Category Archives: Diameties Mellitus and Other Desorder of Metabolism


Hypoglycaemia develops when hepatic glucose output falls below the rate of glucose uptake by peripheral tissues. Hepatic glucose output may be reduced by:
• The inhibition of hepatic glycogenolysis and gluconeogenesis by insulin
• Depletion of hepatic glycogen reserves by malnutrition, fasting, exercise or advanced liver disease
• Impaired gluconeogenesis (e.g. following alcohol ingestion)
In the first of these categories, insulin levels are raised, the liver contains adequate glycogen stores, and the hypoglycaemia can be reversed by injection of glucagon. In the other two situations, insulin levels are low and glucagon is ineffective.
Peripheral glucose uptake is accelerated by high insulin levels and by exercise, but these conditions are normally balanced by increased glucose output. Insulin or sulphonylurea therapy for diabetes accounts for the vast majority of cases of severe hypoglycaemia encountered in an accident and emergency department.
The commonest symptoms and signs of hypoglycaemia are neurological. The brain consumes about 50% of the total glucose produced by the liver. This high energy  requirement is needed to generate ATP used to maintain the potential difference across axonal membranes.


Insulinomas are pancreatic islet cell tumours that secreteinsulin. Most are sporadic but some patients have multiple  tumours arising from neural crest tissue (multiple endocrine neoplasia). Some 95% of these tumours are benign. The classic presentation is with fasting hypoglycaemia, but early symptoms may also develop in the late morning or afternoon. Recurrent hypoglycaemia is often present for months or years before the diagnosis is made,
and the symptoms may be atypical or even bizarre; the presenting features in one series are given. Common misdiagnoses include psychiatric disorders, particularly pseudo dementia in elderly people, epilepsy and cerebrovascular disease.


Whipple’s triad remains the basis of clinical diagnosis. This is satisfied when:
1 Symptoms are associated with fasting or exercise
2 Hypoglycaemia is confirmed during these episodes
3 Glucose relieves the symptoms
A fourth criterion – demonstration of inappropriately high insulin levels during hypoglycaemia-may usefully be added to these.
In practice, the diagnosis is confirmed by the demonstration of hypoglycaemia in association with inappropriate and excessive insulin secretion. Hypoglycaemia is demonstrated by:
MEASUREMENT OF OVERNIGHT FASTING (16 HOURS) GLUCOSE AND INSULIN LEVELS ON THREE OCCASIONS. About 90% of patients with insulinomas will have low glucose and non-suppressed (normal or elevated) insulin levels.
PERFORMING A PROLONGED 72 HOUR SUPERVISED FAST if overnight testing is inconclusive and symptoms persist.
Autonomous insulin secretion is demonstrated by lack of the normal feedback suppression during hypoglycaemia. This may be shown by measuring insulin, C-peptide or proinsulin during a spontaneous episode of hypoglycaemia.  Many patients also have an abnormal (diabetic) glucose tolerance test, but this has no diagnostic value.


Sweating, palpitations, weakness
Confusion or abnormal behaviour
Loss of consciousness
Grand mal seizures


The most effective therapy is surgical excrsion of the tumour. Insulinomas are often very small and difficult to localize. Highly selective angiography and contrastenhanced CT scanning are used in the first instance. Highly selective angiography is very operator dependent and referral to a skilled operator is warranted if initial attempts at localization fail. ‘Blind’ laparotomy runs the risk that the surgeon may be unable to find the insulinoma and resorts to partial pancreatectomy, with unpleasant consequences if the tumour remains in the unexcised pancreas. Localization is also possible using a rapid insulin assay on blood sampled at different levels from the pancreatic vein.
Medical treatment with diazoxide is useful when the insulinoma is malignant, in patients in whom a tumour cannot be located, and in elderly patients with mild  symptoms. Symptoms may remit on treatment with the somatostatin analogue octreotide.

Hypoglycaemia with other tumours

Hypoglycaemia may develop in the course of advanced neoplasia and cachexia, and has been described in association with many tumour types. Certain massive turnours, especially sarcomas, may produce hypoglycaemia due to secretion of insulin-like growth factor 1. True ectopic insulin secretion is extremely rare.

Postprandial hypoglycaemia

If frequent venous blood glucose samples are taken following a prolonged glucose tolerance test, about one in four subjects will have at least one value below 3 mmollitre-1
• The arteriovenous glucose difference is quite marked during this phase, so that very few are truly hypoglycaemic in terms of arterial (or capillary) blood  glucose content. Failure to appreciate this simple fact led some authorities to believe that postprandial (or reactive) hypoglycaernia was a potential ‘organic’ explanation for a variety of complaints that might otherwise have been considered psychosomatic. An epidemic of false ‘hypoglycaernia’ followed, particularly in the USA. Later work showed a poor correlation between symptoms and biochemical hypoglycaemia. Even so, a number of otherwise normal people occasionally become pale, weak and sweaty at times when meals are due, and report benefit from advice to take regular snacks between meals. True postprandial hypoglycaemia may develop in the presence of alcohol, which ‘primes’ the {3cells to produce an exaggerated insulin response to carbohydrate. The person who substitutes alcoholic beverages for lunch is particularly at risk. Postprandial hypoglycaemia sometimes
occurs after gastric surgery, owing to rapid gastric emptying and mismatching of food and insulin. This is referred to as ‘dumping’ but it is now rarely encountered.
Hepatic and renal causes of hypog Iycaemia The liver can maintain a normal glucose output despite extensive damage, and hepatic hypoglycaemia is uncommon. It is particularly a problem with fulminant hepatic failure.
The kidney has a subsidiary role in glucose production (via gluconeogenesis in the renal cortex), and hypoglycaemia is sometimes a problem in terminal renal failure. Hereditary fructose intolerance occurs in 1 in 20000 live births and can cause hypoglycaemia.Endocrine causes of hypoglycaemia.

Endocrine disorders resulting in deficiencies of hormonesantagonistic to insulin are rare but well-recognized causes
of hypoglycaemia. These include hypopituitarism, isolated adrenocorticotrophic hormone (ACT H) deficiency and Addison’s disease.

Drug-induced hypoglycaemia

Many drugs have been reported to produce isolated cases of hypoglycaemia, but usually only when other predisposing factors are present. The following are among the more important:
SULPHONYLUREAS may be used in the treatment of diabetes or may be taken by non-diabetics in suicide attempts.
QUININE may produce severe hypoglycaemia in the course of treatment for faJciparum malaria. SALICYLATES may cause hypoglycaemia following accidental ingestion by children, but this complication is very rare in adults.
PROPRANOLOL has been reported to induce hypoglycaemia in the presence of strenuous exercise or starvation.
PENTAMIDINE may cause hypoglycaemia when used in the treatment of resistant Pneumocystis pneumonia in people with AIDS.

Alcohol-induced hypoglycaemia

Alcohol inhibits gluconeogenesis. Alcohol-induced hypoglycaemia was first described in poorly nourished chronic alcoholics but may also present in binge drinkers and in children who have taken relatively small amounts of alcohol, since they have a diminished hepatic glycogen reserve. The clinical presentation is with coma and hypothermia. Factitious hypoglycaemia This is a relatively common variant of self-induced disease and is much more common than an insulinoma. Hypoglycaemia is produced by surreptitious self-administration of insulin or sulphonylureas. Many patients in this category have been extensively investigated for an insulinoma. Measurement of C-peptide levels during hypoglycaemia should identify patients who are injecting insulin; sulphonylurea abuse can be detected by chromatography of plasma or urine.


There is no evidence that diabetic patients with good glycaemic control are more prone to infection than normal subjects. However, poorly controlled diabetes entails increased susceptibility to the following infections:

1 Skin
(a) Staphylococcal infections (boils, abscesses, carbuncles)
(b) Mucocutaneous candidiasis
2 Urinary tract
(a) Urinary tract infections (in women)
(b) Pyelonephritis
(c) Perinephric abscess
3 Lungs
(a) Staphylococcal and pneumococcal pneumonia
(b) Gram-negative bacterial pneumonia
(c) Tuberculosis
One reason why poor control lends to infection is that chemotaxis and phagocytosis by polymorphonuclear leucocytes is impaired at high blood glucose concentrations. Conversely, infections may lead to loss of glycaemic control, and are a common cause of ketoacidosis. Insulintreated patients need to increase their dose by up to 25% in the face of infection, and non-insulin-treated patients may need insulin cover while the infection lasts. Patients should be told never to omit their insulin dose, even if they are nauseated and unable to eat; instead they should test their blood glucose frequently and seek urgent medical advice.

Distinguishing features between ischaemia and neuropathy in the diabetic foot.

Distinguishing features between ischaemia and
neuropathy in the diabetic foot.

Skin and joints

Joint contractures in the hands are a common consequence of childhood diabetes. The sign may be demonstrated by asking the patient to join the hands as if in prayer; the metacarpophalangeal and interphalangeal joints cannot be apposed. Thickened, waxy skin can be noted on the backs of the fingers. These features may be due to glycosylation of collagen and are not progressive. The condition is sometimes referred to as diabetic cheiroarthropathy.
Osteopenia in the extremities is also described III IDDM but rarely leads to clinical consequences.



Smooth control of diabetes minimizes the risk of infection and balances the catabolic response to anaesthesia and surgery. The procedure for insulin-treated patients is simple:
1 Long acting and/or intermediate insulin should be stopped the day before surgery with soluble insulin substituted.
2 Whenever possible, diabetic patients should be first on the morning theatre list.
3 An infusion of glucose, insulin and potassium is given during surgery. The insulin can be injected into the glucose solution or administered by syringe pump. A standard combination is 16 U of soluble insulin with 10 mmol KCl in 500 ml of 10% dextrose, infused at 100 rnl hour-I.
4 Postoperatively, the infusion is maintained until the patient is able to eat. Other fluids needed in the perioperative period must be given through a separate intravenous line and must not interrupt the glucose/insulin/potassium infusion. Glucose levels are checked every 2-4 hours and potassium levels are monitored. The amount of insulin and potassium in each infusion bag is adjusted either upwards or downwards according to the results of regular monitoring of the blood glucose and serum potassium concentrations. The same approach is used in the emergency situation, with the exception that a separate variable rate insulin infusion may be needed to bring blood glucose under control before surgery.
Non-insulin-treated patients should stop medication 2 days before the operation. Patients with mild hyperglycaemia (fasting blood glucose below 8 mmol litre””) can be treated as non-diabetic. Those with higher levels are treated with soluble insulin prior to surgery, and with glucose, insulin and potassium during and after the procedure, as for insulin-treated patients.

