Can we retire the oral glucose tolerance test?

The use of HbA_1c as a diagnostic test for diabetes was recommended by the World Health Organization (WHO, 2011) early last year, and after much delay, consensus guidance is now available (John et al, 2011). This should make identification of type 2 diabetes easier, as the test does not require fasting and can be taken at any time of day. 

The advantage of early diagnosis of diabetes is that individuals are able to access treatment and education before they develop complications. These individuals are easier to treat, often managing just with lifestyle changes for a significant period of time. It will also have a favourable effect on attainment of QOF indicators for HbA_1c as the practice will have a greater proportion of people with diabetes who can achieve tight targets safely.

Can we, therefore, now throw away the cumbersome, time-consuming, oral glucose tolerance test (OGTT)? Although the use of HbA_1c is likely to become the main method of diagnosis, there are a number of situations in which it is not recommended and the traditional methods should be used, including OGTT (Table 1).

Despite its approval for use in the UK, HbA_1c as a diagnostic tool has not been agreed with all laboratories so some practices will not be able to use this method to diagnose diabetes if their local laboratory facilities are not offering this service. Also, point-of-care HbA_1c testing is not recommended unless it meets stringent quality assurance requirements. As HbA_1c is a reflection of the previous 12 weeks’ blood glucose levels, it should not be used to diagnose diabetes with acute onset of symptoms (suggesting type 1 diabetes). This is because it may not have reached the diagnostic cut point (≥48 mmol/mol [≥6.5%]) even though the individual has high blood glucose levels. Any condition that includes abnormalities of haemoglobin, severe anaemia (<8 g/dL) or alteration in red blood cell turnover will mean HbA_1c as a diagnostic tool will be unreliable. It should also not be used to diagnose gestational diabetes. 

Where HbA_1c is not recommended, testing for diabetes should be performed using either a fasting or random blood glucose using a venous sample (Table 2). A fasting blood glucose between 6.1 and 6.9 mmol/L should lead to a 75 g oral glucose challenge to confirm or exclude diabetes. Gestational diabetes should only be diagnosed using an OGTT (NICE, 2008). Table 3 looks at how to interpret OGTT results.

The WHO has not recommended the use of HbA_1c in identifying people at high risk of diabetes (often termed “pre-diabetes”) although an HbA_1c of between 42 and 47 mmol/mol (6.0 and 6.4%) may suggest this. Formal identification of impaired glucose tolerance will still require an OGTT. In the increasingly litigious NHS, there may still be a place for the OGTT as the gold standard in the formal diagnosis of diabetes.