCance et al. (129) computed the FPG level equivalent (in sensitivity and specificity for retinopathy) to the 1985 WHO criterion of the 2-h PG ≥200 mg/dl (11.1 mmol/l) and found it to be an FPG of ≥123 mg/dl (6.8 mmol/l) (Table 5). Finch et al. (134) approached the problem in each of 13 Pacific populations surveyed with OGTTs by determining the value in the FPG that, when used alone as a diagnostic criterion, gave the same prevalence of diabetes as did 2-h PG ≥200 mg/dl (11.1 mmol/l). The summary estimate from all these populations was a cutpoint of 126 mg/dl (7.0 mmol/l). The same method was applied to data derived from the Pima Indians and resulted in an FPG cutpoint of 120 mg/dl (6.7 mmol/l). In NHANES III, the corresponding cutpoint was 121 mg/dl (6.7 mmol/l) (Table 5). These values and the 2-h PG cutpoint of 200 mg/dl (11.1 mmol/l) are also quite similar to the values of FPG 129 mg/dl (7.2 mmol/l) and 2-h PG 207 mg/dl (11.5 mmol/l) that separated the components of the bimodal frequency distributions and identified individuals with a high prevalence of retinopathy among Egyptians (130). Because the standard errors of these estimates are not known, the small differences in the estimates shown in Table 5 may be consistent with sampling variability.

We chose a cutpoint at the upper end of these estimates (FPG ≥126 mg/dl, 7.0 mmol/l). This value is slightly higher than most of the estimated cutpoints that would give the same prevalence of diabetes as the criterion of 2-h PG ≥200 mg/dl (11.1 mmol/l). That is, slightly fewer people will be diagnosed with diabetes if the new FPG criterion is used alone than if either the FPG or the OGTT is used and interpreted by the previous WHO and NDDG criteria (Table 4).

As noted above, although the OGTT is an acceptable diagnostic test and has been an invaluable tool in research, it is not recommended for routine use. Because of its inconvenience to patients and the perception by many physicians that it is unnecessary, the OGTT is already not widely used for diagnosing diabetes. In addition, it is more costly and time-consuming than the FPG, and the repeat test reproducibility of the 2-h PG is worse than that of the FPG (133). If the OGTT is used, either for clinical or research purposes, the test procedure methods recommended by the WHO (2) and the diagnostic criterion in Table 3 should be employed.

HbA_1c measurement is not currently recommended for diagnosis of diabetes, although some studies have shown that the frequency distributions for HbA_1c have characteristics similar to those of the FPG and the 2-h PG. Moreover, these studies have defined an HbA_1c level above which the likelihood of having or developing macro- or microvascular disease rises sharply (Fig. 2) (129,130,131,132). Furthermore, HbA_1c and FPG (in type 2 diabetes) have become the measurements of choice in monitoring the treatment of diabetes, and decisions on when and how to implement therapy are often made on the basis of HbA_1c. These observations have led some to recommend HbA_1c measurement as a diagnostic test (126,135).

On the other hand, there are many different methods for the measurement of HbA_1c and other glycosylated proteins, and nationwide standardization of the HbA_1c test has just begun (136). Studies of the utility of the test compared with the FPG and 2-h PG have used different assays, thereby making it difficult to assign an appropriate cutpoint. Also, the FPG, 2-h PG, and HbA_1c tests are imperfectly correlated. In most clinical laboratories, a “normal” HbA_1c is usually based on a statistical sampling of healthy, presumably nondiabetic individuals. In conclusion, HbA_1c remains a valuable tool for monitoring glycemia, but it is not currently recommended for the diagnosis of diabetes.

The revised criteria are for diagnosis and are not treatment criteria or goals of therapy. No change is made in the American Diabetes Association’s recommendations of FPG <120 mg/dl (6.7 mmol/l) and HbA_1c <7% as treatment goals (137). The new diagnostic cutpoint (FPG ≥126 mg/dl [7.0 mmol/l]) is based on the observation that this degree of hyperglycemia usually reflects a serious metabolic abnormality that has been shown to be associated with serious complications. The treatment of nonpregnant patients with hyperglycemia near the cutpoint should begin with an individualized lifestyle-modification regimen (i.e., meal planning and exercise). Initiation of pharmacological therapy in these patients has not yet been shown to improve prognosis and may lead to an unacceptably high incidence of hypoglycemic reactions with certain drugs (e.g., sulfonylureas, insulin).

