Diagnosis and monitoring

Diagnosis of diabetes

In the United Kingdom, diagnosis of diabetes follows the criteria published by the World Health Organisation (WHO) in 1999.

Because of its relatively rapid onset and the frequency of clinical symptoms, type 1 diabetes is unlikely to go undiagnosed.

• In the presence of symptoms (thirst, frequent urination, unexplained weight loss, drowsiness or - in extreme cases - coma), a glucose level in whole venous blood in the non-fasting state greater than 10.0 mmol/L is sufficient to establish a diagnosis of diabetes.
• Confirmatory testing with an oral glucose tolerance test (see below) is only required if the blood glucose level is in the uncertain zone of 4.4 - 10.0 mmol/L (whole venous blood), 4.4 - 11.1 mmol/L (whole capillary blood).
• The diagnostic criteria in children are the same as for adults.
• Severe hyperglycaemia under conditions of acute infective, traumatic, circulatory or other stress may be transitory, and is not in itself diagnostic of diabetes.

Type 2 diabetes is more likely to go undiagnosed if onset is slow and symptoms absent or minimal, when they may be missed, or attributed to another cause.

There is at present no established national screening programme for type 2 diabetes. And unless there is a suspicion motivated by family history that someone has diabetes, the diagnosis may first be suggested by the finding of hyperglycaemia during a blood test for some other purpose.

The WHO states that “The diagnosis of diabetes in an asymptomatic subject should never be made on the basis of a single abnormal blood glucose value”, and so a second, confirmatory test on a separate occasion will be needed. This is most usually a blood glucose determination done on a sample taken after overnight fasting, or the oral glucose tolerance test.

The oral glucose tolerance test

This test is carried out after at least three days of unrestricted diet and usual physical activity. After an evening meal containing 30-50g of carbohydrate, the subject fasts overnight (min. 8 hours). On the following morning, a fasting blood sample is taken and the subject consumes a drink containing 75g of glucose. A second blood sample is taken exactly two hours later.

Gestational diabetes is diagnosed using the oral glucose tolerance test, usually at 24-28 weeks after conception. Subjects meeting WHO criteria for diabetes or IGT are classified as having gestational diabetes.

Monitoring and self-management

Once diabetes (or IGT or IFG) has been diagnosed, it is important to monitor blood glucose level on an ongoing basis, since tight control of blood glucose decreases the risk of developing long-term complications in both type 1 and type 2 diabetes.

Measurement of the level of glycosylated haemoglobin (HbA1C) in the blood gives a good indication of the average blood glucose level over the previous 2-3 months. An individual target level will be set by the doctor, taking into account any known risk factors for complications. Guidelines recommend that this target level should be set in the range 6.5 to 7.5 per cent, with lower values in people with risk factors for complications. However, with certain treatments, lower values are associated with a greater risk of experiencing hypoglycaemia, and so a balance must be found. It is important that HbA1C levels are tested regularly, and current guidelines state that this should be at least twice a year and up to six times, depending on how much the person's condition varies. The test is available through a clinician, or a commercially available test kit at a retail pharmacy.

Monitoring average blood glucose levels with HbA1C is not sufficient for good management of the disease, as there may still be large short-term variations after meals, for example, that can have damaging consequences. This is particularly likely in type 1 diabetes; by contrast, levels in type 2 diabetes tend to show less variation.

Short-term monitoring is most conveniently carried out by the patient as a part of self-management, and education will be offered where self-testing is thought to be appropriate.

The simplest form of self-testing is to use glucosesensitive dipsticks to measure glucose in the urine. However, values measured in urine show a poor match with blood levels. Also, this method does not give a warning of developing hypoglycaemia. Its use is usually restricted to type 2 diabetes treated with agents that do not provoke hypoglycaemia and where the individual is not able or willing to carry out the more usual pinprick blood testing.

Most people find they can measure glucose levels in capillary blood, obtained by pricking a finger with a small lancet or needle. A small drop of blood is obtained and spotted onto a reagent strip. Originally, the test strips changed colour to indicate the glucose concentration and the result was then read either manually or by a meter. As many people with diabetes have poor eyesight, newer meters detect glucose by electrochemical means, with the results displayed on a clearly readable screen.

Many people use the finger-prick method of blood sampling successfully over many years, but a noninvasive method of testing would be preferable if it were sufficiently accurate. Various devices have been tried, but none has yet proved entirely satisfactory.

One commercially available device, worn around the wrist like a watch, uses a small electric current to draw out fluid from the skin to test its glucosecontent with a sensor. However, the glucose level in this fluid can differ significantly from that in blood, and the instrument requires careful calibration, using the finger-prick method, each time a new sensor is fitted. Hence, it is not yet a truly reliable, non-invasive monitoring alternative to blood sampling, although it may be useful as a complement.