Pregnancy and diabetes

Modern management has transformed the outcome of pregnancy in women with diabetes. Thirty years ago one  pregnancy in three ended with the death of the fetus or neonate. Today, the results in specialized centres approach those of non-diabetic pregnancy. This improvement is due to meticulous glycaemic control and careful medical and obstetric management. When the pregnancy is planned, optimal glycaemic control is sought prior to conception.

Glycaemic control in pregnancy

The patient should perform daily home blood glucose profiles; the renal threshold falls in pregnancy and urinetests are therefore of little or no value. Insulin requirements rise, and intensified insulin regimens may become necessary. The aim is to maintain blood glucose and Hb Ale or fructosamine levels within the normal range.
General management The patient is seen at intervals of 2 weeks or less at a clinic managed jointly by obstetrician and physician. Circumstances permitting, the aim should be outpatient management with a spontaneous vaginal delivery at term. Retinopathy and nephropathy may deteriorate during pregnancy. Expert fundoscopy and urine testing for protein should be undertaken at booking, 28 weeks and before delivery.

Obstetric problems associated with diabetes

Poorly controlled diabetes is associated with macrosomia, hydrarnnios, pre-eclampsia and intrauterine death. Ketoacidosis in pregnancy carries a 50% fetal mortality,but maternal hypoglycaemia is relatively well tolerated.  Neonatal problems Maternal diabetes, especially when poorly controlled, is associated with fetal macrosomia. The infant of a diabetic mother is more susceptible to hyaline membrane disease than non-diabetic infants of similar maturity. In addition, neonatal hypoglycaemia may occur. The mechanism is as follows: maternal glucose crosses the placenta, but insulin does not; the fetal islets hypersecrete to combat maternal hyperglycaemia, and a rebound to hypoglycaemic levels occurs when the umbilical cord is severed. These complications are due to hyperglycaemia in the third trimester. Poor glycaemic control in the first trimester carries an increased risk of congenital malformations, particularly of the cardiovascular and central nervous systems.

Gestational diabetes

This term refers to glucose intolerance that develops in the course of pregnancy and remits following delivery.The condition is typically asymptomatic and is demonstrated biochemically on the basis of random testing in  each trimester and by oral glucose tolerance testing if the plasma glucose concentration is 7 mmol litre ” or more.
Since the renal threshold for glucose falls during normal pregnancy and glucose tolerance deteriorates, the condition may easily be misdiagnosed.
Treatment is with diet in the first instance, but most patients require insulin cover during the pregnancy. Insulin does not cross the placenta. Oral agents are avoided because of the potential risk to the fetus as they do cross the placenta.
Gestational diabetes has been associated with a higher frequency of obstetric problems, fetal macrosomia and neonatal hypoglycaemia. It is likely to recur in subsequent pregnancies. Gestational diabetes is often the harbinger of NIDDM in later life.
Not all diabetes presenting in pregnancy is gestational. True IDDM may develop, and swift diagnosis is essential to prevent the development of ketoacidosis. Hospital admission is required if the patient is symptomatic, or has ketonuria or a markedly elevated blood glucose level.

Brittle diabetes

There is no precise definition for this term, which is used to describe patients with recurrent ketoacidosis and/or recurrent hypoglycaemic coma. Of these, the largest group is made up of those who experience recurrent severe hypoglycaemia.

Recurrent severe hypoglycaemia

This affects 1-3% of insulin-dependent patients. Most are adults who have had diabetes for more than 10 years. By this stage endogenous insulin secretion is negligible in the great majority of patients. Pancreatic a cells are still present in undiminished numbers, but the glucagon response to hypoglycaemia is virtually absent. Long-term patients are thus subject to fluctuating hyperinsulinaemia due to erratic absorption of insulin from injection sites, and lack a major component of the hormonal defence against hypoglycaemia. In this situation adrenaline secretion becomes vital, but this too may become impaired in the course of diabetes. Loss of adrenaline secretion has been attributed to autonomic neuropathy, but this is unlikely to be the sole cause; central adaptation to recurrent hypoglycaemia may also be a factor.

The following factors may also predispose to recurrent hypoglycaemia:
OVERTREATMENT WITH INSULIN. Frequent biochemical hypoglycaemia lowers the glucose level at which symptoms develop. Symptoms often reappear when overall glucose control is relaxed.
AN UNRECOGNIZED LOW RENAL THRESHOLD FOR GLUCOSE. Attempts to render the urine sugar-free will inevitably produce hypo glycaemia.
EXCESSIVE INSULIN DOSES. A common error is to increase the dose when a patient needs more frequent injections to overcome a problem of timing.
ENDOCRINE CAUSES. These include pituitary insufficiency, adrenal insufficiency and premenstrual insulin sensitivity.
ALIMENTARY CAUSES. These include exocrine pancreatic failure and diabetic gastroparesis.
RENAL FAILURE. The kidneys are important sites for the clearance of insulin which tends to accumulate if renal function is lost.
PATIENT CAUSES. Patients may be unintelligent, uncooperative or may manipulate their therapy.

Recurrent ketoacidosis

This usually occurs in adolescents or young adults, and the most severe form is more common in girls. Although metabolic decompensation may develop very rapidly, it is often impossible to pin-point an underlying abnormality. Many theories exist concerning the causes of this condition, but all agree that it is heterogeneous. The following categories have been suggested:
IATROGENIC. Inappropriate insulin combinations may be a cause of swinging glycaemic control. For example, a once-daily regimen may cause hypo glycaemia during the afternoon or evening and pre-breakfast hyperglycaemia due to insulin deficiency.
INTERCURRENT ILLNES. Unsuspected infections, including urinary tract infections and tuberculosis, may be present. Thyrotoxicosis can also manifest as unstable glycaemic control.
PSYCHOSOCIAL CAUSES. These certainly form the largest category. It has been suggested that neuroendocrine mechanisms such as catecholamine secretion might
mediate the metabolic disturbance. Other patients undoubtedly manipulate their illness, whether consciously or unconsciously.
UNKNOWN AETIOLOGY. The most ‘brittle’ patients of all are usually female, aged 15-25 years and often overweight, and typically suffer from amenorrhoea. The insulin requirement is variable but is often high. Sophisticated ‘cheating’ has been detected in some of these patients, but this should never be assumed without convincing proof.

The diabetic kidney

The kidney may be damaged by diabetes in three main ways:
1 Glomerular damage
2 Ischaemia due to hypertrophy of afferent and efferent arterioles
3 Ascending infection

Diabetic glomerulosclerosis

Clinical nephropathy secondary to glomerular disease usually manifests 15-25 years after diagnosis and affects 30-40% of patients diagnosed under the age of 30 years. It is the leading cause of premature death in young diabetic patients. Older patients may also develop nephropathy, but the proportion affected is much smaller. The earliest functional abnormality in the diabetic kidney is renal hypertrophy associated with a raised glomerular filtration rate; this appears soon after diagnosis and is related to poor glycaemic control. The initial structural lesion in the glomerulus is thickening of the basement membrane. Associated changes may result in disruption of the protein cross-linkages that make the membrane an effective filter. In consequence, a progressive leak of protein into the urine occurs. The earliest evidence of this is ‘micro albuminuria’ (i.e. amounts of urinary albumin so small as to be undetectable by dipsticks), which in turn may, after some years, progress to intermittent albuminuria followed by persistent proteinuria. Light-microscopic changes of glomerulosclerosis become manifest; both diffuse and nodular glomerulosclerosis can occur. The latter is sometimes known as the Kimmelstiel-Wilson lesion. At a later stage still, the glomerulus is replaced by hyaline material. At the stage of persistent proteinuria, the plasma creatinine is normal but the average patient is only some 8-10 years from end-stage renal failure. The proteinuria may become so heavy as to induce a transient nephrotic syndrome, with peripheral oedema and hypoalbuminaernia. Patients with nephropathy typically show a normochromic normocytic anaemia and a raised erythrocyte sedimentation rate (ESR). Hypertension is a common development and may itself damage the kidney still further. A rise in plasma creatinine is a late feature that progresses inevitably to renal failure, although the rate of progression may vary widely between individuals. The natural history of this process. A curious feature is that almost all patients with long-established diabetes have abnormalities on renal biopsy but only a proportion develop the features of a progressive renal disease.

Schematic representation of the natural history of nephropathy.

Schematic representation of the natural history of nephropathy.

Ischaemic lesions

Arteriolar lesions, with hypertrophy and hyalinization of the vessels, affect both afferent and efferent arterioles. The appearances are similar to those of hypertensive disease but are not necessarily related to the blood pressure in patients with diabetes.

