Criteria for diagnosis

The current World Health Organization criteria for diagnosing diabetes mellitus are shown in Table 32.1. The figures shown apply to the concentrations found in venous plasma; slightly different figures (not shown) apply to whole blood or capillary samples. Glucose is routinely measured in blood specimens that have been collected into tubes containing fluoride, an inhibitor of glycolysis. Because of the need sometimes to obtain rapid blood glucose results and the widespread self-monitoring of diabetic patients, blood glucose may also be assessed outside the laboratory using devices such as those shown in Figure 32.1. The fasting and 2-hour criteria define similar levels of glycaemia above which the risk of diabetic complications increases substantially.

Fasting blood glucose

A fasting blood glucose concentration of ≥7.0 mmol/L is regarded as diagnostic of diabetes, whether or not hyperglycaemic symptoms are present. The patient should be fasted overnight (at least 10 hours). If the result falls between 6.0 and 6.9 mmol/L, the patient is said to have ‘impaired fasting glycaemia’ (see below). Interpretation of fasting glucose results is shown in Table 32.1.

Other glucose measurements

Two other tests are still widely used, although their role in the diagnosis of diabetes is increasingly peripheral. The first is random blood glucose measurement, often performed opportunistically, particularly if a patient has hyperglycaemic (osmotic) symptoms like thirst, frequency and polyuria. A result of ≥11.1 mmol/L requires confirmation with a fasting glucose. The second is the oral glucose tolerance test (Fig 32.2). This was often done where diagnostic confusion still existed despite repeat glucose measurement, e.g. discrepant results from fasting and random blood glucose measurement. As the name suggests, the patient consumes an oral glucose load, with blood glucose measured (fasting) at the beginning of the test, and two hours later. The oral glucose tolerance test is difficult to standardise, and the correct procedure is often not followed. It is no longer widely advocated in the diagnosis of diabetes. However, the oral glucose tolerance test criteria for diagnosis of diabetes and of impaired glucose tolerance (see below) are included in Table 32.1 for completeness.

Impaired glucose tolerance and impaired fasting glycaemia

Impaired fasting glycaemia (IFG) and impaired glucose tolerance (IGT) are intermediate categories of glycaemia that fall short of the diagnosis of diabetes, but which define an increased risk of developing diabetes. IGT can only be diagnosed after an oral glucose tolerance test. The risks associated with IGT (e.g. risk of developing diabetes) are well-defined and have been characterized over many years. By contrast, IFG is diagnosed from a single fasting sample. Its existence as a diagnostic category arose out of the 1997 recommendation of the American Diabetes Association that oral glucose tolerance tests be abandoned in favour exclusively of fasting samples (this recommendation was not adopted by the World Health Organization when it updated its 1985 diagnostic criteria in 1999). The criteria for IFG are arbitrary and the associated risks less well-defined than for IGT.

Glycated haemoglobin

Hyperglycaemia leads to the non-enzymatic attachment of glucose to a variety of proteins (glycation), which is virtually irreversible under physiological conditions and the concentration of glycated protein is therefore a reflection of mean blood glucose level during the life of that protein.

Glycated haemoglobin (HbA_1c) reflects the mean glycaemia over 2 months prior to its measurement, the half-life of haemoglobin. The HbA_1c concentration is expressed as mmol glycated haemoglobin per mol total haemoglobin (mmol/mol). Although there is disagreement about the exact optimal concentration, it is probably somewhere in the region of 50 mmol/mol. HbA_1c is widely used in diabetic clinics and primary care to complement the information from single or even serial blood glucose measurements. Spurious results may sometimes be obtained in patients with inherited structurally abnormal haemoglobins (haemoglobinopathies).

Self-monitoring

Most diabetic patients are taught how to monitor their own diabetic control. Home urine testing is the simplest form of monitoring and may be acceptable in patients with mild (early) Type 2 diabetes. If more accurate monitoring is required, patients may be supplied with a point-of-care glucose monitor (Fig. 32.1). This requires finger-prick whole blood samples (a drop is usually adequate); the patient is usually advised to vary the time of self-testing in order to get an overall picture of their glycaemic control.

Glycosuria

The presence or absence of glycosuria has no role in the screening or the diagnosis of diabetes. For historical reasons many occupational and other health checks include urine glucose testing, which may reveal unsuspected diabetes. However, the false-negative rate for this test is unacceptably high.

Ketones in urine or blood

The term ‘ketone bodies’ refers to acetone and the keto-acids acetoacetate and β-hydroxybutyrate. These are frequently found in uncontrolled diabetes (diabetic ketoacidosis or DKA – see p. 66). They are also found in normal subjects as a result of starvation or fasting, and sometimes in alcoholic patients with poor dietary intake (alcoholic ketoacidosis).