Chapter 1

Colin, a 51-year-old construction worker, saw the Practice Nurse at his GP surgery for a review of his hypertension. Amongst the results of his pre-arranged blood test was an HbA_1c of 67 mmol/mol.

Is this result enough to make a diagnosis of diabetes for Colin?

Chapter 2

The threshold for diagnosing diabetes is an HbA_1c of ≥48 mmol/mol (1,2). First, however, you should ask Colin if he has experienced symptoms of thirst, polyuria or weight loss. If osmotic symptoms of diabetes are present, then a single measurement above the threshold is sufficient to diagnose diabetes. If, however, Colin is asymptomatic then the test should be repeated, ideally within 2 weeks (without change of lifestyle or diet). The repeat test that is performed should be the same as the first test, in this case HbA_1c (3).

On repeat testing, an HbA_1c of 65 mmol/mol was recorded. Colin had a BMI of 29.2 kg/m² and his mother was known to have type 2 diabetes. With this information, Colin was diagnosed as having type 2 diabetes.

Since up to 50% of people with type 2 diabetes have complications at the time of diagnosis, Colin will be assessed for the presence of nephropathy, retinopathy and neuropathy, and his cardiovascular risk estimated.

Chapter 3

Rao, a 42-year-old accountant of Asian origin, on a routine visit to the GP surgery, mentioned to the Practice Nurse that there was a strong family history of type 2 diabetes. Rao was asymptomatic with regard to diabetes symptoms. The Practice Nurse arranged for a venous blood sample to be taken, and a fasting plasma glucose (FPG) level came back at 6.7 mmol/L together with an HbA_1c of 52 mmol/mol.

How would you interpret these results? Does Rao have a diagnosis of diabetes?

Chapter 4

The relevant diagnostic thresholds for diagnosing diabetes are an FPG level of ≥7.0 mmol/L or an HbA_1c of ≥48 mmol/mol or more (1,2). Thus, Rao meets the threshold criteria for HbA_1c, but not for FPG.

In this situation, where the individual is asymptomatic and one test is above the threshold but the other is not, the test that indicates diabetes (in this case HbA_1c) should be repeated. If the repeat test is above the threshold, then the diagnosis of diabetes is made. If not (i.e. HbA_1c <48 mmol/mol), then Rao would be deemed to have non-diabetic hyperglycaemia (NDH). This should be coded on his records and follow-up arranged, with repeat testing at appropriate intervals. In England, he should be referred to the Healthier You: NHS Diabetes Prevention Programme (4) for further advice and support. It is important to recognise that the diagnostic overlap of different tests for diabetes is not identical (3).

If the situation is such that two different tests have been performed (such as FPG and HbA_1c) and both are above the diagnostic threshold, then the diagnosis of diabetes is established, even if asymptomatic.

What other measurements, apart from HbA_1c and fasting plasma glucose, can be used to diagnose diabetes?

Chapter 5

Diabetes can be diagnosed by several methods, either based on plasma glucose (FPG, random plasma glucose or oral glucose-tolerance test) or glycated haemoglobin (HbA_1c) (1–3,5).

Historically, FPG and the 75-g oral glucose tolerance test (OGTT) have been used to diagnose diabetes. With improving standardisation of HbA_1c assays, the World Health Organization (WHO) recommended using HbA_1c as a more pragmatic means of diagnosing diabetes (6). NICE guidelines for type 2 diabetes have incorporated this advice and the switch from glucose-based measurements to HbA_1c as the principal means of diagnosing type 2 diabetes has been implemented in recent years (7).

The criteria for diagnosis of diabetes and NDH (impaired glucose regulation, intermediate hyperglycaemia or “prediabetes”) are shown below (1–3,5). The American Diabetes Association (ADA) criteria for NDH are slightly different than WHO and other guidelines (see below). The choice of thresholds for diagnosing diabetes has been based on the occurrence of retinopathy, a microvascular complication of diabetes (1).

Over what period of time does HbA_1c reflect raised plasma glucose levels?

Chapter 6

HbA_1c represents the proportion of haemoglobin within red blood cells that is glycated. Glucose binds to the N-terminal valine on the beta chain of haemoglobin A (in normal adults HbA predominates), with the level of glycation reflecting prevailing blood glucose levels over the lifetime of a red blood cell, which is approximately 120 days. Once the haemoglobin is glycated it remains so until the red blood cell is destroyed.

Thus, HbA_1c represents a 3-month average of plasma glucose levels. Importantly, glucose levels in the days nearer the HbA_1c test contribute significantly more than those from days further from the test. HbA_1c is, therefore, weighted to more recent plasma glucose levels (because newer red blood cells glycated more recently will better survive until the time of the test than older red blood cells).