Infective lesions

Urinary tract infections are more common in women but not men with diabetes. Ascending infection may occur because of bladder stasis due to autonomic neuropathy, and infections more easily become established in damaged renal tissue. Autopsy material frequently reveals interstitial changes suggestive of infection, but ischaemia may produce similar changes and the true frequency of pyelonephritis in diabetes remains uncertain. Untreated infections in diabetics can result in renal papillary necrosis, in which renal papillae are shed in the urine, but this complication is rare.
Diagnosis and management of diabetic nephropathy The urine of all diabetic patients should be checked regularly for the presence of protein. Many centres also screen for microalbuminuria since there is some evidence that  meticulous glycaemic control or early antihypertensive treatment at this stage may delay the onset of frank proteinuria. Once proteinuria is present, other possible causes for this should be considered (see below), but once these are excluded, a presumptive diagnosis of diabetic nephropathy can be made. For practical purposes this implies inevitable progression to end-stage renal failure, although the time course can be very markedly slowed by early aggressive antihypertensive therapy. Clinical suspicion may be provoked by an atypical history, the absence of diabetic retinopathy (usually but not invariably present with diabetic nephropathy) and the presence of haematuria. Renal biopsy should be considered in such cases, but in practice is rarely necessary or helpful. The risk of intravenous urography is increased in diabetes, especially if patients are allowed to become dehydrated prior to the procedure, and a renal ultrasound is preferable. Other investigations include repeated microscopy and culture of the urine, 24-hour urine collections to quantify protein loss and to measure creatinine clearance, and regular measurement of the plasma creatinine level. Management of diabetic nephropathy is similar to that of other causes of renal failure, with the following
AGGRESSIVE TREATMENT OF BLOOD PRESSURE with a target below 140190 mmHg has been shown to slow the rate of deterioration of renal failure considerably. Angiotensin-converting enzyme inhibitors are the  drugs of choice.
ORAL HYPOGLYCAEMIC AGENTS partially excreted via the kidney (e.g. chlorpropamide) must be avoided.
INSULIN SENSITIVITY INCREASES and drastic reductions in dosage may be needed.
DIABETIC RETINOPATHY tends to progress rapidly and frequent ophthalmic supervision is essential. Management of end-stage disease is made more difficult by the fact that patients often have other complications of diabetes such as blindness, autonomic neuropathy or peripheral vascular disease. Vascular shunts tend to calcify rapidly and hence chronic ambulatory peritoneal dialysis may be preferable to haemodialysis. The failure rate of renal transplants is somewhat higher than in non-diabetic patients. A segmental pancreatic graft is sometimes performed at the same time as a renal graft. Although pancreatic grafts have a limited viability, owing to progressive fibrosis within the graft, they may give the patient a year or so of freedom from insulin injections. Diabetic nephropathy is becoming less common as diabetic care improves.

The neuropathic man.

The neuropathic man.

Diabetic neuropathy

Diabetes can damage peripheral nervous tissue in a number of ways. The vascular hypothesis postulates occlusion of the vasa nervorum as the prime cause. This seems likely in isolated mononeuropathies but the diffuse symmetrical nature of the common forms of neuropathy implies a metabolic cause. Since hyperglycaemia leads to  increased formation of sorbitol and fructose in Schwarm cells, accumulation of these sugars may disrupt function and structure.
The earliest functional change in diabetic nerves is delayed nerve conduction velocity; the earliest histological change is segmental demyelination, due to damage to Schwarm cells. In the early stages axons are preserved, implying prospects of recovery, but at a later stage irreversible axonal degeneration develops.
The following varieties of neuropathy may occur :
1 Symmetrical mainly sensory polyneuropathy (distal) 2 Acute painful neuropathy
3 Mononeuropathy and multiple mononeuropathy:
(a) Cranial nerve lesions
(b) Isolated peripheral nerve lesions
4 Diabetic amyotrophy
5 Autonomic neuropathy

Symmetrical mainly sensory polyneuropathy

This is often unrecognized by the patient in its early stages. Early clinical signs are loss of vibration sense, pain sensation (deep before superficial) and temperature sensation in the feet. At later stages patients may complain of a feeling of ‘walking on cotton wool’ and can lose their balance when washing the face or walking in the dark owing to impaired proprioception. Involvement of the hands is less common and results in a ‘stocking and glove’ sensory loss. Complications include unrecognized trauma, beginning as blistering due to an ill-fitting shoes. or a hot water bottle, and leading to ulceration.
SEQUELAE OF NEUROPATHY. Involvement of motor nerves to the small muscles of the feet gives rise to interosseous wasting. Unbalanced traction by the long flexor muscles leads to a characteristic shape of the foot, with a high arch and clawing of the toes, which in turn leads  to abnormal distribution of pressure on walking, resulting in callus formation under the first metatarsal head or on the tips of the toes and perforating neuropathic ulceration.
Neuropathic arthropathy (Charcot’s joints) may sometimes develop in the ankle. The hands show small-muscle wasting as well as sensory changes but it is important to differentiate these signs and symptoms from those of the carpal tunnel syndrome, which occurs with increased frequency  in diabetes and may be amenable to surgery.

Acute painful neuropathy

A diffuse, painful neuropathy is rare. The patient describes burning or crawling pains in the feet, shins and anterior thighs. These symptoms are typically worse at night, and pressure from bedclothes may be intolerable. It may present at diagnosis or develop after sudden improvement in glycaemic control (e.g. when insulin is started). It usually remits spontaneously after 3-12 months if good control is maintained. A more chronic form, developing later in the course of the disease, is sometimes resistant to almost all forms of therapy. Neurological assessment is difficult because of the hyperaesthesia experienced by the patient, but muscle wasting is not a feature and objective signs can be minimal. Mononeuropathy and multiple mononeuropathy
CRANIAL NERVE LESIONS. Isolated or multiple palsies of nerves to the external eye muscles, especially the third and sixth nerves, are more common in diabetes. A characteristic feature of diabetic third nerve lesions is that pupillary reflexes are retained owing to sparing of pupillomotor fibres. Full spontaneous recovery is the rule.
ISOLATED PERIPHERAL NERVE LESIONS. Manifestations may be sensory, motor, or mixed, and multiple nerves may be involved (‘mononeuritis multiplex’). The onset is usually abrupt and sometimes painful; recovery is typically slow and incomplete. Lesions are more likely to occur at common sites for external pressure palsies or nerve entrapment, e.g. the median nerve in the carpal tunnel.

Diabetic amyotrophy

This condition is usually seen in older men with diabetes. Presentation is with painful wasting, usually asymmetrical,
of the quadriceps muscles. The wasting may be very marked and knee reflexes are diminished or absent. The affected area is often extremely tender. Extensor plantar responses sometimes develop and CSF protein content is elevated. Diabetic amyotrophy is usually associated with periods of poor glycaemic control and may be present at diagnosis. It often resolves in time with careful control of the blood glucose.

Autonomic neuropathy

Asymptomatic autonomic disturbances can be demonstrated on laboratory testing in many patients, but symptomatic autonomic neuropathy is rare. It affects both the sympathetic and parasympathetic nervous system and can be disabling.
THE CARDIOVASCULAR SYSTEM. Vagal neuropathy results in tachycardia at rest and loss of sinus arrhythmia. At a later stage the heart may become denervated (resembling a transplanted heart). Cardiovascular reflexes such as the Valsalva manoeuvre are impaired. Postural hypotension occurs owing to loss of sympathetic tone to peripheral arterioles. A warm foot with a bounding pulse is sometimes seen in a polyneuropathy as a result of peripheral vasodilatation.
GASTROINTESTINAL TRACT. Vagal damage can lead to gastroparesis, often asymptomatic, but sometimes leading to intractable vomiting. Diarrhoea often occurs at night accompanied by urgency and incontinence. Diarrhoea and steatorrhoea may occur owing to bacterial overgrowth and treatment is with antibiotics.
BLADDER INVOLVEMENT. Loss of tone, incomplete emptying, and stasis (predisposing to infection) can occur, and may ultimately result in an atonic, painless, distended bladder.
IMPOTENCE. This is common. The first manifestation is incomplete erection which may in time progress to total impotence; retrograde ejaculation also occurs. However, impotence in diabetes is not always due to autonomic neuropathy. Other causes include anxiety, depression, alcohol excess, drugs, primary or secondary gonadal failure and inadequate vascular supply due to atheroma in pudendal arteries. Treatment should ideally include sympathetic counselling of both partners. Some patients may benefit from intracavernous injection of papaverine or the use of vacuum devices to produce an erection which is then maintained by slipping a tight rubber band over the base of the penis until intercourse is complete.

The diabetic foot

Many amputations in diabetes could be delayed or prevented by more effective patient education and medical supervision. Ischaemia, infection and neuropathy combine to produce tissue necrosis. Although these factors may coexist, it is important to distinguish between the ischaemic and the neuropathic foot.

Management of the diabetic foot

Many diabetic foot problems are avoidable, so patients need to learn the principles of foot care and should be advised concerning appropriate footwear and the risks of smoking. Older patients should visit a chiropodist regularly and should not cut their own toe-nails. Once tissue damage has occurred in the form of ulceration
or gangrene, the aim is preservation of viable tissue. The two main threats are:
INFECTION. This rapidly takes hold in a diabetic foot, and early effective antibiotic treatment is essential. Collections of pus are drained and excision of infected bone is needed if osteomyelitis develops and does not respond to appropriate antibiotic therapy. Regular Xrays of the foot are needed to check on progress.
ISCHAEMIA. The blood flow to the feet is assessed clinically or with the Doppler ultrasound stethoscope. Femoral arteriography may be performed, since localized areas of occlusion may be amenable to bypass surgery or angioplasty.
Foot problems are the major cause of hospital bed occupancy by diabetic patients. Good liaison between surgeon and physician is essential if this period in hospital is to be used efficiently. When irreversible arterial insufficiency is present, it is often quicker and kinder to opt for an early major amputation rather than subject the patient to a debilitating sequence of conservative procedures.

Diabetic Eye Disease

Diabetes can affect the eyes in a number of ways. The most common and characteristic form of involvement is diabetic retinopathy. About one in three young patients is likely to develop visual problems, and in the UK 5% have in the past become blind after 30 years of diabetes; diabetes is the commonest cause of blindness in the population as a whole up to the age of 60 years. Other forms of eye disease may also occur: THE LENS may be affected by reversible osmotic changes in patients with acute hyperglycaemia-causing blurred vision-or by cataracts.

NEW VESSEL FORMATION IN THE IRIS (rubeosis iridis) may develop as a late complication of diabetic retinopathy and can cause glaucoma.
EXTERNAL OCULAR PALSIES, especially of the sixth nerve, can occur (a mononeuritis). The natural history of retinopathy.

