Mark Evans
Synopsis
Diabetes mellitus affects at least 2% of many national populations. Its successful management requires close collaboration between the patient and health-care professionals.
• History of pharmacological treatment of diabetes.
• Insulins in current use (including choice, formulations, adverse effects, hypoglycaemia, insulin resistance).
• Oral antidiabetes drugs.
• Treatment of diabetes mellitus.
• Diabetic ketoacidosis.
• Surgery in diabetic patients.
• Obesity and overweight.
Diabetes overview
Diabetes is best regarded as a group of related conditions in which blood glucose levels tend to rise. Diabetes is common and increasing. In large part, this global increase in diabetes may be related to increased levels of obesity. Diabetes can lead to serious medical complications – blindness from retinopathy, renal failure, gangrene and limb amputation, cardiovascular disease and premature death. Aggressive therapy with a combination of pharmacological therapies aimed at lowering blood glucose and blood pressure and optimising lipids can reduce risk of these complications.
History of insulin therapy in diabetes
Diabetes was known to ancient Greek medicine with the description of ‘a melting of the flesh and limbs into urine … the patients never stop making water but the flow is incessant … their mouth becomes parched and their body dry’.1
Insulin (as pancreatic islet cell extract) was first administered to a 14-year-old insulin-deficient patient on 11 January 1922 in Toronto, Canada. R.D. Lawrence, an adult sufferer from diabetes who developed the disease in 1920 and who, because of insulin, lived until 1968, has told how:
Many doctors, after they have developed a disease, take up the speciality in it … But that was not so with me. I was studying for surgery when diabetes took me up. The great book of Joslin said that by starving you might live four years with luck. [He went to Italy and, whilst his health was declining there, he received a letter from a biochemist friend which said] there was something called ‘insulin’ appearing with a good name in Canada, what about going there and getting it. I said ‘No thank you; I've tried too many quackeries for diabetes; I'll wait and see’. Then I got peripheral neuritis … So when [the friend] cabled me and said, ‘I've got insulin – it works – come back quick’, I responded, arrived at King's College Hospital, London, and went to the laboratory as soon as it opened … It was all experimental for [neither of us] knew a thing about it … So we decided to have 20 units a nice round figure. I had a nice breakfast. I had bacon and eggs and toast made on the Bunsen. I hadn't eaten bread for months and months … by 3 o'clock in the afternoon my urine was quite sugar free. That hadn't happened for many months. So we gave a cheer for Banting and Best.2
But at 4 pm I had a terrible shaky feeling and a terrible sweat and hunger pain. That was my first experience of hypoglycaemia. We remembered that Banting and Best had described an overdose of insulin in dogs. So I had some sugar and a biscuit and soon got quite well, thank you.3
Diabetes mellitus is classified broadly as:
• Type 1 (formerly, insulin dependent diabetes mellitus, IDDM) which typically occurs in younger people who cannot secrete insulin.
• Type 2 (formerly, non-insulin dependent diabetes mellitus, NIDDM), which usually occurs in older people who are typically (although not always) obese. Type 2 diabetes is best thought of as a group of conditions characterised by a variable combination of reduced insulin secretion and resistance to insulin's blood glucose lowering action.
• Other causes: including gestational diabetes, disease processes affecting the liver or pancreas such as cystic fibrosis causing a ‘secondary’ diabetes, monogenic forms (maturity onset diabetes of the young, MODY).
These terms and abbreviations are used in this chapter.
Sources of insulin
Insulin is synthesised and stored (bound to zinc) in granules in the β-islet cells of the pancreas. Daily secretion typically amounts to 30–40 units, which is about 25% of total pancreatic insulin content. The principal factor that evokes insulin secretion is a high blood glucose concentration. In addition, insulin release following oral intake of carbohydrate is facilitated by the action of ‘incretin factors’ such as GLP-1 (see later) released from specialised neuroendocrine cells in the small bowel.
Insulin is a polypeptide with two peptide chains (A chain, 21 amino acids; B chain, 30) linked by two disulphide bridges. The basic structure having metabolic activity is common to all mammalian species but there are minor species differences:
• Bovine insulin differs from human insulin by three amino acids.
• Porcine insulin differs from human by only one amino acid.
• Human insulin is made either by enzyme modification of porcine insulin, or by using recombinant DNA to synthesise the pro-insulin precursor molecule for insulin. This is done by artificially introducing the DNA into either Escherichia coli or yeast.4
• Insulin analogues are now widely used and have modifications introduced to the A and/or B chains, which result in more rapid onset and offset of action (rapidly acting analogues), or slower offset (long acting analogues) than naturally occurring insulin.
Insulin receptors
Insulin receptors (comprising 2 α and 2 β subunits) are present on the surface of target cells such as liver, muscle and fat. Insulin binding results in tyrosine autophosphorylation of the β subunit. This then phosphorylates other substrates so that a signalling cascade is initiated and biological responses ensue. Downstream effects of stimulation of the insulin receptor include both immediate/short-term actions (for example translocation of the glucose transporter GLUT4 to the surface of target cell) and longer-term actions (for example increased expression of glucokinase and reduced expression of gluconeogenic and ketogenic enzymes in the liver).
Actions of insulin
• Reduction in blood glucose is due to increased glucose uptake in peripheral tissues (which oxidise glucose or convert into glycogen or fat), and a reduction in hepatic output of glucose (diminished breakdown/increased synthesis of glycogen and diminished gluconeogenesis).
• Other metabolic effects. Although, therapeutically, insulin is thought of as a blood glucose lowering hormone, it has a number of other cellular actions. Insulin is an anabolic hormone, enhancing protein synthesis (which has resulted in cases of misuse by bodybuilders). It also inhibits both breakdown of fats (lipolysis) and ketogenesis. Insulin also has actions on electrolytes, stimulating potassium uptake into cells and renal sodium retention (anti-natriuretic effect). Within brain, insulin may have actions to stimulate memory and act as a nutritional signal to help control appetite/food intake.
Uses
• Diabetes mellitus is the main indication.
• Insulin promotes the passage of potassium into cells by stimulating cell surface Na/K ATPase action, and this effect is utilised to correct hyperkalaemia (see p. 458).
• Insulin-induced hypoglycaemia can also be used as a stress test of anterior pituitary function (growth hormone and corticotropin and thus cortisol are released).