The new criteria have implications for estimates of the prevalence of diabetes. Although an FPG ≥126 mg/dl (7.0 mmol/l) and a 2-h PG ≥200 mg/dl (11.1 mmol/l) have similar predictive value for adverse outcomes, the two tests are not perfectly correlated with each other. A given person may have one glucose value above one cutpoint and another value below the other cutpoint. Thus, simultaneous measurement of both FPG and 2-h PG will inevitably lead to some diagnostic discrepancies and dilemmas. Although diagnosing diabetes by either test will result in a similar number of “cases,” different individuals in different hyperglycemic stages may be identified. (This situation would be even more complicated if a third diagnostic test, such as HbA_1c, were used.) However, according to the data reviewed above, there is no basis for concluding that the 2-h PG is more reliable than the FPG. Thus, the FPG alone should be used for estimating the comparative prevalence of diabetes in different populations.

Table 4 shows the effect of the new diagnostic criteria on the estimated prevalence of diabetes in the U.S. population aged 40–74 years using data from NHANES III. Diagnosing diabetes in those without a medical history of diabetes by using only the FPG test would result in a lower prevalence of diabetes than would using WHO criteria (4.35 vs. 6.34%). The total prevalence of diabetes (including those with a medical history) would be 12.27%, or 14% lower than the prevalence of 14.26% by the WHO criteria. Of note, these prevalence estimates refer to results of testing on one occasion. The prevalence of diabetes confirmed by a second test will be lower regardless of which criteria are used.

Widespread adoption of the new criteria may, however, have a large impact on the number of people actually diagnosed with diabetes. Presently, about half the adults with diabetes in the U.S. are undiagnosed (127), but many might now be diagnosed if the simpler FPG test were always used.

Type 1 diabetes is usually an autoimmune disease, characterized by the presence of a variety of autoantibodies to protein epitopes on the surface of or within the β-cells of the pancreas. The presence of such markers before the development of overt disease can identify patients at risk (138). For example, those with more than one autoantibody (i.e., ICA, IAA, GAD, IA-2) are at high risk (139,140,141). At this time, however, many reasons preclude the recommendation to test individuals routinely for the presence of any of the immune markers outside of a clinical trials setting. First, cutoff values for some of the assays for immune markers have not been completely established for clinical settings. Second, there is no consensus yet as to what action should be taken when a positive autoantibody test is obtained. Thus, autoantibody testing may identify people at risk of developing type 1 diabetes without offering any proven measures that might prevent or delay the clinical onset of disease. Of note, however, is that there are a number of ongoing well-controlled clinical studies testing various means of preventing type 1 diabetes. These studies conducted in high-risk subjects may one day offer an effective means to prevent type 1 diabetes, in which case screening may become appropriate. Last, because the incidence of type 1 diabetes is low, routine testing of healthy children will identify only the small number (<0.5%) who at that moment may be “prediabetic.” Thus, the cost-effectiveness of such screening is questionable, at least until an effective therapy is available. For the above reasons, the clinical testing of individuals for autoantibodies related to type 1 diabetes, outside of research studies, cannot be recommended at this time. Similarly, antibody testing of high-risk individuals (e.g., siblings of type 1 patients) is also not recommended until the efficacy and safety of therapies to prevent or delay type 1 diabetes have been demonstrated. On the other hand, the autoantibody tests may be of value to identify which newly diagnosed patients have immune-mediated type 1 diabetes in circumstances where it is not obvious, particularly when therapies become available to preserve β-cell mass.

Undiagnosed type 2 diabetes is common in the U.S. As many as 50% of the people with the disease, or about 8 million individuals, are undiagnosed (127). Of concern, there is epidemiological evidence that retinopathy begins to develop at least 7 years before the clinical diagnosis of type 2 diabetes is made (142). Because hyperglycemia in type 2 diabetes causes microvascular disease and may cause or contribute to macrovascular disease, undiagnosed diabetes is a serious condition. Patients with undiagnosed type 2 diabetes are at significantly increased risk for coronary heart disease, stroke, and peripheral vascular disease. In addition, they have a greater likelihood of having dyslipidemia, hypertension, and obesity (143).

Thus, early detection, and consequently early treatment, might well reduce the burden of type 2 diabetes and its complications. However, to increase the cost-effectiveness of testing undiagnosed, otherwise healthy individuals, testing should be considered in high-risk populations. Suggested criteria for testing are given in Table 6. Factors leading to these recommendations include: 1) the steep rise in the incidence of the disease after age 45 years, 2) the negligible likelihood of developing any of the complications of diabetes within a 3-year interval of a negative screening test, and 3) knowledge of the well-documented risk factors for the disease. Although the OGTT and FPG are both suitable tests, in clinical settings, the FPG is strongly recommended because it is easier and faster to perform, more convenient and acceptable to patients, more reproducible, and less expensive.

We gratefully acknowledge the invaluable assistance of Robert Misbin, MD, in the development of the manuscript; Katherine Flegal, PhD, for her analysis of the NHANES III data set; Reubin Andres, MD, for sharing unpublished data from the Baltimore Longitudinal Study of Aging; and Michael Engelgau, MD, for providing the raw data from the Egyptian Study (130).