Diabetes causes increased thickness of the basement membrane and increased permeability of the retinal capillaries. Aneurysmal dilatation may occur in some vessels while others become occluded. These changes are first detectable by fluorescein angiography: a fluorescent dye is injected into an arm vein and photographed in transit through the retinal vessels. This technique is not necessary to screen effectively for retinal disease. After 20 years of IDDM, almost all patients have some retinopathy- 60% progress to proliferative retinopathy. After 20 years of NIDDM >80% have some retinopathy, 20% have proliferative changes.
BACKGROUND RETINOPATHY. The first abnormality visible through the ophthalmoscope is the appearance of dot ‘haemorrhages’, which are actually due to capillary microaneurysms. Leakage of blood into the deeper layers of the retina produces the characteristic ‘blot’ haemorrhage,  while exudates of fluid rich in lipids and protein give rise to hard exudates. These have a bright yellowwhite colour and are often irregular in outline with a sharply defined margin. These changes rarely develop in young patients with a duration of diabetes under 10 years, but by 20 years virtually all eyes will at least manifest the occasional dot haemorrhage on careful ophthalmoscopy. In contrast, retinopathy may be present at diagnosis or shortly thereafter in older patients.
Background retinopathy does not in itself constitute a threat to vision but may progress to two other distinct forms of retinopathy: maculopathy or proliferative retinopathy. Both are the consequence of damage to retinal blood vessels and resultant retinal ischaemia.

Background retinopathy and maculopathy.

Background retinopathy and maculopathy.

Prevalance of retinopathy in relation to duration of the disease in patients with insulin-dependent diabetes mellitus diagnosed under the age of 33 years.

Prevalance of retinopathy in relation to duration
of the disease in patients with insulin-dependent diabetes mellitus diagnosed under the age of 33 years.

DIABETIC MACULOPATHY. This may lead to blindness in the absence of proliferation and particularly affects the older patient with IDDM. Macular oedema is the first feature of maculopathy and may in itself result in permanent  macular damage if not treated early. The first, and only, sign of this is deteriorating visual acuity and the condition cannot be diagnosed with standard ophthalmoscopy.
This is why it is essential to screen patients with diabetes regularly for changes in visual acuity. In most cases, however, maculopathy does not generate sufficient oedema to cause early loss of acuity. The process may  then be detected in its later stages as encroachment of hard exudates and haemorrhages on the macular area. These changes are easily visible on ophthalmoscopy, but only through fully dilated pupils.
PREPROLIFERA TIVE RETINOPATHY. Progressive retinal ischaemia leads to further changes which herald proliferative, sight-threatening retinopathy. The earliest sign is the appearance of ‘cotton-wool spots’, representing oedema resulting from retinal infarcts. Unlike hard exudates they may also occur in severe hypertensive retinopathy. The term ‘soft exudate’ is often used synonymously  but is best avoided. Cotton-wool spots are greyish-white, have indistinct margins and a dull matt surface, unlike the glossy appearance of hard exudates. Venous beading and/or venous loops are other recognized preproliferative changes.
PROLIFERATIVE RETINOPATHY. Hypoxia is thought to be the signal for formation of new vessels. These lie superficially or grow forward into the vitreous, resembling fronds of seaweed. They branch repeatedly, are fragile, bleed easily (because they lack the normal supportive tissue) and may give rise to a fibrous-tissue reaction. With advanced retinopathy, haemorrhages can be preretinal or into the vitreous. A vitreous haemorrhage presents as a loss of vision in one eye, sometimes noticed on waking, or as a floating shadow affecting the field of vision. Ophthalmoscopy gives the appearance of a featureless, grey haze. Partial recovery of vision is the rule, as the blood is reabsorbed, but repeated bleeds may occur. Loss of vision may also result from fibrous proliferation associated with new vessel formation. This may give rise to traction bands that contract with the course of time, producing retinal detachment.


Senile cataracts develop some 10-15 years earlier in diabetic patients than in the remainder of the population. Juvenile or ‘snowflake’ cataracts are much less common. These are diffuse, rapidly progressive cataracts associated with very poorly controlled diabetes. They should be distinguished from temporary lens changes that occasionally appear during hyperosmolar coma and resolve when the coma is brought under control.

Examination of the eye

Careful systematic examination of the eye is essential. Visual acuity and eye movements are tested, and the pupils are dilated with a quick-acting mydriatic such as tropicamide 0.5%. Dilating drugs should not be used in patients  with a history of glaucoma, except with the advice of an ophthalmologist.
The examination begins at arm’s length. At this distance, cataracts are silhouetted against the red reflex of the retina. The ophthalmoscope is advanced until the retina is in focus. The examination begins at the optic disc, moves through each quadrant in turn, and ends with the macula (since this is least comfortable for the patient). The ophthalmoscope is then adjusted to the +10 dioptre lens for examination of the cornea, anterior chamber and lens. The location of abnormalities should always be sketched in the notes for future reference. Management of diabetic eye disease.
There is no specific medical treatment for background retinopathy, but patients are advised not to smoke and hypertension should be treated. Rapid progression may occur in pregnant patients and in those with nephropathy, and these groups need frequent monitoring. All  patients with retinopathy should be examined regularly by a diabetologist or ophthalmologist. Early referral to an ophthalmologist is essential in the following circumstances:
• Deteriorating visual acuity
• Hard exudates encroaching on the macula
• Preproliferative changes (cotton-wool spots or venous beading)
• New vessel formation
The ophthalmologist may perform fluorescein angiography to define the extent of the problem. Maculopathy and proliferative retinopathy are treatable by retinal laser photocoagulation; in the latter condition early effective therapy reduces the risk of visual loss by about 50%. The value of photocoagulation is particularly marked in those with disc (as against peripheral) new vessels. In one trial only 15% of treated, as against 50% of untreated, eyes with disc new vessels progressed to legal blindness. Treatment in this case is by panretinal photocoagulation with 2000–5000 laser burns to each eye.

Problems of management

HYPOTENSION. This may lead to renal shutdown.
Plasma expanders (or whole blood) are therefore given if the systolic blood pressure is below 80 mmHg. A central venous pressure line is useful in this situation. A bladder catheter is inserted if no urine is produced within 2 hours, but routine catheterization is unnecessary.
COMA. The usual principles apply. It is essential to pass a nasogastric tube to prevent aspiration since gastric stasis is common and the rare, but fatal, complication of acute gastric dilatation may result.
CEREBRAL OEDEMA. This rare, but feared, complication has mostly been reported in children or young adults. Excessive rehydration and use of hypertonic fluids such as 8.4% bicarbonate may sometimes be responsible. The mortality is high.
HYPOTHERMIA. Severe hypothermia with a core temperature below 33°C may occur and may be overlooked unless a rectal temperature is taken with a lowreading thermometer.
LATE COMPLICATIONS. These include stasis pneumonia and deep-vein thrombosis, and occur especially in the comatose or elderly patient.
COMPLICATIONS OF THERAPY. These include hypoglycaemia and hypokalaemia. Overenthusiastic fluid replacement may precipitate pulmonary oedema in older patients. Hyperchloraemic acidosis may develop in the course of treatment since patients have lost a large variety of negatively charged electrolytes, which are replaced with chloride. The kidneys usually correct this spontaneously within a few days.


Hyperglycaemia-measure blood glucose Ketonaemia-test plasma with KetostixlAcetest Acidosis-measure blood gases.


Blood glucose
Urea and electrolytes
Full blood count
Blood gases
Blood and urine culture
Chest X-ray

Subsequent management

Intravenous fluids and insulin are continued until the patient feels able to eat and keep food down. The drip is then taken down and a similar amount of insulin is given as three or four soluble subcutaneous doses per day until a maintenance regimen can be restarted. Sliding-scale regimens are often unnecessary and may even delay the establishment of stable blood glucose levels.
The treatment of diabetic ketoacidosis is incomplete without a careful enquiry into the causes of the episode and advice as to how to avoid its recurrence.

Non-ketotic hyperosmolar state

This condition, in which severe hyperglycaemia develops without significant ketosis, is the metabolic emergency characteristic of uncontrolled NIDDM. Patients present in middle or later life, often with previously undiagnosed diabetes. Common precipitating factors include consumption of glucose-rich fluids (e.g. Lucozade), concurrent medication such as thiazide diuretics or steroids, and intercurrent illness.
Non-ketotic coma and ketoacidosis represent two ends of a spectrum rather than two distinct disorders. The biochemical differences.May partly be explained as follows:
AGE. The extreme dehydration characteristic of nonketotic coma may be related to age. Old people experience thirst less acutely, and more readily become dehydrated. In addition, the mild renal impairment associated with age results in increased urinary losses of fluid and electrolytes.
THE DEGREE OF INSULIN DEFICIENCY is less severe in non-ketotic coma. Endogenous insulin levels are sufficient to inhibit hepatic ketogenesis, whereas glucose production is unrestrained.

Electrolyte changes in diabetic ketoacidosis and non-ketotic hyperosmolar state.

Electrolyte changes in diabetic
ketoacidosis and non-ketotic hyperosmolar state.


The characteristic clinical features are dehydration and stupor or coma. Impairment of consciousness is directly related to the degree of hyperosmolality. Evidence of underlying illness such as pneumonia or pyelonephritis may be present, and the hyperosmolar state may predispose to stroke, myocardial infarction or arterial insufficiency in the lower limbs.


These are according to the guidelines for ketoacidosis with some exceptions. Many patients are extremely sensitive to insulin and the glucose concentration may plummet The resultant change in osmolality may cause cerebral damage. It is sometimes useful to infuse insulin at a rate of 3 U hour-I for the first 2-3 hours, increasing to 6 U hour-I if glucose is falling too slowly. Normal saline is the standard fluid for replacement. Avoid halfnormal saline (0.45%) except in exceptional circumstances, since rapid dilution of the blood may cause more cerebral damage than a few hours of exposure to hypernatraemia.


The reported mortality is around 20-30%, mainly because of the advanced age of the patients and the frequency of intercurrent illness. Unlike ketoacidosis, nonketotic hyperglycaemia is not an absolute indication for subsequent insulin therapy, and survivors may do well on diet and oral agents.