Pharmacokinetics
In health, insulin is secreted by the pancreas, enters the portal vein and passes straight to the liver, where half of it is taken up. The rest enters and is distributed in the systemic circulation so that its concentration (in fasting subjects) is only about 15% of that entering the liver. Insulin is released continuously and rhythmically from the healthy pancreas with additional increases following carbohydrate ingestion. As described below, modern insulin regimens in diabetes aim to match this pattern as far as possible.
In contrast to the natural pancreatic release, when insulin is injected subcutaneously during the treatment of diabetes, it enters the systemic circulation so that both liver and other peripheral organs receive the same concentration. It is inactivated in the liver and kidney; about 10% appears in the urine. The plasma t½ is only 5 min although clearance of ‘tissue’ insulin levels lags behind this; this is noteworthy when stopping intravenous insulin infusions as it may take 60 min for effects to wear off.
Most commonly, insulin is self-delivered by patients using either a syringe with a fixed needle (after drawing up insulin from a vial) or an insulin pen device (supplied as a preloaded disposable pen or with replaceable cartridges). Within hospital, soluble insulin may be delivered by intravenous infusion. Typically, 50 units soluble insulin is dissolved in 50 mL isotonic saline (i.e. insulin concentration 1 unit/mL).
An alternative and increasingly popular method for delivering insulin, especially in type 1 diabetes, is for patients to use continuous subcutaneous insulin infusion devices (‘insulin pumps’). These small cellphone-size personal devices provide a continuous basal delivery of soluble insulin (usually analogue, see below) with an additional insulin bolus when needed to cover meals or to correct high blood glucose values. Insulin pumps have become more sophisticated over the last decade, with patients able to set multiple pre-programmed basal insulin rates, and/or temporary infusion rates for such things as exercise or illness. Most pumps now have inbuilt software to calculate bolus doses from blood glucose/carbohydrate data. Some of the currently available insulin pumps link to subcutaneous continuous glucose monitors and, excitingly, the expectation is that this hardware will allow the development of an ‘artifical pancreas’ with insulin delivery controlled partially or totally by real-time glucose sensing.
Preparations of insulin (Table 36.1)
Dosage is measured in international units standardised by chemical assay. There are three major factors:
• Strength (concentration).
• Source (human, porcine, bovine).
• Formulation (short acting vs delayed action)
Table 36.1 Insulin preparations
Broadly speaking, four different types of insulin with differing time-courses of action are available for treating diabetes (illustrated in Fig. 36.1):
1. Short duration of action (and rapid onset). Soluble insulin (also called neutral or regular insulin). The most recent additions to this class of insulin, lispro, aspart and glulisine, are modified human insulins with changes in the B chain resulting in more rapid absorption after subcutaneous injection and thus a faster onset and shorter duration of action.
2. Intermediate duration of action (and slower onset). Preparations in which the insulin has been modified physically by combination with protamine or zinc to give an amorphous or crystalline suspension; this is given subcutaneously and slowly dissociates to release insulin in its soluble form. Isophane (NPH) insulin, a suspension with protamine, is still widely used. Insulin Zinc Suspensions (amorphous or a mixture of amorphous and crystalline) are now rarely used.
3. Longer duration of action. Newer analogues glargine and detemir (Fig. 36.2) have become widely used, especially in type 1 diabetes. Small changes in the amino acid structure of glargine result in a significant slowing of absorption from subcutaneous depots. In contrast, detemir owes its protracted action to fatty acylation. After absorption, detemir is thus bound to circulating albumin which delays its action.
4. A biphasic mixture of soluble or short acting analogue insulin with isophane insulin.
Fig. 36.1 Approximate pharmacokinetic profiles of human insulin and insulin analogues. The relative duration of action of the various forms of insulin is shown. The duration will vary widely both between and within persons.
(From Hirsch I B 2005 Insulin analogues. New England Journal of Medicine 352:174–183.)
Fig. 36.2 Amino acid alterations in insulin analogues. The diagram shows the structure of native insulin, and the modifications of this structure in a number of commercially available alternatives.
Notes for prescribing insulin
Allergy
to purified or analogue insulins is very rare.
Antibodies
to insulin do occur, but are largely of no clinical significance.
Compatibility
Soluble insulin may be mixed in the syringe with insulin zinc suspensions (amorphous, crystalline) and with isophane and mixed (biphasic) insulin, and used at once. Long acting analogue insulins, and protamine insulin suspensions, should not be mixed in a syringe with short acting insulins.
Intravenous insulin
Only soluble (neutral) insulin should be used. Analogue and regular insulin have identical action profiles when given i.v. although the latter tends to be more widely used.
The standard strength
of insulin preparations is 100 units/mL (U100). Solutions of 40 and 80 units/mL remain available in some countries, and health-care providers should be aware of this. Biological standardisation of insulin has been replaced by physicochemical methods (high-performance liquid chromatography, HPLC).
Choice of insulin regimen
There are three common regimens incorporating the insulin types described above for patients requiring insulin:
1. ‘Basal bolus’ therapy: multiple injections of short acting insulin are given during the day to mimic prandial secretion of insulin by the pancreas, combined with once or twice daily intermediate or long acting insulin to provide the background insulin. This approach aims to mimic the non-diabetic pattern of insulin release. The total insulin dose is usually apportioned to be 40–60% background and 40–60% prandial.
When choosing the short acting insulin in a basal bolus regimen, soluble insulin is given 30 min before meals. Short acting analogues may be given immediately before, during or even after the meal, although recent data suggest that even these insulins may be more effective if given 15 min prior to eating. The more rapid waning of action profile also means that the risk of hypoglycaemic reactions before the next meal may be lower with the analogues. For choice of background insulin, long acting analogues may give less risk of nocturnal hypoglycaemia than NPH insulin (see Fig. 36.1) although NPH insulins offer greater flexibility if patients need to change background insulin from day to day (as with some sportsmen or pregnant women, for example).
Insulin pump therapy uses the same principles as basal bolus insulin but uses only fast acting (usually analogue) insulin. In this case, the ‘background’ action comes from the fact that insulin is delivered continuously, analogous to insulin release from the non-diabetic pancreas.
2. Twice daily therapy involves two injections of biphasic insulin. Although less ‘physiological’ than basal bolus, it is simpler, with fewer insulin injections. The available mixtures are listed in Table 36.1. The most commonly used is 30:70 (soluble: NPH). Typically half to two-thirds of the daily dose may be given in the morning before breakfast and half to one-third before the evening meal. A combination of biphasic insulin with breakfast and fast acting insulin with evening meal and bedtime background insulin may be useful in some children with type 1 diabetes to avoid having to inject insulin at school.
