Home / Resources / Clinical Gems / Handbook of Diabetes, 4th Edition, Excerpt #2: Diagnosis and Classification of Diabetes

Handbook of Diabetes, 4th Edition, Excerpt #2: Diagnosis and Classification of Diabetes

Jul 14, 2014
Rudy Bilous MD, FRCP
Richard Donnelly MD, PHD, FRCP, FRACP


Diabetes mellitus is diagnosed by identifying chronic hyperglycemia. The World Health Organization (WHO) and the American Diabetes Association (ADA) have used a fasting plasma glucose (FPG) of 126mg/dL.(7 mmol/L) or higher to define diabetes (Table 3.1 ). This originated from epidemiological studies in the 1990s which appeared to show that the risk of microvascular complications (e.g. retinopathy) increases sharply at a FPG threshold of 126mg/dL.(7 mmol/L) (Figure 3.1 ). Lately, however, the notion of a clear glycemic threshold separating people at high and low risk of diabetic microvascular complications has been called into question. Part of the rationale for switching to HbA1c > 6.5% (48 mmol/mol) as a diagnostic test is that moderate retinopathy, in more recent trials, is rare below this HbA1c threshold….

There are currently 23.6 million people in the USA with diabetes (7.8% of the population). The total number of people with diabetes worldwide is projected to increase from 171 million in 2000 to 366 million in 2030. A key demographic change to the rising prevalence of diabetes worldwide is an increasing proportion of people > 65 years of age. Diabetes can be diagnosed in several ways.

  • HbA 1c ≥ 6.5% of 126mg/dL. (48 mmol/mol).
  • A casual (random) plasma glucose level ≥199mg/dL.(11.1 mmol/L) in someone with typical symptoms of diabetes.
  • A fasting plasma glucose level ≥ 126mg/dL.(7.0 mmol/L)
  • A plasma glucose level ≥ 200mg/dL.(11.1 mmol/L) 2 hours after a 75 g load of glucose given by mouth (the oral glucose tolerance test – OGTT).





Intermediate categories of hyperglycemia: prediabetes

During the natural history of all forms of diabetes, the disease passes through a stage of impaired glucose tolerance (IGT), defined as a plasma glucose of 140 – 200 mg/dL (7.8 – 11.0 mmol/L) 2 hours after an OGTT. Impaired fasting glycemia (IFG) is an analogous category based on fasting glucose levels, and is defined as a FPG of 110-126mg/dL.( 6.1 – 6.9 mmol/L) (Box 3.1 ). Impaired glucose tolerance and IFG are intermediate metabolic stages between normal glucose homeostasis and diabetes.

They are both risk factors for future diabetes and cardiovascular disease, but the 2-hour plasma glucose concentration is a particularly strong predictor of cardiovascular risk and mortality (Figure 3.2 ).


A proportion ofpatients with IFG and/or IGT (5 – 10% per annum) will deteriorate metabolically into overt diabetes. Lifestyle modification (diet, exercise and weight loss) is the best approach to diabetes prevention for these patients. More recently, some genetic markers have been associated with an increased risk of progression from IGT to diabetes, e.g. common polymorphisms of the transcription factor 7-like 2 gene (TCF7L2).

For an OGTT, the subject is tested in the morning after an overnight fast, in the seated position. After taking a fasting blood sample, 75 g of glucose is given by mouth, often in the form of a glucose drink such as Lucozade (388 mL). For children, the glucose dose is calculated as 1.75 g/kg. A further blood sample is taken at 2 hours, and the fasting and 2-hour glucose values are interpreted as in Figure 3.3.

Screening for diabetes by the FPG level does not identify exactly the same population as that diagnosed by the plasma glucose 2 hours after an OGTT or by HbA1c (Table 3.2 ). For example, in the US NHANES study, 1.6% of the population had HbA1c ≥ 6.5% but 5% of these would be undiagnosed using FPG or 2h criteria. Only 55% of patients with FPG ≥ 126mg/dL.( 7mmol/L) and 2h glucose ≥ 200mg/dL.(11.1 mmol/L) had an HbA1c≥ 6.5% (Cowie et al. 2010 ).

Glycosuria (the presence of glucose in the urine) cannot be used to diagnose diabetes because of the poor relationship between blood and urine glucose. This is for several reasons: the renal threshold for glucose reabsorption varies considerably within and between individuals, the urine glucose concentration is affected by the subject’s state of hydration and the result reflects the average blood glucose during the period that urine has accumulated in the bladder. The average renal threshold is 10 mmol/L (i.e. blood glucose concentration above this level will ‘spill over’ into the urine), but a negative urine test can be associated with marked hyperglycemia.


Longer term indices of hyperglycemia include the glycated hemoglobin percentage (HbA1c), a measure of integrated blood glucose control over the preceding few weeks. HbA1c is used primarily to assess glycemic control among people with diabetes on treatment (aiming for HbA 1c 6 – 7%). HbA1c analyses are now being calibrated to the IFCC assay (rather than the NGSP DCCT HPLC assay). Thus, the units of HbA1c are changing from percent to mmol/mol (Table 3.3).