Lactic acidosis

Lactic acidosis may occur in diabetic patients on biguanide therapy. Phenformin, the agent responsible in the great majority of reported cases, has now been withdrawn in the UK. The risk in patients taking metformin is extremely low provided that the therapeutic dose is not exceeded and the drug is withheld in patients with advanced hepatic or renal dysfunction.
Patients present with a severe metabolic acidosis, usually without significant hyperglycaemia or ketosis, and treatment is by rehydration and infusion of isotonic 1.26% bicarbonate. The mortality is in excess of 50%.


When insulin was introduced it was assumed that it would provide complete and adequate replacement therapy, just as thyroxine does in hypothyroidism. Time proved that insulin-treated patients still have a considerably reduced life expectancy. Those diagnosed before the age of 20 years have only a 60-70% chance of living past the age of 45 years, although there are recent indications of improved survival. The excess deaths are mainly due to diabetic nephropathy, but there is also a considerable excess cardiovascular mortality. Heart disease, peripheral vascular disease and stroke are the major causes of death in patients over the age of 50 years.

Macrovascular complications

Diabetes is a risk factor in the development of atherosclerosis. This risk is related to that of the background population. For example, Japanese diabetics are much less likely to develop atherosclerosis than patients in Europe but are much more likely to develop it than non-diabetic Japanese. The excess risk to diabetics compared with the general population increases as one moves down the body:

STROKE is twice as likely.
MYOCARDIAL INFARCTION is three to five times as likely and women with diabetes lose their prerneno pausal protection from coronary artery disease.
AMPUTA TION OF A FOOT for gangrene is 50 times as likely.
Diabetes is additive with other risk factors for large-vessel disease. In other words, the diabetic who smokes or is obese, hypertensive or hyperlipidaemic adds the risks conferred by these conditions to that of diabetes itself.

Duration of diabetes


Systolic hypertension

Hyperinsulinaemia due to insulin resistance associated with obesity and syndrome X Hyperlipidaemia, particularly hypertriglyceridaemia Proteinuria (including microalbuminuria) Other factors are the same as for the general population.

Insulin resistance

Hyperinsulinaemia due to insulin resistance associated with obesity is sometimes known as ‘syndrome X. Confusingly the same term is used by cardiologists for a rare variant of angina. Syndrome X includes glucose intolerance, hypertension, central obesity and dyslipoproteinaemia (increased very low density lipoprotein and reduced high-density lipoprotein). It is found in patients with NIDDM and carries a high risk of coronary artery disease.

Microvascular complications

In contrast to macrovascular disease, which is prevalent in Western populations as a whole, microvascular disease is specific to diabetes. Small blood vessels throughout the body are affected but the disease process is of particular danger in three sites:
• Retina
• Renal glomerulus
• Nerve sheath
Diabetic retinopathy, nephropathy and neuropathy tend to manifest 10-20 years after diagnosis in young patients. They present earlier in older patients, probably because these have had unrecognized diabetes for months or even years prior to diagnosis.


Diabetic ketoacidosis


Diabetic ketoacidosis is the hallmark of IDDM. Its main causes can be grouped as follows:
• Previously undiagnosed diabetes
• Interruption of insulin therapy
• The stress of intercurrent illness
The majority of cases reaching hospital could have been prevented by earlier diagnosis, better communication between patient and doctor, and better patient education. The most common error of management is for patients to reduce or omit insulin because they feel unable to eat owing to nausea or vomiting. This is a factor in at least 25% of all hospital admissions. Insulin should never be stopped.


Ketoacidosis is a state of uncontrolled catabolism associated with insulin deficiency. Insulin deficiency is a necessary precondition since only a modest elevation in insulin levels is sufficient to inhibit hepatic ketogenesis. Even so,  stable patients do not readily develop ketoacidosis when insulin is withdrawn. Other factors include counteriregulatory hormone excess and fluid depletion. The combination of insulin deficiency with excess of its hormonal antagonists leads to the parallel processes

Dehydration occurs during ketoacidosis as a consequence of two parallel processes.

Dehydration occurs during ketoacidosis as a
consequence of two parallel processes.

In the absence of insulin, hepatic glucose production accelerates and peripheral uptake by tissues such as muscle is reduced. Rising glucose levels lead to an osmotic diuresis, loss of fluid and electrolytes, and dehydration. Plasma osmolality rises and renal perfusion falls. In parallel, rapid lipolysis occurs, leading to elevated circulating free fatty-acid levels. The free fatty acids are broken down to fatty acyl-CoA within the liver cells, and this in turn is converted to ketone bodies within the mitochondria.
Accumulation of ketone bodies produces a metabolic acidosis. This is typically associated with nausea and vomiting, leading to further loss of fluid and electrolytes. The excess ketones are excreted in the urine but also appear in the breath, producing a distinctive smell similar to that of acetone. Respiratory compensation for the acidosis leads to hyperventilation, graphically described as ‘air hunger’. Progressive dehydration impairs renal excretion of hydrogen ions and ketones, aggravating the acidosis. As the pH falls below 7.0 ([H+] >100 nmol Iitre “), pHdependent enzyme systems in many cells function less effectively. Untreated, severe ketoacidosis is invariably fatal.


The features of ketoacidosis are those of uncontrolled diabetes with acidosis, and include prostration, hyperventilation (Kussmaul respiration), nausea, vomiting and, occasionally, abdominal pain. The latter is sometimes so severe as to cause confusion with a surgical acute abdomen.

Some patients are mentally alert at presentation, but confusion and stupor are common. Up to 5% present in coma. Evidence of marked dehydration is present and the eyeball is lax to pressure in severe cases. Hyperventilation is present but becomes less marked in very severe acidosis owing to respiratory depression. The smell of ketones on the breath allows an instant diagnosis to be made by those able to detect the odour. The skin is dry and the body temperature is often subnormal, even in the presence of infection; in such cases, pyrexia may develop later.

Ketogenesis. During insulin deficiency, lipolysis accelerates and free fatty acids taken up by liver cells form the substrate for ketone formation (acetoacetate, acetone and 13- hydroxybutyrate) within the mitochondrion.

Ketogenesis. During insulin deficiency, lipolysis accelerates and free fatty acids taken up by liver cells form the substrate for ketone formation (acetoacetate, acetone and 13-
hydroxybutyrate) within the mitochondrion.


This is confirmed by demonstrating hyperglycaemia with ketonaemia or heavy ketonuria, and acidosis. No time should be lost and treatment is started as soon as the first blood sample has been taken. Hyperglycaemia is demonstrated by dipstick, while a blood sample is sent to the laboratory for confirmation. Ketonaemia is confirmed by centrifuging a blood sample and testing the plasma with a dipstick that measures ketones. An arterial blood sample is taken for blood gas analysis.

Further investigations are detailed below.


The principles of management are as follows.
REPLACE THE ELECTROLYTE LOSSES. Potassium levels need to be monitored with great care to avoid the cardiac complications of hypokalaemia or hyperkalaemia.
RESTORE THE ACID-BASE BALANCE. A patient with healthy kidneys will rapidly compensate for the metabolic acidosis once the circulating volume is restored. Bicarbonate is seldom necessary, although it is usual for this to be given as isotonic 1.26% bicarbonate solution (not 8.4% which is grossly hyperosmolar) if the pH is below 7.0 ([H+] >100 nmollitre-‘).
REPLACE THE DEFICIENT INSULIN. Modern treatment is with low doses of insulin, which lower blood glucose by suppressing hepatic glucose output rather than by stimulating peripheral uptake, and are therefore much less likely to produce hypoglycaemia. Soluble insulin is given as an intravenous infusion where facilities for adequate supervision exist, or as hourly intramuscular injections. The subcutaneous route is avoided because subcutaneous blood flow is reduced in shocked patients.
MONITOR BLOOD GLUCOSE CLOSELY. Hourly measurement is needed in the initial phases of treatment.
REPLACE THE ENERGY LOSSES. When plasma glucose falls to near-normal values (12 mrnol litre “), saline infusion should be replaced with 5% dextrose containing 20 mmol KCl litre “. The insulin infusion rate is reduced and adjusted according to blood glucose.
SEEK THE UNDERLYING CAUSE. Physical examination may reveal a source of infection, e.g. a perianal abscess. Two common markers of infection are misleading: fever is unusual even when infection is present and polymorpholeucocytosis is present even in the absence of infection. Relevant investigations include a chest Xray, urine and blood cultures, and an ECG (to exclude myocardial infarction). If infection is suspected, broadspectrum antibiotics are started once the appropriate cultures have been taken.

Guidelines for the diagnosis and management of diabetic ketoacidosis.

Guidelines for the diagnosis and
management of diabetic ketoacidosis.

Social implications

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.


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.

Terms used in uncontrolled diabetes.

Terms used in uncontrolled diabetes.


Guidelines to therapy

All patients with diabetes require diet therapy. Good glycaemic control is unlikely to be achieved with insulin or oral therapy when diet is neglected, especially when the patient is also overweight. Insulin is always indicated in a patient who has been in ketoacidosis, and is usually indicated in patients who present under the age of 40 years. Insulin is also indicated in older patients following primary or secondary failure of oral therapy (see below). Tablets should be avoided in younger patients and are contraindicated in pregnancy. In older patients the approach to therapy is empirical. Diet alone should be tried in the first instance, and dietary knowledge and compliance should always be reassessed with care before proceeding to the next step. This is of particular importance in the obese patient who fails to lose weight.
When diet fails to achieve satisfactory control, thin patients are usually treated with a sulphonylurea drug, and obese patients with a biguanide. Primary failure of  treatment occurs when these agents (alone or in combination) never achieve the desired level of control.
Other patients may show a good initial response followed by progressive loss of control over the succeeding months or years; this is referred to as secondary failure of  treatment.
Since this approach is largely empirical, it is not surprising that practice differs from one country to another. For example> metformin (the only biguanide in common  use) is very widely employed in France, tends to be used less in the UK, and is not licensed in the USA. Criteria of control also vary, so that what is classed as primary failure at one centre may be seen as a success in another. The most widespread error in management is procrastination;
the patient whose control is inadequate on tablets should start insulin without undue delay.