3. Background or prandial insulin alone may be sufficient in type 2 diabetes when patients progress from oral therapy on to insulin. In this situation, oral therapy is usually continued in combination with insulin.
Dose and injection technique
A typical insulin-deficient patient with type 1 diabetes needs 0.5–0.8 units/kg insulin per day with approximately 50% as background. Increasingly, patients with type 1 diabetes are not being prescribed fixed insulin doses. Instead patients are being trained in how to self-adjust insulin doses, to allow for factors which will influence how much insulin is needed: meals with differing carbohydrate contents, digesting and skipping meals, exercise/activity, illness/stress, alcohol, travel, menstrual cycle. These same principles apply to insulin delivered by an insulin pump although many patients require lower total insulin doses by this route.
Initial treatment dose for a patient with type 1 diabetes, without ketoacidosis, is usually 0.3 units/kg daily. This initial management is aimed at introducing patients to regular insulin injections and blood glucose testing and aiming to tighten glycaemic control gradually over the first few weeks/months. Some patients with type 1 diabetes may have a significant residual insulin secretory capacity and may require no insulin for some months after diagnosis, often termed the ‘honeymoon’ period. Others may be started initially on low doses of either background insulin alone or prandial insulin, depending on their clinical status and whether they have any residual endogenous insulin secretion at diagnosis/presentation.
For type 2 diabetes, glycaemic targets have become lower over the last decade so that increasing numbers are treated with insulin. Although dosing calculators have been used, particularly in some clinical trials, in practice patients are often started on low doses of insulin using a simple regimen and then the dose/regimen is built up as indicated by blood glucose response. Most of these patients are insulin resistant and a useful therapeutic strategy is to combine oral insulin-sensitising therapy with metformin or pioglitazone (see below) with injected insulin. Severe insulin resistance merits specialist investigation for a possible underpinning cause.
Injection technique has pharmacokinetic consequences according to whether the insulin is delivered into the subcutaneous tissue or (inadvertently) into muscle and patients should standardise their technique. The introduction of a range of needles of appropriate length and pen-shaped injectors has enabled patients to inject perpendicularly to the skin without risk of intramuscular injection. The absorption of insulin is as much as 50% more rapid from shallow intramuscular injection. Clearly, factors such as heat or exercise that alter skin or muscle blood flow can markedly alter the rate of insulin absorption.
Sites of injection should be rotated to minimise local complications (lipodystrophy). Absorption is faster from arm and abdomen than it is from thigh and buttock.
Adverse effects of insulin
Hypoglycaemia
Hypoglycaemia is the main adverse effect of the therapeutic use of insulin. It occurs with excess insulin dosing. Common causes are misjudging or missing meals, activity/exercise and alcohol. Hypoglycaemia is problematic because the brain relies largely, if not exclusively, on circulating glucose as its source of fuel. A significant fall in blood glucose can result in impaired cognition, lethargy, coma, convulsions and perhaps even death (hypoglycaemia was implicated in one series in 4% of deaths aged less than 50 years in patients with type 1 diabetes). Hypoglycaemia is a major factor for insulin-treated patients, with fear of hypoglycaemia being rated as highly as fear of other complications of diabetes such as blindness or limb amputation. Hypoglycaemia is a particular problem for some patients who lose symptomatic awareness of (and associated counterregulatory neurohumoral defences against) hypoglycaemia.
When human insulin first became available, a number of patients reported that they had less symptomatic awareness of hypoglycaemic episodes. Although the bulk of the subsequent scientific studies examining this failed to detect any significant differences in responses to hypoglycaemia between human and animal insulins, the possibility remains that some patients do react differently and a small number of patients still prefer to use porcine insulin. In practice, the debate about human vs animal insulin has become less topical as non-human analogue insulins are being increasingly used in routine clinical practice.
Prevention of hypoglycaemia depends largely upon patient education, but regular mild episodes of hypoglycaemia are an almost unavoidable aspect of intensive glycaemic control, at least with currently available insulin replacement regimens. Patients should be vigilant, particularly if they have reduced symptomatic awareness of hypoglycaemia, carry rapid acting carbohydrates with them and monitor blood glucose regularly, especially with exercise and before driving a motor vehicle.
Treatment of hypoglycaemia is to give 20 g of rapidly acting carbohydrates by mouth (e.g. dextrose tablets, fruit juice or glucose drinks) if the patient is not cognitively obtunded, repeated after 10 min if needed. Where the conscious level is impaired, rescue needs to be non-oral therapy with either i.v. glucose (dextrose) or glucagon. For i.v. glucose, current advice is to avoid using 50% dextrose which is irritant, especially if extravasation occurs. Administration of 50–100 mL of 20% glucose (i.e. 10–20 g), is less thrombogenic. Glucagon (t½ 4 min) is a polypeptide hormone (29 amino acids) from the β-islet cells of the pancreas. It is released in response to hypoglycaemia from the non-diabetic pancreas (although not in type 1 diabetes for reasons that are unclear) and is a physiological regulator of insulin effect, acting by causing the release of liver glycogen as glucose. Glucagon is used as a ‘stopgap’ treatment for insulin-induced hypoglycaemia although is ineffective in prolonged or repeated hypoglycaemia where hepatic glycogen will be exhausted. The main advantage of glucagon is that is available in kits for home use so that 1.0 mg s.c. or i.m. can be useful where rescue is needed by parents/carers/partners without waiting for paramedic assistance.
The response to rescue is usually rapid. After initial therapy, the patient should be given a snack containing slowly absorbable ‘starchy’ carbohydrate to avoid relapse. The patient's treatment regimen should also be carefully reviewed with appropriate educational input. In particular, it is useful to ask whether this is part of a pattern of repeated episodes or a ‘one-off’ event with a clear precipitant.
After large overdoses of insulin (particularly long acting) or sulfonylurea, 20% glucose may be needed by continuous i.v. infusion for hours or days. With very large overdoses, for example where several hundred units have been administered to self-harm, it may be possible surgically to excise the depot of insulin from the injection site if it can be clearly identified. After prolonged hypoglycaemia, cerebral oedema may occur. Full recovery of cognitive function generally lags behind restoration of blood glucose but if the patient does not respond clinically to restoration of blood glucose within 30 min, cerebral oedema and i.v. dexamethasone therapy should be considered. Although the brain appears to be more resilient to hypoglycaemia than to other insults such as anoxia or trauma, very severe and prolonged hypoglycaemia can undoubtedly result in permanent brain damage.
