Patients starting on insulin need to inform the driving licence authority and their insurance companies. They are also wise to inform their employers. Certain types of work are unsuitable for insulin-treated patients, including driving heavy goods or public service vehicles, working at heights, piloting an aircraft or working close to dangerous machinery in motion. Certain professions such as the police and the armed forces are barred to all diabetic patients but there are few other limitations, although a considerable amount of ill-informed prejudice still exists.
Complications
AT THE INJECTION SITE. Shallow injections result in intradermal insulin delivery and painful, reddened lesions or even scarring. Injection site abscesses occur but are extremely rare.
Local allergic responses sometimes occur early in therapy but usually resolve spontaneously. Generalized allergic responses are exceptionally rare.
Lipodystrophies that may occur include lipoatrophy, a local allergic response now virtually abolished by the use of highly purified insulins, and lipohypertrophy, occurring as a result of overuse of a single injection site with any type of insulin.
INSULIN RESISTANCE. The most common cause of mild insulin resistance is obesity. Occasional unstable patients require massive insulin doses, often with a fluctuating requirement. There are often associated behavioural problems. Insulin resistance associated with antibodies directed against the insulin receptor has been reported in patients with acanthosis nigricans.
WEIGHT GAIN. Patients who are non-compliant with their diet and predisposed to weight gain may show progressive weight gain on treatment, especially if the insulin dose is increased inappropriately.
HYPOGLYCAEMIA. This is the most common complication of insulin therapy and is a major cause of anxiety for patients and relatives. Symptoms develop when the blood glucose level is below 2.5 mrnol litre ” and typically develop over a few minutes, with most patients experiencing ‘adrenergic’ features of sweating, tremor and a pounding heart beat. Physical signs include pallor and a cold sweat. Many patients with long-standing diabetes report loss of these warning symptoms and are at a greater risk of drifting into severe hypoglycaemia. Such patients appear pale, drowsy or detached, signs that their relatives quickly learn to recognize. Behaviour is clumsy or inappropriate, and some become irritable or even aggressive. Others slip rapidly into hypoglycaemic coma. Occasionally, patients develop convulsions during hypoglycaemic coma, especially at night. It is important not to confuse this with idiopathic epilepsy, especially since patients with frequent hypoglycaemia often have abnormalities on the EEG. Another presentation is with a hemiparesis that resolves within a few minutes when glucose is administered.
Hypoglycaemia is a common problem. Virtually all patients experience intermittent symptoms and one in three will go into a coma at some stage in their lives. A minority suffer attacks that are so frequent and severe as to be virtually disabling.
Hypoglycaemia results from an imbalance between injected insulin and a patient’s normal diet, activity and basal insulin requirement. The times of greatest risk are before meals and during the night. Irregular eating habits, unusual exertion and alcohol excess may precipitate episodes; others appear to be due simply to variation in insulin absorption.
A further problem is that diabetic patients have an impaired ability to counter-regulate glucose levels after hypoglycaemia. The glucagon response is invariably deficient, even though the a cells are preserved and respond normally to other stimuli. The adrenaline response may also fail in patients with a long duration of diabetes.
Nocturnal hypoglycaemia is commonly caused by attempts to compensate for the slight increase in insulin requirements from 4 a.m. (the ‘dawn phenomenon’). This is related to the nocturnal peak of growth hormone secretion. Since injected insulin inevitably peaks and declines, increasing the evening dose of insulin to combat fasting hyperglycaemia increases the risk of hypoglycaemia in the early hours of the morning. It was widely believed that this hypoglycaemia caused a rebound hyperglycaemia (the ‘Somogyi effect’) owing to an unbalanced counter-regulatory response, but in practice fasting hyperglycaemia is usually due to insulin deficiency.
The diagnosis of hypoglycaemia is simple and can usually be made on clinical grounds. Patients should carry a card or wear a bracelet or necklace identifying themselves as diabetic, and these should be looked for in unconscious patients. If real doubt exists, it will do no harm to administer glucose whilst a laboratory blood glucose result is awaited.
Any form of rapidly absorbed carbohydrate will relieve the early symptoms, and patients should always carry glucose or sweets. Drowsy patients will often be able to take carbohydrate in liquid form, e.g. a spoonful of sugar in water. Milk should be avoided since fat delays gastric emptying and slows recovery. Unconscious patients should be given intravenous glucose (50 ml of 50% extrose solution) followed by a flush of normal saline to preserve the vein, or intramuscular glucagon (1 mg). Glucagon acts by mobilizing hepatic glycogen, and works almost as rapidly as glucose. It is simple to administer and can be given at home by relatives. Glycogen reserves should be replenished with oral glucose once the patient revives.
Measuring control
The ‘artificial pancreas’ is a system of blood glucose control that works by continuous blood glucose analysis. This is fed into a computer, which delivers an appropriate amount of insulin into the circulation. Patients on insulin need to devise their own simplified form of this feedback loop.
Urine tests
Urine tests (dipstix) are simple to perform, and it can usually be assumed that a patient with consistently negative tests and no symptoms of hypoglycaemia is well controlled. Even so, the correlation between urine tests and simultaneous blood glucose is poor for three reasons:
1 Changes in urine glucose lag behind changes in blood glucose.
2 The mean renal threshold is around 10 mmol litre-‘ but the range is wide (7-13 rnrnol litre “). The threshold also rises with age.