The diet for a diabetic patient is no different from the diet considered healthy for the population as a whole.


This should consist of unrefined carbohydrate rather than simple sugars such as sucrose. Carbohydrate is absorbed relatively slowly from fibre-rich foods, preventing the rapid swings in circulating glucose seen when refined sugars are ingested. For example, the glucose peak seen after eating an apple is much flatter than that seen after drinking the same amount of carbohydrate as apple juice.
Calories Calories should be tailored to the needs of the patient. The total amount of carbohydrate in the diet should provide  50-55% of the total calories with fat 30-35% and protein 15%.
THE OVERWEIGHT DIABETIC PATIENT is started on a reducing diet of approximately 1000-1600 kcal daily (4000-6000 kJ).
THE LEAN PATIENT is put on an isocaloric diet.
PATIENTS WHO ARE UNDERWEIGHT because of untreated diabetes require energy supplementation.

Prescribing a diet

Most people find it extremely difficult to modify their eating habits, and repeated advice and encouragement are needed if this is to be achieved. A diet history is taken,  and the diet prescribed should involve the least possible interference with the life-style of the patient. It is important to stress that patients on insulin or oral agents should eat the same amount at the same time each day. Patients on insulin require snacks between meals and at bedtime to buffer the effect of injected insulin. Alcohol is not forbidden, but its energy content should be taken into account. Patients on insulin should be warned to avoid alcoholic binges since these may precipitate severe hypoglycaemia.
The role of patient education and community care The care of diabetes is based on self-management by the patient, who is helped and advised by those with specialized  knowledge. The quest for improved glycaemic control has made it clear that whatever the technical expertise applied, the outcome depends on willing cooperation by the patient. This in turn depends on an understanding of the risks of diabetes and the potential benefits of glycaemic control and other measures such as maintaining a lean weight> stopping smoking and taking care of the feet. If accurate information is not supplied, misinformation from friends and other patients will take its place. For this reason, many patients have exaggerated fears of, for example, blindness (about 1 patient in 20 is blind after 30 years of diabetes), death during hypoglycaemia (extremely rare), or the risk of passing diabetes on to their children (2-5% of offspring develop IDDM). Organized training programmes involving all healthcare workers including nurse specialists, dietitians and chiropodists are now a recognized part of good diabetes care.

Tablet treatment


These have two main actions:
1 They increase basal and stimulated insulin secretion.
2 They reduce peripheral resistance to insulin action.
The effects upon insulin secretion are most marked in the early stages of treatment, but the peripheral effects are more important for maintenance therapy. The sulphonylureas should be avoided in young ketotic patients, who require early insulin therapy, and are contraindicated in pregnancy. Insulin should be substituted during major surgery or severe intercurrent illness. These drugs have similar actions and potency. All should be used with care in patients with liver disease, and only those primarily excreted by the liver should be given to patients with renal impairment. Sulphonylureas all encourage weight gain and are not the first choice in obese patients. Tolbutamide is the safest drug in the very elderly because of its short duration of action. Chlorpropamide has the disadvantages of a long duration of action and a wider range of side-effects, and is now less widely used.
DRUG INTERACTIONS. All sulphonylureas bind to circulating albumin and may be displaced by other drugs, such as sulphonamides, that compete for their binding sites. Their clinical effect may be reduced by thiazide diuretics or steroid therapy. SIDE-EFFECTS. Hypoglycaemia is the most common and dangerous side-effect. Because the action of many  sulphonylureas persists for more than 24 hours, recurrent or prolonged hypoglycaemia is likely, and hospital admission is usually necessary. Skin rashes and other sensitivity reactions may occur.
Chlorpropamide use is often associated with a facial flush when alcohol is taken. It may also cause a cholestatic jaundice and a syndrome of inappropriate antidiuretic hormone (ADH) secretion in 2-4% of patients.

Drugs used in non-insulin-dependent diabetes mellitus.

Drugs used in non-insulin-dependent diabetes mellitus.


Metforrnin acts by reducing glucose absorption from the gut and by increasing insulin sensitivity. Unlike the sulphonylureas it does not induce hypoglycaemia in normal volunteers. It is usually reserved for patients in middle or old age, particularly for the overweight since it does not promote weight gain. It may be given in combination with sulphonylureas when a single agent has proved to be ineffective.
Its side-effects include anorexia, epigastric discomfort and diarrhoea. Lactic acidosis has occurred in patients with severe hepatic or renal disease, and metformin is contraindicated when these are present.

Other drugs

Acarbose, an a-glucosidase inhibitor, inhibits intestinal amylase, sucrase and maltase activity, thereby reducing carbohydrate absorption. It is being used in NIDDM patients who are inadequately controlled on diet alone or on diet with oral hypoglycaemic agents. Its long-term value is unproven.

Insulin treatment


The needles used to inject insulin are very fine and sharp. Even though most injections are virtually painless, patients are understandably apprehensive and treatment begins with a lesson in injection technique. Insulin is either drawn up into special plastic insulin syringes marked in units (100 U in 1 ml), or is administered by a pen injection device. Injections are given at 90° to the skin of the thighs or abdomen, and the needle is usually inserted to its full length. Most patients starting insulin injection prefer pen devices when given a choice.


• Useful in the visually impaired (audible clicks, clear numbering)
• Easy to use
• Convenient to carry around
• Can be used discreetly in public places (e.g. restaurants)
• Some psychological benefit in ‘needle phobias’
• Available free in the UK


• Zinc insulins cannot be used because they aggregate in the pen cartridges.
• Pen needles are not yet available on prescription.
• Pens produced by one manufacturer cannot be used with insulin from another.
The injection site used should be changed regularly to prevent areas of lipohypertrophy. The rate of insulin absorption depends on local subcutaneous blood flow, and is accelerated by exercise, local massage or a warm  environment. Absorption is more rapid from the abdomen than from the arm, and is slowest from the thigh. All these factors can influence the shape of the insulin profile. All patients need careful training for a life with insulin, but routine hospital admission to begin insulin treatment is unnecessary where facilities for community support exist.

Choice of insulin

SPECIES. Insulin is found in every creature with a backbone, and the central part of the molecule shows few species differences. (For example, fish insulin produces hypoglycaemia in humans.) Small differences in the amino acid sequence may alter the antigenicity of the molecule. Beef insulin differs from human insulin by three amino acids and pork from human by one. Both may induce antibody formation, beef more readily than pork.
Human insulin is produced by DNA coding of cultured yeast or bacterial cells to produce proinsulin, with subsequent enzymatic cleavage to insulin. All forms of injected insulin, even human, may result  in antibody formation, but insulin antibodies usually have little clinical importance. Human insulin has largely replaced the other varieties, mainly as the result of market forces. Some patients report altered perception of hypoglycaemia on human insulins. This effect has not been reproduced in double blind trials, but such patients should be offered the opportunity to try a different species of insulin.

Human insulin preparations.

Human insulin preparations.

PURITY. Insulins used in the Western World are now of very high purity, but older products are still widely distributed elsewhere. FORMULATION (Table 17.5). There are two main types
of insulin:
1 Insulin prepared in a clear solution (soluble or crystalline). These insulins are short-acting and are the only insulins to be used in emergencies such as ketoacidosis  or for surgical operations.
2 Insulins premixed with retarding agents (either protamine or zinc) that precipitate crystals of varying size according to the conditions employed. These insulins are intermediate or long acting.

Clinical use

In normal subjects a sharp increase in insulin occurs after meals; this is superimposed on a constant background of secretion. Insulin therapy attempts to reproduce this pattern. In order to achieve this, a common strategy is to give intermediate-acting insulin to control the afternoon and night blood sugar level, and shortacting insulins morning and evening to match meal times.
Ideal control is often hard to achieve for four reasons:
In normal people, insulin is secreted directly into the portal circulation and passes directly to the liver in high concentration. The insulin injected by diabetics passes into the systemic circulation before passage to the liver.
2 Subcutaneous soluble insulin takes 60-90 min to achieve peak plasma levels-the onset and offset of action are too slow.
3 The absorption of subcutaneous insulin into the circulation is variable; the longer acting the preparation, the more erratic the absorption.
4 Basal insulin levels are constant in the normal state, but injected insulin invariably peaks and declines, with resulting swings in metabolic control.

Glucose and insulin profiles in normal subjects.

Glucose and insulin profiles in normal subjects.

Individuals vary and therapy must be tailored accordingly. One approach to therapy is outlined here.
YOUNG PATIENTS are started on two injections daily of an intermediate insulin at a dose of 8-10 U twice daily. Some recovery of endogenous insulin secretion may occur over the first few months (the ‘honeymoon period’) and the insulin dose may need to be reduced. Requirements rise thereafter and a multiple injection regimen is then appropriate for most younger patients. This is flexible and usually highly acceptable.
PATIENTS WITH NIDDM. Twice-daily injections of premixed soluble and isophane insulins, e.g. Mixtard, are effective in the majority of patients with NIDDM. OLDER PATIENTS may sometimes manage adequately on a single daily injection.
If glycaemic control is inadequate with the standard approach, the alternatives are multiple insulin injections or continuous subcutaneous insulin infusion (CSII). Both methods require a planned approach to life, with special attention to diet and exercise and frequent bloodglucose testing.

Insulin regimens. Profiles of soluble insulins are shown as blue lines and intermediate- or long-acting insulin as purple lines. The arrows indicate when the injections are given. (a) Twice daily soluble and intermediate. (b) Splitting the evening injection. (c) Three times daily soluble with additional intermediate or long-acting insulin given before bedtime. B, breakfast; L, lunch; S, supper; Sn, snack (bedtime).