Lipohypertrophy may occur if an injection site is repeatedly used, because of the local anabolic effects of insulin. Aesthetics aside, lipohypertrophy is a practical issue as insulin absorption from injection into areas of fatty hypertrophy becomes more variable and unpredictable, resulting in both hyperglycaemia and hypoglycaemia. Lipoatrophy at injection sites is rare (but still occurs) with modern, purified insulins and is thought to be related to a local immune reaction to insulin. More generalised allergic reactions to insulin are fortunately rare. If either lipodystrophy or lipoatrophy are present, the site should be avoided.
Oral antidiabetes drugs
Oral antidiabetes drugs are either (i) secretagogue therapy to increase endogenous insulin release, or (ii) insulin sensitisers to reduce insulin resistance, or (iii) drugs aimed at modifying absorption of glucose.
(i) Insulin secretagogues
Sulfonamide derivatives (sulfonylureas) act to increase endogenous insulin secretion by blocking ATP-sensitive potassium channels on the β-islet cell plasma membrane. This results in the release of stored insulin in response to glucose. The discovery of sulfonylureas was serendipitous. In 1930 it was noted that sulfonamides could cause hypoglycaemia, and in 1942 severe hypoglycaemia was found in patients with typhoid fever during a therapeutic trial of sulfonamide. Sulfonylureas were introduced into clinical practice in 1954 and continue to be widely used in type 2 diabetes. Sulfonylureas are ineffective in totally insulin-deficient patients; successful therapy probably requires at least 30% of normal β-cell function to be present. Secondary failure (after months or years) occurs due to declining β-cell function.
Their main adverse effects are hypoglycaemia and weight gain. Hypoglycaemia can be severe and prolonged (for days), and may be fatal in 10% of cases, especially in the elderly and patients with heart failure in whom long-acting agents should be avoided. Erroneous alternative diagnoses such as stroke may be made. Sulfonamides, as expected, potentiate sulfonylureas both by direct action and by displacement from plasma proteins.
Several sulfonylureas are available (see also Table 36.2). Choice is determined by the duration of action as well as the patient's age and renal function, and unwanted effects. The long acting sulfonylureas, e.g. glibenclamide, are associated with a greater risk of hypoglycaemia; for this reason they should be avoided in the elderly, for whom shorter acting alternatives, such as gliclazide, are preferred. In patients with impaired renal function, gliclazide, glipizide and tolbutamide are preferred as they are not excreted by the kidney. Gliclazide is a commonly used second-generation agent. If the dose exceeds 80 mg, the drug should be taken twice daily before meals, or once daily if prescribed as a modified-release preparation. Glimepiride is designed to be used once daily and provokes less hypoglycaemia than glibenclamide.
Table 36.2 Principal oral antidiabetes drugs
Meglitinides
such as repaglinide (t½ 1 h) are short-acting oral hypoglycaemic agents that have not been widely used in clinical practice. Like sulfonylureas, they act by blockade of ATP-dependent potassium channels (Table 36.2). The shorter action profile of meglitinides compared with sulfonylureas should in theory reduce risk of hypoglycaemia.
Incretin analogues and mimetics
Exenatide and liraglutide are functional analogues of incretin, a glucagon-like peptide-1 (GLP-1), and naturally occurring peptide that enhances insulin secretion in response to rise in blood glucose. Aside from boosting insulin secretion, they have other actions which offer potential advantages over other insulin secretagogues in diabetes: reducing glucagon secretion, slowing gastric emptying and acting on brain GLP-1 receptors to reduce appetite. Animal studies suggest that there may be a stimulatory effect of GLP-1 analogues on beta cell mass although this has not yet been demonstrated reliably in humans. GLP-1 therapy is currently aimed at overweight patients with type 2 diabetes where sulfonylureas or insulin therapy may promote weight gain. Exenatide and liraglutide are administered subcutaneously twice or once daily although longer-acting analogues are likely to be available soon, allowing a regimen with injections no more often than once weekly. The main adverse effect is nausea, which may be sufficient to prevent use. Rarely, pancreatitis has been reported in patients using GLP-1 agonists.
Dipeptidyl peptidase-4 (DPP-4) inhibitors
(vildagliptin and sitagliptin) offer an alternative strategy for targeting the GLP-1 pathway, in this case by inhibiting breakdown of native GLP-1. The main advantage these agents offer over injectable GLP-1 agonists is that they are orally active but are generally less potent.
(ii) Insulin sensitisers
Biguanides
(see also Table 36.2) have been available since 1957. Metformin is now the only biguanide in use, and is a major agent in the management of type 2 diabetes. The most important physiological effect appears to be an increase in hepatic insulin sensitivity/reduction of hepatic glucose production. Recent studies have suggested that the intracellular target of metformin in the liver is the enzyme adenosine monophosphate-activated protein kinase (AMPK) system. AMPK is a conserved regulator of the cellular response to low energy, being activated when intra-cellular ATP levels decrease and AMP concentrations increase.5
Metformin (t½ 5 h) is taken with or after meals. Metformin can be used in combination with either insulin or other oral hypoglycaemic agents. Its chief use is in the obese patient with type 2 diabetes, either alone or in combination with a sulfonylurea or insulin. The drug is ineffective in the absence of insulin.
Minor adverse reactions are common, including nausea, diarrhoea and a metallic taste in the mouth. These symptoms are usually transient or subside after reduction of dose and can be minimised by building doses up slowly and ensuring that metformin is taken with or after food. A modified release preparation, Metformin MR, is reported to be better tolerated in some patients who suffer gastrointestinal side-effects with regular metformin.
More serious, but rare, is lactic acidosis. When this condition does occur, it is usually against the background of significant medical illnesses which tend to increase circulating lactic acid levels, particularly renal impairment, liver failure, or cardiogenic or septic shock. Metformin is therefore contraindicated in these conditions, including relatively mild renal impairment and use should be reviewed when plasma creatinine is > 130 mmol/L (or eGFR < 45 mL/min/1.73 m2) and stopped when plasma creatinine is > 150 mmol/L (or eGFR < 30 mL/min/1.73 m2).