What advantages does HbA_1c offer as a diagnostic test over other methods?

Chapter 7

An important advantage of using HbA_1c to diagnose diabetes is that a venous blood sample can be taken at any time, in contrast to FPG and OGTT, both of which require the individual to be starved and, in the case of the OGTT, detained for a prolonged period. A further benefit of using HbA_1c is that it is inherently less subject to the day-to-day variation of other diagnostic tests (1,3,5). 

When should HbA_1c not be relied on to diagnose diabetes?

Chapter 8

There are, however, circumstances where HbA_1c should not be used as a diagnostic test for diabetes (see below), notably when blood glucose levels have risen rapidly and, crucially, when type 1 diabetes is suspected. In these situations, HbA_1c will not accurately reflect current levels of glycaemia, and diagnoses of diabetes in these cases will depend on measurement of plasma glucose levels, as outlined earlier in these case studies in Chapter 5 (1,3,5).

• Symptoms suggesting type 1 diabetes.
• All symptomatic children and young adults (<18 years).
• Acute onset of diabetes symptoms.
• People who are at high risk of diabetes who are acutely unwell.
• People taking medication that can increase glucose levels (e.g. corticosteroids, antipsychotics).
• Pregnancy.

HbA_1c can also be misleading in a variety of medical conditions. A reduced red cell lifespan may lead to falsely lowered HbA_1c. This situation may arise with a haemolytic anaemia, splenomegaly or chronic liver disease, such as cirrhosis. In contrast, extended erythrocyte survival can lead to a falsely elevated HbA_1c, such as following splenectomy. HbA_1c values are generally unreliable in the setting of significant anaemias (Hb <100 g/L). In the case of haemoglobinopathies, specialist advice should be sought (1,5).

Chapter 9

Rachael, 43-years-old, complained to her GP of fatigue. Amongst her blood test results was an HbA_1c of 46 mmol/mol. She had a BMI of 28.4 kg/m² and her mother has a diagnosis of type 2 diabetes.

How would you interpret Rachael’s HbA_1c measurement?

Chapter 10

Rachael’s HbA_1c does not meet the threshold for diagnosing diabetes, but is within the range indicating NDH in the UK (42–47 mmol/mol) (2,3). In NDH, glucose levels are raised but not in the diabetic range.

NDH is a situation in which individuals do not meet the criteria for diabetes, but have results that place them at increased risk of developing type 2 diabetes. NDH ranges of HbA_1c, impaired fasting glucose and impaired glucose tolerance arising in an OGTT are shown in the table in Chapter 4 above (2,3). Slightly different criteria for NDH are recommended by the ADA: FPG 5.6–6.9 mmol/L; HbA_1c 39–47 mmol/mol (5.7–6.4%) (1).

Why is it important to identify Rachael’s non-diabetic hyperglycaemia?

Chapter 11

Rachel is at increased risk of developing type 2 diabetes. One in seven adults in the UK is estimated to have NDH and, of these, 5–12% are thought to progress to type 2 diabetes every year (8).

Cardiovascular risk in individuals with NDH is two to three times higher than that of individuals with normoglycaemia (9). The extent to which microvascular complications accumulate in NDH is uncertain, but it is notable that a quarter of patients have complications by the time they are diagnosed with diabetes (10).

HbA_1c has been found to more accurately predict risk of subsequent diabetes and cardiovascular events than fasting glucose (11). It is appropriate for those with NDH to undergo annual monitoring (including HbA_1c measurement) and to have their cardiovascular risk assessed and risk factors addressed.

Chapter 12

Misdiagnosis of diabetes is a frequent finding in primary care, and errors in classification of the type of diabetes are also common (12).

These case histories illustrate the criteria for diagnosing diabetes and NDH. One positive test is sufficient for diagnosis if the patient has osmotic symptoms (increased thirst and micturition, weight loss). Otherwise, a further test should be undertaken to confirm the diagnosis.

HbA_1c has been recommended as a pragmatic choice for diagnosing diabetes. It should not be used in situations when blood glucose levels are changing rapidly, notably in children and young people where type 1 diabetes is the concern.

NDH identifies a group of individuals at higher risk of developing type 2 diabetes and already at higher cardiovascular risk than the normoglycaemic population.

NICE has recommended a two-stage process of identifying people at high risk of diabetes (13). The first step involves risk stratification from a validated questionnaire or a computer-based assessment from primary care records. High-risk individuals from this first step then undergo a blood test for hyperglycaemia, and subsequent management then depends on whether the result is normal or indicates NDH or diabetes.

People can now perform their own initial assessment (without a blood test) using an online validated risk tool for type 2 diabetes (Know Your Risk) that can be found on the Diabetes UK website (14). They are subsequently signposted to receive further support, as necessary.