Insulin regimens. Profiles of soluble insulins are
shown as blue lines and intermediate- or long-acting insulin as
purple lines. The arrows indicate when the injections aregiven. (a) Twice daily soluble and intermediate. (b) Splitting the evening injection. (c) Three times daily soluble with
additional intermediate or long-acting insulin given before bedtime. B, breakfast; L, lunch; S, supper; Sn, snack (bedtime).

MULTIPLE INJECTIONS. The introduction of ‘pen injection’ devices has made this approach much more acceptable to patients. Two variants are shown diagrammatically. Multiple injection regimens and infusion devices have the advantage of flexibility concerning meal times, which is of great value to patients with busy jobs, shift workers and those who travel regularly. The amount eaten at each meal can be chosen at meal time and an appropriate dose of insulin given. With twice daily regimens, the size and timing of meals is fixed more rigidly.

INFUSION DEVICES. CSII is delivered by a small pump strapped around the waist that infuses a constant trickle of insulin via a needle in the subcutaneous tissues. Mealtime doses are delivered when the patient touches a button on the side of the pump.
This approach is particularly useful in the overnight period. Disadvantages include the nuisance of being attached to a gadget, skin infections, and the risk of ketoacidosis if the flow of insulin is broken (since these patients have no protective reservoir of depot insulin). Infusion pumps should only be used by specialized centres able to offer a round-the-clock service to their patients.


Genetic susceptibility

100M is not genetically predetermined, but an increased susceptibility to the disease may be inherited.

Identical twins

The identical twin of a patient with mOM has a 30-35% chance of developing the disease. This implies that nongenetic factors must also be involved.


The child of an insulin-dependent diabetic patient has anincreased chance of developing mOM. This risk, curiously, is greater with a diabetic father (between 1 in 20 and 1 in 40) than with a diabetic mother (1 in 40-80).
If one child in a family has IDDM, each sibling has a 1 in 20 risk of developing diabetes. If a sibling is HLAidentical, the risk rises to 1 in 6.
H LA SYSTEM. More than 90% of IDDM patients carry HLA-DR3 and/or DR4 compared with 40% of the general population. The relative risk conferred by DR3 is about 7, DR4 about 9, but the highest risk of 14 is with DR3/DR4 heterozygote. Since the risk is additive a model based on two susceptibility genes (one associated with DR3 and one with DR4) has been proposed. Stronger associations have been reported with the DQ region. Most people have two alleles with aspartic acid at position 57 on the HLADQ {3chain which confer resistance to the development of diabetes. Substitution of aspartate at position 57 by another amino acid considerably increases susceptibility to IDDM as do alleles coding for arginine at position 52 on the a chain. These DQ polymorphisms determine the degree of autoimmune response against the pancreatic islet cells but other genes, probably interacting with  environmental factors, are necessary for the development of diabetes.
In contrast, individuals with HLA-DR2 have a considerably reduced risk (0.12 times normal) of developing diabetes. The reasons for this protective effect are unclear.
THE INSULIN GENE. Associations between IDDM and other chromosomes (apart from chromosome 6) have been described. A polymorphous DNA region close to the  insulin gene on chromosome 11 has been studied, and short, intermediate and long insertions (see p. 109) have been reported. Homozygosity of the short (class 1) allele is found in some 80% of patients with IDDM as against 40% of controls.

Autoimmunity and insulin-dependent diabetes mellitus.
Several pieces of evidence suggest that autoimmune processes are involved in the pathogenesis of IDDM. ASSOCIATION WITH OTHER AUTOIMMUNE DISEASE.  Autoimmune thyroid disease, Addison’s disease and pernicious anaemia are more common in patients and their relatives.
IMMUNOGENETIC ASSOCIATIONS. The association between HLA-DR3, DR4 and diabetes, and the protective  effect of DR2 might represent idiosyncracies in theimmune process that result in increased (or reduced)  susceptibility.
THE INSULITIS PROCESS. Autopsies of patients who died soon after diagnosis are characterized by infiltration  of the pancreatic islets by mononuclear cells. A similar pattern occurs in other autoimmune diseases, e.g. thyroiditis.
IMMUNE ABNORMALITIES AT DIAGNOSIS. About 70% of newly presenting patients have islet-cell antibodies.  These react with human islets and can be detected by immunofluorescence. They usually become undetectable  within a few years of diagnosis. In such patients increased numbers of activated T lymphocytes may also be present in the circulation at diagnosis.
Of newly diagnosed patients with IDDM, 80% have antibodies to the enzyme glutamic acid decarboxylase (GAD). Pancreatic {3 cells have high levels of this enzyme and the GAD autoantigen may be critical for the initiation of {3cell destruction.
IMMUNOSUPPRESSION with agents such as cyclosporin at or soon after diagnosis prolongs {3-cell survival.

Environmental factors

A viral aetiology has been suspected for many years. This is based on the seasonal incidence of the condition, anecdotal associations, and analysis of viral antibody titres at diagnosis. IgM antibodies to Coxsackie B4 have been  reported in 20-30% of new cases. However, in view of the long prodromal period (see below), it seems likely that viruses precipitate rather than initiate the onset of diabetes.

The diabetes prodrome

Prospective study of first-degree relatives of children with diabetes has revealed that islet-cell antibodies may appear in the circulation months or even years before diagnosis. Insulin antibodies have also been reported in this period, and abnormalities of insulin secretion in response to intravenous glucose may develop. The sequence of events leading to diagnosis may be as shown. Better understanding of this sequence may in time permit strategies of prevention to be tested.

Genetic predisposition

Initiating event (? environmental)
Immune activation

Progressive fI cell 1055
Abnormalities of insulin secretion
7 Second trigger (7 viral)
Clinical onset
‘Remission’ (partial recovery of fI cell function)
Loss of insulin secretion

Non-insulin-dependent diabetes mellitus (type II diabetes)


Unlike IDDM this is relatively common in all populations enjoying an affluent life-style. Large differences in prevalence have been reported. The disease may be present in a subclinical form for years before diagnosis, and the incidence  increases markedly with age and degree of obesity. The onset may be accelerated by the stress of pregnancy, drug treatment or intercurrent illness. Estimates of prevalence using the WHO criteria would suggest an overall prevalence of around 2% in the UK. NIDDM is twice as prevalent in people of Afro-Caribbean ancestry and three to five times more prevalent in people from South Asia than in white Europeans.
The rates have been reported in various populations. Epidemiological surveys suggest that indolent well-fed populations are two to twenty times as likely to develop NIDDM as lean populations of the same race.



Identical twins of a patient with NIDDM have an almost 100% chance of developing diabetes and about 25% of other patients have a first-degree relative with NIDDM. Certain families exist in which diabetes appears to travel  as an autosomal dominant characteristic, but NIDDM is almost certainly a polygenic disorder. Some families show abnormalities of the gene which codes for the enzymeglucokinase on chromosome 7. The defect differs from one fam ily to another and more than 20 mutations have been described so far. Studies in other family groups have  shown linkage between the inheritance of diabetes and markers on chromosomes 3 and 20. The next decade is likely to see the identification of many of the gene abnormalities which predispose to NIDDM.


There is no evidence of immune involvement in its pathogenesis.

Prevalence of non-insulin-dependent diabetes mellitus in various populations (WHO criteria).

Prevalence of non-insulin-dependent diabetes
mellitus in various populations (WHO criteria).

Insulin secretion and action

Patients with NIDDM, unlike those with IDDM, retain about 50% of their f:l-cell mass at autopsy. Abnormalities of insulin secretion develop early in the course of the disease. Normal subjects have a biphasic  insulin response to intravenous glucose. In NIDDM the first-phase insulin response to intravenous glucose is lost, and insulin secretion in response to oral glucose is delayed and exaggerated. The majority of patients manifest reduced insulin secretion relative to the prevailing glucose concentration, and progressive f:l-cell loss occurs in many patients, although not to the extent seen in IDDM. It is not known whether this is due to ‘exhaustion’ of surviving f:lcells or to some independent process of damage. Islet amyloid deposits are commonly seen in NIDDM, destroying the cells and interfering with glucose and hormone transport. The amyloid deposits are derived  from islet amyloid polypeptide (IAPP), which may oppose the action of insulin, possibly explaining the insulin resistance that is also present. Obesity is present in 80% of patients with NIDDM, but insulin resistance may also be marked in lean individuals.

Impaired glucose tolerance

If an oral glucose tolerance test  is administered at random to a large population, 1-2% will be found to have unsuspected diabetes. A much larger group-S% or more (depending on the age, race and nutritional state of the population) – fall into an intermediate category now referred to as impaired glucose tolerance (IGT). The criteria for this category are given below. Follow-up shows that some (2-4% yearly) go on to develop diabetes, but that the abnormality does not progress in the majority. Obesity and lack of regular  physical exercise make progression to frank diabetes more likely. Classification is complicated by the poor reproducibility of the oral glucose tolerance test and the group is certainly heterogeneous. Some are obese, some have liver disease, and others are on medication that impairs glucose tolerance; individuals in this category have a risk  of cardiovascular disease that is twice that of people with normal glucose tolerance, but do not develop the specific microvascular complications of diabetes.

Tropical diabetes

A distinct variety of diabetes has been described. This is found only in developing countries on or near the equator. The following features have been described:
• Onset is before the age of 30 years.
• There is a history of severe malnutrition.
• There is insulin dependence, sometimes with severe but fluctuating insulin resistance.
• Ketoacidosis does not develop when insulin is withdrawn.
There are two main variants:
FIBROCALCULOUS PANCREATIC DIABETES. This is associated with exocrine pancreatic deficiency, pancreatic fibrosis (often leading to calcification) and the presence of stones in  the pancreatic duct. There may be a history of recurrent abdominal pain, and in 75% of cases there is evidence of pancreatic calcification on plain abdominal X-ray. Most populations in which this condition arises are subject to malnutrition and have  a diet based on cassava. Cyanates are present in the cassava root and may be a factor in the pancreatic damage.
PROTEIN-DEFICIENT PANCREATIC DIABETES. This form appears to be a direct consequence of malnutrition. The main differences from the fibrocalculous variant are that exocrine pancreatic function is unimpaired  and there is no evidence of pancreatic fibrosis or calcification. Abdominal pain is not a feature.
In both forms of tropical diabetes, insulin secretion is preserved, although impaired; this is the likely explanation for the observed resistance to ketosis. After an overnight fast, 75 g of glucose is taken in 250-350 ml of water. Blood samples are taken in the fasting state and 2 hours after the glucose has been given.
A specific enzymatic glucose assaymust be used. Note:
The concentration of glucose measured in plasma is 10% greater than that of whole blood.