Metformin should also be withdrawn temporarily before general anaesthesia and administration of iodine-containing contrast media, which might precipitate renal impairment. During pregnancy, metformin use is unlicensed and current advice is that it should be substituted by insulin. Lactic acidosis may require treatment with i.v. isotonic sodium bicarbonate.
Apart from diabetes, the insulin-sensitising effects of metformin may also be useful in polycystic ovary syndrome, a condition in which insulin resistance occurs and may contribute to the hyperandrogenism and consequent hirsutism and disordered menstrual cycles which characterise this condition.
Thiazolidinediones
Pioglitazone reduces peripheral insulin resistance, leading to a reduction of blood glucose concentration. This class of drugs stimulate the nuclear hormone receptor, peroxisome proliferator-activated receptor (PPARγ), which causes differentiation of adipocytes. The major action is to stimulate peripheral insulin sensitivity and thus glucose uptake in skeletal muscle, although the mechanism by which this ‘cross-talk’ from adipocytes to muscle occurs is unclear. They are slower to act than either metformin or sulfonylureas.
The main adverse effects of thiazolidinediones are weight gain (typically 3–4 kg of weight gain in the first year of use), fluid retention (with peripheral oedema in 3–4% of patients) and decreased bone density.
(iii) Agents which reduce glucose absorption
Acarbose
is an α-glucosidase inhibitor that reduces the digestion of complex carbohydrates and slows their absorption from the gut and thus reduces postprandial glycaemia. The usual dose is 50–300 mg/day. Adverse effects are common, mainly flatulence and diarrhoea, which lead to a high discontinuation rate. In high doses it may cause frank malabsorption. Acarbose may be combined with a sulfonylurea.
Dietary fibre and diabetes
The addition of gel-forming (soluble) but unabsorbable fibre (guar gum, a hydrocolloidal polysaccharide of galactose and mannose from seeds of the ‘cluster bean’) to the diet of diabetics reduces carbohydrate absorption and flattens the postprandial blood glucose curve. Reduced need for insulin and oral agents has been reported, but adequate amounts (taken with lots of water) are unpleasant (flatulence) and patient compliance is therefore poor.
Choice of oral antidiabetic drugs in type 2 diabetes
In general terms, a hierarchy of therapies exists for type 2 diabetes, progressing from diet and lifestyle alone (described later), through monotherapy with oral agents, combinations of oral therapies and then onto insulin/injection therapy either alone or in combination with oral treatment. The nature of ‘typical’ type 2 diabetes is that glucose intolerance tends to progress so that many patients will need to escalate therapy with time to avoid worsening glycaemia (and warning patients of this early after diagnosis helps avoid subsequent disappointment and demotivation). It is also worth emphasising that this ‘typical’ time-course is not universal. Analogous to type 1 diabetes, some patients may present with marked symptomatic hyperglycaemia requiring immediate insulin therapy (and indeed some of these may have an unrecognised late onset of type 1 diabetes).
The evidence base for the most effective strategies for using antidiabetic therapy continues to evolve and the following is UK NICE guidance set out as a guide for readers. Current advice is that metformin(where not contraindicated and if tolerated) is useful primary monotherapy for overweight patients and many who are not overweight. Sulfonylurea therapy is now less often used as first-line therapy but is an alternative to metformin or can be added in to dual therapy as needed. Thiazolidinediones can also be used in combination with the above. GLP-1 agonists may be considered where overweight is a consideration. The therapeutic target is to maintain the HbA1c (glycosylated haemoglobin) below 7.0% and insulin may eventually be required, either alone (see earlier section on insulin regimens) or in combination with metformin (and/or sulfonylurea or pioglitazone).
Diet and diabetes
Specialised diet and lifestyle advice is of paramount importance in managing diabetes. Patients should be allowed to follow their own preferences as far as is practicable. They should receive dietary advice on a high complex carbohydrate diet (approx. 65% of total calories) with low fat (< 30% of calories) and an emphasis on reduction in saturated fat in favour of mono- and polyunsaturates. Caloric intake may need to be restricted and patients encouraged to achieve an ideal body-weight. Although certain foods are marketed as ‘diabetic’, there are concerns about whether these may be low in glucose but high in calories. Diet should be high in fibre with plenty of fresh fruit and vegetables. Advice about alcohol intake and smoking should be given (and where appropriate, information about what to do when/after drinking alcohol because of the effects causing delayed hypoglycaemia). This should be combined with advice about activity/exercise levels.
As earlier indicated, increasing numbers of patients with type 1 diabetes are now taught how to count dietary carbohydrates, allowing them to adjust their insulin doses around dietary intake/activity/alcohol.
The advice above may need further modifying in those with ischaemic heart disease and/or established nephropathy.
Interactions with non-diabetes drugs
Some examples are listed below to show that the possibility of interactions of practical clinical importance is a real one. In general, whenever a patient with diabetes takes other drugs it is prudent to be on the watch for disturbance of glycaemic control.
β-adrenoceptor-blocking drugs may impair the sympathetically mediated (β2 receptor) release of glucose from the liver in response to hypoglycaemia and also reduce the adrenergically mediated symptoms of hypoglycaemia (except sweating). Insulin hypoglycaemia may thus be more prolonged and/or less noticeable. Ideally, a patient with diabetes needing a β-adrenoceptor blocker should be given a β1-selective member, e.g. bisoprolol.
Thiazide diuretics at a higher dose than those now generally used in hypertension can precipitate/worsen diabetes, probably by reducing insulin secretion.
Hepatic enzyme inducers may enhance the metabolism of sulfonylureas that are metabolised in the liver (tolbutamide). Cimetidine, an inhibitor of drug-metabolising enzymes, increases metformin plasma concentration and effect.
Monoamine oxidase inhibitors potentiate oral agents and perhaps also insulin. They can also reduce appetite and so upset control.
Interaction may occur with alcohol (hypoglycaemia with any antidiabetes drug).
Salicylates and fibrates can increase insulin sensitivity, resulting in lower blood glucose.
The action of sulfonylureas is intensified by heavy sulfonamide dosage, and some sulfonamides increase free tolbutamide concentrations, probably by competing for plasma protein-binding sites.
The use of glucagon as rescue therapy for hypoglycaemia is described above. Adrenaline/epinephrine raises the blood sugar concentration by mobilising liver and muscle glycogen (a β2-adrenoceptor effect), and suppressing secretion of insulin (an α-adrenoceptor effect). Hyperglycaemia may occur in patients with phaeochromocytoma, and is usually reversed by α-adrenoceptor blockade (see p. 383).