The potential value of screening for diabetes is to facilitate early diagnosis and treatment. About 20% of newly diagnosed subjects with types 2 diabetes already have evidence of vascular complications. This suggests that complications begin about 5 – 6 years before a diagnosis is made, and that the actual onset of (type 2) diabetes may occur several years before the clinical diagnosis.

In most countries, there is no systematic screening policy for diabetes, yet there are estimates that up to 50% of patients with diabetes are undiagnosed. Ad hoc screening of high-risk groups is becoming more common. The FPG is quick and cheap, but can miss those with isolated post challenge hyperglycemia. In future, HbA1c will be increasingly used for screening and diagnosis. Screening policies should target high-risk groups (Box 3.2 ).


Classification of diabetes

The current classification of diabetes is based on the aetiology of the disease (Box 3.3).

  • Type 1 diabetes (caused by pancreatic islet cell destruction).
  • Type 2 diabetes (caused by a combination of insulin resistance and β cell insulin secretory dysfunction).
  • Other specific types of diabetes (caused by conditions such as endocrinopathies, diseases of the exocrine pancreas, genetic syndromes, etc.; see below).
  • Gestational diabetes (defined as diabetes that occurs for the first time in pregnancy).

Type 1 diabetes is subdivided into two main types: 1a or autoimmune (about 90% of type 1 patients in Europe and North America, in which immune markers, such as circulating islet cell antibodies, suggest autoimmune destruction of the β-cells) and 1b or idiopathic (where there is no evidence of autoimmunity). A steady increase (2.5 – 3% per annum) in the incidence of type 1 diabetes has been reported worldwide, especially among young children < 4 years old. There are large differences between countries in the incidence of type 1 diabetes, e.g. up to 10-fold difference among European countries.

This classification has now replaced the earlier, clinical classification into ‘insulin-dependent diabetes mellitus’ (IDDM) and ‘non-insulin dependent diabetes mellitus’ (NIDDM), which was based on the need for insulin treatment at diagnosis. IDDM is broadly equivalent to type 1 diabetes and NIDDM to type 2 diabetes (Table 3.4). One of the disadvantages of the old classification according to treatment was that subjects could change their type of diabetes – for example, some type 1a patients diagnosed after the age of 40 years masquerade as NIDDM, before eventually becoming truly insulin dependent (this is now classified as latent autoimmune diabetes in adults; LADA).


Various clinical and biochemical features can be used to decide whether the patient has type 1 or type 2 diabetes (Box 3.4). The distinction may be difficult in individual cases. The category of ‘other specific types of diabetes ‘is a large group of conditions, which includes genetic defects in insulin secretion (such as in maturity-onset diabetes of the young (MODY) and insulinopathies), genetic defects in insulin action (e.g. syndromes of severe insulin resistance), pancreatitis and other exocrine disorders, hormone-secreting tumors such as acromegaly (growth hormone) and Cushing’s syndrome (cortisol). Some cases are caused by the administration of drugs such as glucocorticoids. Some genetic syndromes are sometimes associated with diabetes (e.g. Down’s syndrome, Klinefelter’s syndrome and many more).


For more information and to purchase this book, just follow this link:
Rudy Bilous MD, FRCP, Professor of Clinical Medicine, Newcastle University, Honorary Consultant Endocrinologist, South Tees Foundation Trust, Middlesbrough, UK
Richard Donnelly MD, PHD, FRCP, FRACP, Head, School of Graduate Entry Medicine and Health, University of Nottingham, Honorary Consultant Physician, Derby Hospitals NHS Foundation Trust, Derby, UK 

A John Wiley & Sons, Ltd., Publication

This edition first published 2010, © 2010 by Rudy Bilous and Richard Donnelly. Previous editions: 1992, 1999, 2004
Note: The contents of this work are intended to further general scientific research, understanding, and discussion only and are not intended and should not be relied upon as recommending or promoting a specific method, diagnosis, or treatment by physicians for any particular patient. The publisher and the authors make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of fitness for a particular purpose. In view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of information relating to the use of medicines, equipment, and devices, the reader is urged to review and evaluate the information provided in the package insert or instructions for each medicine, equipment, or device for, among other things, any changes in the instructions or indication of usage and for added warnings and precautions. Readers should consult with a specialist where appropriate. The fact that an organisation or website is referred to in this work as a citation and/or a potential source of further information does not mean that the authors or the publisher endorse the information the organisation or website may provide or recommendations it may make. Further, readers should be aware that Internet websites listed in this work may have changed or disappeared between when this work was written and when it is read. No warranty may be created or extended by any promotional statements for this work. Neither the publisher nor the authors shall be liable for any damages arising herefrom.