This is present when the fasting blood glucose is over 6.7 mmol litre-I and/or when the 2-hour value is over 10 mmol litre’ Corresponding values for plasma glucose are 7.8 mmol litre-I and 11.1 mmol litre:

Impaired glucose tolerance

This is present when the fasting blood glucose is below 6.7 mmol litre ” and when the 2-hour value is between 6.7 and 10 mmol litre I. Corresponding values for plasma glucose are 7.8 and 11.1 mmol litre-I. Impaired glucose tolerance can only be diagnosed using the oral glucose tolerance test. Intermediate sampling times (e.g. 30 min and 60 min) are not needed for the diagnosis of diabetes by WHO criteria. However, simultaneous blood and urine glucose measurements can be used to define a low renal threshold for glucose. Diabetes can usually be diagnosed on the basis of fasting or random blood glucose measurements (see text). The glucose tolerance test should be reserved for borderline casesonly.


Acute presentation

Young people often present with a brief 2-4 week history and report the classic triad of symptoms:
POLYURIA, due to the osmotic diuresis that results when blood glucose levels exceed the renal threshold THIRST, due to the resulting loss of fluid and electrolytes WEIGHT LOSS, due to fluid depletion and the accelerated breakdown of fat and muscle secondary to insulin deficiency. Ketoacidosis may be the presenting feature if these early symptoms are not recognized and treated.

Subacute presentation

The clinical onset may be over several months, particularly in older patients. Thirst, polyuria and weight loss are usual features but medical attention is sought for such symptoms as lack of energy, visual blurring due to glucose-  induced changes in refraction, or pruritus vulvae or balanitis due to Candida infection .

Complications may be the presenting feature. These include:
• Staphylococcal skin infections
• Retinopathy noted during a visit to the optician
• A polyneuropathy causing tingling and numbness in the feet
• Impotence
• Arterial disease, resulting in myocardial infarction or peripheral gangrene

Asymptomatic diabetes

Glycosuria or a raised blood glucose may be detected on routine examination (e.g. for insurance purposes) in individuals who have no symptoms of ill health.

Physical examination

This is often unrewarding in younger patients, but evidence of weight loss and dehydration may be present, and the breath may smell of ketones. Older patients may present with established complications, and the presence of the characteristic retinopathy is diagnostic of diabetes.


The diagnosis is usually simple. Blood glucose is so closely controlled by the body that even small deviations become important.
1 In symptomatic patients, a single elevated blood glucose, measured by a reliable method, indicates diabetes.
2 In asymptomatic or mildly symptomatic patients, the diagnosis is made on:
(a) One, preferably two, fasting venous blood glucose levels above 6.7 mmol litre-I (120 mg dI-I); the equivalent venous plasma level is 7.8 mmol litre”” (140 mg dI-I), or (b) One, preferably two, random values above 10 mmol litre-I (180 mg dr ‘) in venous whole blood or ILl mmol litre-I (200 mg dI-l) in venous plasma.
3 A glucose tolerance test (GTT) is unnecessary when thecriteria above are satisfied, and should be reserved for  true borderline cases.
4 Glycosuria is measured using sensitive glucose-specific dipstick methods. Glycosuria is not diagnostic of diabetes  but indicates the need for further investigation. About 1% of the population have renal glycosuria. This is an inherited low renal threshold for glucose, transmitted either as a Mendelian dominant or recessive trait.

Other investigations

No further tests are needed to diagnose diabetes. Other routine investigations include screening the urine for proteinuria, a full blood count, urea and electrolytes, liver biochemistry and a fasting blood sample for cholesterol and triglycerides. The latter test is useful to exclude an associated hyperlipidaemia but should only be performed after blood glucose has been brought under control. It is important to remember that diabetes may be secondary to other conditions, may be precipitated by underlying illness and be associated with autoimmune disease or hyperlipidaemia. Hypertension is present in one-third of European patients with NIDDM and 50% of Afro-Caribbeans.

Diameties Mellitus and Other Desorder of Metabolism

Diabetes Mellitus


Diabetes mellitus is a group of metabolic disorders characterized by chronic hyperglycaemia due to relative insulin deficiency, or resistance or both. It is common and affects approximately 30 million people worldwide. Diabetes is usually irreversible and, although patients can have a reasonably normal life-style, its late complications result in reduced life expectancy and considerable uptake of health resources. Macrovascular disease leads to an increased prevalence of coronary artery disease, peripheral vascular disease and stroke, while microvascular damage results in diabetic retinopathy and contributes to nephropathy.

Insulin secretion

Insulin is the key hormone involved in the storage and controlled release within the body of the chemical energy available from food. It is synthesized in the f3 cells of the pancreatic islets in the form of proinsulin, which is stored in secretory granules close to the cell membrane. A biochemically inert peptide fragment known as connecting (C) peptide breaks off from proinsulin in the secretory process, so that equimolar quantities of insulin and Cpeptide are released into the circulation. Insulin enters the portal circulation and is carried to the liver, its prime target organ. About 50% of secreted insulin is extracted and degraded in the liver; the residue is broken down by the kidney. C-peptide is only partially extracted by the liver (and hence provides a useful index of the rate of insulin secretion), but is mainly degraded by the kidney. An outline of glucose metabolism Blood glucose levels are closely regulated in health and rarely stray outside the range of 3.5-8.0 mmollitre -I (63- 44 mg dr ‘), despite the varying demands offood, fasting and exercise. The principal organ of glucose homeostasis is the liver, which absorbs and stores glucose (as glycogen) in the postabsorptive state and releases it into the circulation between meals to match the rate of glucose utilization by peripheral tissues. The liver also manufactures glucose (6 carbons) from 3-carbon molecules derived from breakdown of fat and protein by the process of gluconeogenesis.
GLUCOSE PRODUCTION. About 200 g of glucose is produced and utilized each day. More than 90% is derived from the liver, three-quarters from glycogen and onequarter from gluconeogenesis. The remaining 5-10% derives from renal gluconeogenesis.
GLUCOSE UTILIZATION. The brain is the major consumer of glucose. Its requirement is 1 mg kg-I body weight per minute, or 100 g daily in a 70 kg man. Glucose uptake by the brain is obligatory and is not dependent on insulin, and the glucose used is oxidized to carbon dioxide and water.
Other tissues, such as muscle and fat, are facultative glucose consumers. The effect of insulin peaks associated with meals is to lower the threshold for glucose entry into cells; at other times, energy requirements are largely met by fatty-acid oxidation. Glucose taken up by muscle is stored as glycogen or broken down to lactate, which reenters the circulation and becomes an important substrate for hepatic gluconeogenesis. Glucose is used by fat tissue as a source of energy and as a substrate for triglyceride synthesis; lipolysis releases fatty acids from triglyceride together with glycerol, another substrate for hepatic gluconeogenesis.
HORMONAL REGULATION. Insulin is the major regulator of intermediary metabolism, although its actions are modified in important respects by other hormones. Dose-response curves for the production and utilization of glucose.
At low insulin levels, glucose production is maximal and utilization is minimal; at high levels the situation is reversed. At intermediate plasma insulin levels of 40- 50 mU litre-I, hepatic glucose production is largely suppressed but peripheral utilization remains low. This observation forms the theoretical basis for the low-dose insulin regimen used to treat diabetic ketoacidosis.
The effect of counter-regulatory hormones (glucagon, adrenaline, cortisol and growth hormone) is to shift the dose-response curves to the right, resulting in greater production of glucose and less utilization for a given level of insulin.

The insulin receptor is a glycoprotein (400 kDa) which straddles the cell membrane of many target cells. It consists of a dimer with two ex subunits, which include the binding sites for insulin, and two f3 subunits, which traverse the cell membrane and initiate at least some of the intracellular actions of insulin. The DNA sequence coding for the receptor has been isolated and sequenced and is located on the short arm of chromosome 19.
Insulin molecules bind to these receptors forming complex that promotes glucose uptake. This insulinreceptor complex is internalized by the cell with subsequent degradation of insulin and recycling of the receptor to the cell surface.


Diabetes may be primary or secondary.
Although insulin-dependent diabetes mellitus (IDDM, type I diabetes) and non-insulin-dependent diabetes mellitus (NIDDM, type II diabetes) represent two distinct diseases from the epidemiological point of view, clinical distinction may sometimes be difficult. The two disease processes should, in clinical terms, be visualized as opposite ends of a continuous spectrum.
Insulin-dependent diabetes mellitus (type I diabetes)


Approximately one person in 300 in the UK is treated with insulin, but some of these would be considered to have NIDDM by the criteria shown in Table 17.2. IDDM is most common in populations of European extraction, and within Europe there is a marked increase in incidence as one moves north. The highest incidence occurs in northern Scandinavia, but there is an unexplained hotspot in the island of Sardinia which has the second highest rate in the world. The frequency of IDDM in various countries.
The incidence in childhood is maximal at 10-13 years of age. Presentation is more common in the spring and autumn than in the summer, and it has been suggested that this might be related to the greater prevalence of viral infections at these times. The incidence of mOM is rising, in Europe, with an approximate doubling over the past 20-30 years.

Liver disease
Pancreatic disease
Cystic fibrosis
Chronic pancreatitis
Tropical diabetes
Carcinoma of the pancreas
Endocrine disease
Cushing’s syndrome
Drug-induced disease
Thiazide diuretics
Corticosteroid therapy
Insulin-receptor abnormalities
Congenital lipodystrophy
Acanthosis nigricans
Genetic syndromes, e.g. Friedreich’s ataxia, myotonic