Adrenal steroids
either endogenous or exogenous, antagonise the actions of insulin. Although this effect is only slight with mineralocorticoids, glucocorticoid hormones increase gluconeogenesis and reduce glucose uptake and utilisation by the tissues. The therapeutic use of high dose glucocorticoids may precipitate frank diabetes in some patients or worsen blood glucose control in those with established diabetes. Steroid-induced diabetes often requires insulin rather than oral hypoglycaemic therapy, particularly for flexibility if steroid doses are being tapered down. Similarly, patients with Cushing's syndrome may also develop ‘secondary diabetes’ with a marked resistance to insulin. In contrast, patients with hypoadrenalism from Addison's disease, hypopituitarism, or following steroid withdrawal after prolonged glucocorticoid therapy may be abnormally sensitive to insulin action and prone to recurrent hypoglycaemia.
Growth hormone
antagonises the actions of insulin in the tissues. Like Addison's disease, growth hormone deficiency may cause increased insulin sensitivity and/or a tendency for hypoglycaemia. Acromegalic patients may develop insulin-resistant diabetes.
Oral contraceptives
can impair carbohydrate tolerance although the effects are usually relatively mild.
Thyroid hormone
excess may increase the requirements for insulin.
Drug-induced diabetes
Diazoxide
(see p. 401) is chemically similar to thiazide diuretics, but stimulates the ATP-dependent K+ channel that is blocked by the sulfonylureas. Although formerly used as an antihypertensive agent, its current use in therapeutics is confined to the rare indication of treating hypoglycaemia due to islet cell tumour (insulinoma). Adrenocortical steroids are also diabetogenic (see above).
Pregnancy and diabetes
During (and indeed before) pregnancy, close control of diabetes is critical. Ideally, pregnancy should be planned and women should be seen in a pre-conception clinic to optimise care. Glycaemic targets are tight, aiming for HbA1c values as close to the non-diabetic range (4.0–5.9%) as possible. Current practice is for women on oral hypoglycaemic agents who are planning, or starting, a pregnancy to change to insulin and continue on it throughout pregnancy. There is no definitive evidence that oral drugs are associated with fetal malformations. Other drugs (blood pressure and lipid lowering) should be reviewed and stopped or altered to agents judged safe in pregnancy as appropriate. Oral folic acid should be started.
Diabetes can present de novo during pregnancy. Although this is usually gestational diabetes which resolves after delivery, it is worth remembering that rarely type 2 or type 1 diabetes can present in pregnancy.
Risks of pregnancy in diabetes include an increased rate of fetal loss and malformations. Maternal hyperglycaemia can lead to fetal hyperglycaemia with consequent fetal islet cell hyperplasia, high birth-weight babies (leading to mechanical obstetric challenges) and postnatal hypoglycaemia.
Note that glycosuria is not a reliable guide of blood glucose values in pregnancy. The renal threshold for glucose (also of lactose) falls, so that glycosuria and lactosuria may occur in the presence of a normal blood glucose.
Insulin requirements increase steadily after the third month. Some women develop a marked intolerance of oral carbohydrates with a tendency for large postprandial rises in blood glucose. During labour, i.v. insulin infusion may be needed. Use of β2-adrenoceptor agonists and of dexamethasone (to prevent respiratory distress syndrome in the prematurely newborn) causes hyperglycaemia and increased insulin (and potassium) needs.
Of note, insulin requirements reduce immediately after delivery and may remain low during the following weeks, particularly with lactation/breast feeding.
Surgery in diabetic patients
Principles of management
• Surgery constitutes a major stress.
• Insulin needs will often increase with surgery.
The programme for control should be agreed between anaesthetist and physician whenever diabetic patients must undergo general anaesthesia or modify their diets. There are many different techniques that can give satisfactory results.
Type 1 diabetes
The guidelines below may also be useful for insulin-treated type 2 diabetes but suggested doses may need modifying if patients are insulin resistant with a large constitutive insulin requirement.
Elective major surgery
• Ideally admit to hospital the day before surgery and arrange surgery for the morning.
• Evening before surgery: give patient's usual insulin (consider 20% reduction in long-acting insulin if prone to hypoglycaemia).
• Day of operation: omit morning s.c. dose; set up i.v. infusion: glucose 5% + KCl 20 mmol/L, infuse at 100 mL/h; insulin should be infused by pump at an approximate rate of 2 units/h and adjusted according to a sliding scale.
• Modify regimen during and after surgery according to monitoring; insulin doses should be adjusted according to similar scale as that in Table 36.3.
• Stop i.v. infusion 1 h after first post-surgical s.c. soluble insulin (given when eating again).
• Insulin requirements may be high, 10–15 units/h, in cases of major surgery, serious infection, corticosteroid use, obesity.
Table 36.3 Sliding scale of insulin doses according to blood glucose concentrations (not for ketoacidosis)
|
Blood glucose (mmol/L) |
Infusion rate (mL/h = units/h for 50-mL syringe containing 50 units insulin) |
|
≥ 22.0 |
10.0 (and check pump and connections) |
|
19–21.9 |
8.0 |
|
16–18.9 |
6.0 |
|
12–15.9 |
4.0 |
|
8–11.9 |
2.0 |
|
4–7.9 |
1.0 |
|
< 3.9 |
0.5 (and increase glucose infusion) |
Minor surgery/ procedures
For example, simple dental extractions or endoscopy. For short, relatively non-stressful procedures, i.v. insulin is usually not needed. When nil by mouth, patients should omit fast acting insulin normally given for that meal time.
Emergency surgery
When a surgical emergency is complicated by diabetic ketosis, an attempt should be made to control the ketosis before surgery. Management during the operation will be similar to that for major surgery except that more insulin may be needed.
Type 2 diabetes
For minor procedures where diabetes is well controlled on oral hypoglycaemics, it should be possible simply to omit the oral hypoglycaemic agent on the morning of surgery. Otherwise it is safest to follow guidelines for type 1 diabetes above.
Diabetic ketoacidosis
The condition is discussed in detail in medical texts and only the more pharmacological aspects will be considered here.
The best way to consider ketoacidosis is as a severe and life threatening metabolic disorder resulting from a lack of insulin in which hyperglycaemia is present, rather than as a primary hyperglycaemic disorder. The patient with ketoacidosis often remains critically ill during treatment even after blood glucose is normalised. Patients are severely dehydrated and fluid resuscitation is a major priority. Insulin is needed not only to lower blood glucose, but also to suppress ketogenesis. The objective is to supply, as continuously as possible, a moderate amount of insulin.