3 Urine tests can give no guidance concerning blood glucose levels below the renal threshold. Urinary ketones may also be measured by a dipstick test. This is rarely helpful in routine outpatient management, but can be useful in special situations such as intercurrent infections. Heavy ketonuria can inhibit some dipstick tests for glucose.
Blood glucose testing
This provides the best assessment of day-to-day control.The fasting blood glucose concentration is a useful guide to therapy in NIDDM.
A random blood glucose test (e.g. in the clinic) is of limited value, but patients may easily be taught to provide their own profiles by testing finger-prick blood samples with reagent strips and reading these with the aid of a visual scale or reflectance meter. It has been amply demonstrated that most patients are willing and able to provide reasonably accurate results provided they have been properly taught.
Blood is taken from the side of a fingertip (not from the tip, which is densely innervated) using a special lancet, e.g. Monolet, which can be fitted to a spring-loaded device. Patients are asked to take regular profiles (e.g. four daily samples on 2 days each week) and to note these in a diary or record book. Home blood glucose monitoring is essential for good diabetic control. Patients are encouraged to adjust their insulin dose as appropriate and should ideally be able to obtain advice over the telephone when needed.
Glycosylated haemoglobin (HbA, or HbA,J
Glycosylation of haemoglobin occurs as a two-step reaction, resulting in the formation of a covalent bond between the glucose molecule and the terminal valine of the f3-chain of the haemoglobin molecule. The rate at which this reaction occurs is related to the prevailing glucose concentration. Glycosylated haemoglobin is expressed as a percentage of the normal haemoglobin (normal range approximately 4-8% depending on technique of measurement). This test provides an index of the average blood glucose concentration over the life of the haemoglobin molecule (approximately 6 weeks). The figure will be misleading if the life-span of the red cell is reduced or if an abnormal haemoglobin or thalassaemia is present. Although the glycosylated haemoglobin test provides a rapid assessment of the level of glycaemic control in a given patient, blood glucose testing is needed before the clinician can know what to do about it. Glycosylated plasma proteins (,fructosamine’) may also be measured as an index of control. Glycosylated albumin is the major component and fructosamine measurement relates to glycaemic control over the preceding 1-3 weeks. The technique is cheaper and quicker than glycosylated haemoglobin measurement and lends itself to automation. It is useful in patients with haemoglobinopathy and in pregnancy (when haemoglobin turnover is changeable). Correlation between the two tests is weak. This may reflect the greater interindividual variation in plasma proteins than in haemoglobin. It is also less reliable and measurement of Hbx., is often preferred. Does good glycaemic control matter?
The answer to this question involves a number of separate issues.
Is poor control associated with an increased risk of microvascular complications? Retrospective studies have repeatedly shown that those with the worst control have the highest rate of complications.
2 Is this increased risk reversible? There are many difficulties in answering this question:
(a) The gestation of diabetic complications is lengthy, often 10-20 years.
(b) Control was difficult to quantify before HbA,c tests were introduced.
(c) Good control is difficult to achieve, so comparisons have usually been between ‘poor and worse’ control rather than between ‘good and bad’.
(d) Some patients are easier to bring into good control than others, so the groups selected as ‘well’ or ‘poorly’ controlled in previous studies were not strictly comparable.
Studies in experimental animals strongly suggest that improved control is protective and this has now been confirmed in humans. The Diabetes Control and Complications Trial (DCCT) in the USA compared standard versus intensive insulin therapy in a prospective
controlled trial of young patients with IDDM. Even on intensive therapy, mean blood glucose levels were 40% above the non-diabetic range, but this level of control reduced the risk of progression to retinopathy, nephropathy or neuropathy over the 7 years of the study by some 60%. Near-normoglycaemia should, therefore, be the goal for all young patients with IDDM. Unwanted effects of this policy include weight gain and a 2-3 fold increase in the risk of severe hypoglycaemia. Control should be less strict in those with a history of recurrent severe hypoglycaemia. It remains unclear whether equally stringent standards should be applied in patients with NIDDM, particularly since a protective effect upon progression of macrovascular disease has yet to be demonstrated. A large trial in patients with NIDDM is due to report shortly and should allow this question to be answered.
3 Can established complications be halted or reversed by intensive insulin therapy? Insulin infusion devices have made near-normal blood glucose control possible for closely supervised groups of patients. Studies in patients with established retinopathy or nephropathy have shown that patients with early retinopathy benefit from 2-3 years of intensive therapy, but that patients with more advanced retinal changes or proteinuria do not. Retinopathy may show a transient deterioration when strict control is first established. These observations suggest that microvascular lesions may be selfperpetuating once a threshold level of damage has
been reached.
4 Is macrovascular disease influenced by control? Patients with impaired glucose tolerance have an increased rate of large vessel disease but rarely develop microvascular lesions. This might be because large arteries are more sensitive to elevated glucose levels, but it has also been suggested that hyperinsulinaemia (present in many patients with NIDDM and a common consequence of insulin treatment) is a cause of accelerated atherogenesis.
At present there is little evidence that good glycaemic control protects against arterial disease.