Intravenous fluid
A patient with diabetic ketoacidosis may have a fluid deficit of above 5 L. Usual fluid replacement is isotonic saline and a typical regimen might be:
• 1 L in the first hour, followed by
• 2 L in 4 h, then
• 4 L in the next 24 h, watching for signs of fluid overload.
Note that fluid replacement itself causes a fall in blood glucose concentration both by dilution and also by restoring blood volume to perfuse skeletal muscle, a major insulin target tissue.
Soluble insulin
should be given by continuous i.v. infusion of a 1-unit/mL solution of insulin in isotonic sodium chloride. High doses are needed (e.g. 0.1 U/kg/h) and treatment needs to be prolonged beyond restoration of blood glucose to suppress ketogenesis. A reasonable rate of fall of glucose during treatment is 4–5.5 mmol/L (75–100 mg/100 mL) per hour. Once the blood glucose has fallen to 10–15 mM, i.v. dextrose infusion is started to prevent this continued high dose insulin resulting in hypoglycaemia (rather than reducing the insulin infusion rate, which will slow resolution of the metabolic derangement).
Potassium
Even if plasma potassium concentration is normal or high, patients have a substantial total body deficit, and the plasma level will fall briskly with i.v. fluids (dilution) and insulin, which draws potassium into cells within minutes. Potassium chloride should be added to the second and subsequent litres of fluid according to plasma potassium (provided the patient is passing urine).
• < 3.5 mmol/L: add 40 mmol/L.
• 3.5–4.5 mmol/L: add 30 mmol/L.
• 4.5–5.5 mmol/L: add 20 mmol/L.
• > 5.5 mmol/L: none.
Bicarbonate
(isotonic) may be considered for use only if plasma pH is < 7.0 although insulin and fluid resuscitation should rapidly correct acidosis.
Success in treatment
of diabetic ketoacidosis and its complications (hypokalaemia, aspiration of stomach contents, infection, shock, thromboembolism, cerebral oedema) depends on close, constant, informed supervision and repeated monitoring of clinical state and biochemical parameters.
Diabetic ketosis without acidosis
As with full-blown ketoacidosis, this may develop during intercurrent illness. Increasingly, patients with type 1 diabetes are given ‘sick day rules’ for how to manage this when appropriate out of hospital. This relies on the patient being fully conscious, not vomiting and being willing and able to drink fluids and follow instructions, including regular and repeated monitoring of blood glucose and ketone levels. In general, insulin doses are increased with additional injections of soluble insulin (10–20% of total daily insulin dose every 2 h). The rules include ‘bail out’ instructions on seeking admission if ketone and glucose levels are not resolving and/or patients start to vomit or fail to improve clinically.
Hyperosmolar diabetic coma
occurs chiefly in type 2 diabetics who fail to compensate for their continuing osmotic glucose diuresis. It is characterised by severe dehydration, a very high blood sugar level (> 33 mmol/L), and lack of ketosis and acidosis. Treatment is with isotonic (0.9%) saline, at half the rate recommended for ketoacidotic coma, and with less potassium than in severe ketoacidosis. Insulin requirements are less than in ketoacidosis, where the acidosis causes resistance to the actions of insulin, and should generally be half those shown in Table 36.3. Patients may be profoundly dehydrated and liable to thrombosis so that prophylactic low molecular weight heparin should be considered.
Preventing complications other than by glucose lowering
Diabetes is a condition not just of abnormal glucose but also of significantly increased cardiovascular risk. Indeed, most patients with both type 1 and type 2 diabetes succumb to either the macrovascular or microvascular complications – especially ischaemic heart disease and/or diabetic nephropathy.
Aggressive treatment of hypertension and hyperlipidaemia in addition to glycaemia is particularly important in patients with diabetes. For example, the landmark UK Prospective Diabetes Study (UKPDS) of type 2 diabetes confirmed that good glycaemic control and aggressive blood pressure reduction independently improve outcome.6,7 For every 1% reduction in haemoglobin A1c (HbA1c) there was a 21% reduction in diabetes-related deaths and a 37% reduction in microvascular disease. Of highest importance was the finding that effective blood pressure control – regardless of the type of antihypertensive drug – was more influential than glycaemic control in preventing macrovascular complications. Reduction of blood pressure in 758 patients to a mean of 144/82 mmHg achieved a 32% reduction in deaths related to diabetes and a 37% reduction in microvascular endpoints, compared with findings in 390 patients treated to a blood pressure of 154/87 mmHg.
Similarly, aggressive targeting of lipids reduces cardiovascular complications in diabetes. In the Heart Protection Study, addition of simvastatin 40 mg daily to the treatment of 4000 patients with diabetes reduced cardiovascular complications by 30%. Some guidelines have suggested that aspirin may be worth using in primary prevention in diabetes (i.e. in those who have not suffered a cardiovascular event) although current thinking is that the benefits are unproven.
Patients with evidence of diabetic nephropathy should receive either an angiotensin-converting enzyme (ACE) inhibitor or an angiotensin receptor antagonist; the evidence for the superiority of the latter in reducing progression to renal failure compared with other antihypertensive agents is particularly strong.8 Addition of an ACE inhibitor to other drugs may also improve overall outcome in patients with diabetes.9 In addition, diabetic nephropathy is independently associated with an increased risk of macrovascular disease so that aggressive lipid and blood pressure lowering therapy, as described above, should be employed.
Summary
• Diabetes mellitus is important in global terms because of its chronicity, and high incidence and frequency of major complications. It is generally divided into two kinds: type 1 (previously, insulin dependent diabetes mellitus) and type 2 (previously, non-insulin dependent diabetes, essentially an umbrella term for a group of conditions which are non-type 1).
• Type 1 diabetes is commoner in those with onset before age 30 whereas type 2 diabetes prevails in older patients. Increasingly, insulin therapy is required in type 2 diabetes when glycaemic control is not optimised by oral drugs.
• Insulin is usually self-administered subcutaneously to stable patients, with a variety of regimens which can be tailored to the needs of a particular patient. Modern practice in type 1 diabetes is to educate patients in flexible insulin dosing which is adjusted for differing meals, activity levels, etc.
• In the treatment of diabetic ketoacidosis, in the perioperative patient, and during inpatient management of the critically ill patient with diabetes, insulin is best given by intravenous infusion of the soluble form.
• Diet plays a major role in the treatment of type 2 diabetes, particularly where associated with obesity.
• If a drug is required in type 2 diabetes, metformin (a biguanide) is now widely used as first-line therapy, especially for the obese. Other monotherapy options include a sulfonylurea in the non-obese or a thiazolidinedione in patients intolerant of, or uncontrolled by, metformin or sulfonylurea. Many patients with type 2 diabetes will need treatment escalation with time to multiple combination therapy and/or insulin.
• Aggressive blood glucose lowering treatment of type 1 and type 2 diabetes reduces risk of microvascular complications. Close attention to associated risk factors, especially hyperlipidaemia and hypertension, is important in reducing risk of macrovascular disease.
Obesity and appetite control
Overweight and obesity are the commonest nutritional disorders in developed countries. Between 1991 and 1998 the incidence of obesity rose from 12.0% to 17.9% in the USA. Obesity predisposes to several chronic diseases including hypertension, hyperlipidaemia, diabetes mellitus, cardiovascular disease and osteoarthritis, and aspects of these are discussed in the relevant sections of this book.
Individuals whose body mass index10 (BMI) lies between 25 and 30 kg/m2 are considered overweight and those in whom it exceeds 30 kg/m2 are defined as obese. Management of the condition involves a variety of approaches from nutritional advice to lifestyle alteration, drugs and, where available and appropriate, bariatric surgery. In the UK, an evidence-based algorithm coordinates these.11 The present account concentrates on pharmacological interventions.
In general, drugs that have been used for obesity act either on the gastrointestinal tract, lowering nutrient absorption, or centrally, reducing food intake by decreasing appetite or increasing satiety (appetite suppressants). A number of pharmacological agents that have been marketed for obesity have been withdrawn because of concerns about safety. Currently, only one agent, orlistat, is available in the UK:
Orlistat
Orlistat is a pentanoic acid ester that binds to and inhibits gastric and pancreatic lipases; the resulting inhibition of their activity prevents the absorption of about 30% of dietary fat compared with a normal 5% loss. Weight loss is due to calorie loss but drug-related adverse effects also contribute by diminishing food intake. The drug is not absorbed from the alimentary tract.
Clinical trials have shown that patients who adhered to a low-calorie diet and took orlistat lost on average 9–10 kg after 1 year (compared with 6 kg in those taking placebo); in the following year those who remained on orlistat regained 1.5–3.0 kg (4–6 kg with placebo). Orlistat has found a place in the management of obesity in the UK but, not surprisingly, this is subject to stringent guidance from NICE, namely that it be initiated only in individuals with a BMI of 28 kg/m2 or more who also have cardiovascular risk factors, or 30 kg/m2 or more without such co-morbidity.
The dose is 120 mg, taken immediately before, during or 1 h after each main meal, up to three times daily. If a meal is missed, or contains no fat, the dose of orlistat should be omitted.
Treatment should be accompanied by counselling advice and proceed beyond 3 months only in those who have lost more than 5% of their initial weight, beyond 6 months in those who have lost more than 10%, should not normally exceed 1 year, and never more than 2 years.
Adverse effects
include flatulence and liquid, oily stools, leading to faecal urgency, abdominal and rectal pain. Symptoms may be reduced by adhering to a reduced-fat diet. Low plasma concentrations of the fat-soluble vitamins A, D and E have been found. Orlistat is contraindicated where there is chronic intestinal malabsorption or cholestasis.
Leptin
The adipocyte-derived hormone leptin (Greek: leptos, thin) has a limited role in therapeutics for patients with rare genetic defects in the leptin or leptin receptor genes. Leptin acts on the brain to reduce appetite. Most obese patients have raised plasma leptin concentrations, to which they have become relatively resistant. A small number of patients are genuinely deficient in leptin and therapy with recombinant leptin has had dramatic beneficial effects. Leptin may also be of benefit in lipodystrophic patients, a rare group of conditions in which a generalised or partial lack of adipocytes leads to marked metabolic abnormalities and diabetes.
Leptin acts to control satiety via the melanocortin system, predominantly in the basomedial hypothalamus, and a number of agents in development aim to target this.
Obesity and diabetes
Obesity is associated with type 2 diabetes and weight loss improves (and in rare cases ‘cures’) diabetes. Weighed against this is the challenge that certain pharmacologic therapies for diabetes, particularly sulfonylureas, thiazolidinediones and insulin, can promote weight gain. The patient and clinician may find themselves caught in a vicious cycle where dose escalation of these agents leads to weight gain, worsening glycaemic control, which in turn leads to further dose increments.
As for patients without diabetes, diet and exercise are critical factors but diabetes treatment may need to be adjusted for this (for example insulin reductions to avoid hypoglycaemia with increased exercise or reduced carbohydrate intake). Metformin, either alone or as adjuvant therapy, especially with insulin as an ‘insulin-sparing’ agent is useful. GLP-1 analogue therapy offers an alternative to insulin therapy. Ultimately, as for patients without diabetes, orlistat and/or bariatric surgery12 may be appropriate.
Guide to further reading
American Diabetes Association. Standards of medical care in diabetes. Diabetes Care. 2011;34(Suppl. 1):S11.
Bergenstal R.M., Tamborlane W.V., Ahmann A., et al. STAR 3 Study Group Effectiveness of sensor-augmented insulin-pump therapy in type 1 diabetes. N. Engl. J. Med.. 2010;363:311–320.
Chan J.L., Mantzoros C.S. Role of leptin in energy-deprivation states: normal human physiology and clinical implications for hypothalamic amenorrhoea and anorexia nervosa. Lancet. 2005;366:74–85.
Cushman W.C., Evans G.W., Byington R.P., et al. Effects of intensive blood-pressure control in type 2 diabetes mellitus. N. Engl. J. Med.. 2010;362:1575–1585.
Daneman D. Type 1 diabetes. Lancet. 2006;367:847–858.
Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) Study Research Group. Intensive diabetes treatment and cardiovascular disease in patients with type 1 diabetes. N. Engl. J. Med.. 2005;353:2643–2653.
Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) Research Group. Modern-day clinical course of type 1 diabetes mellitus after 30 years’ duration. Arch. Intern. Med.. 2009;169:1307–1316.
Dornhorst A. Insulinotropic meglitinide analogues. Lancet. 2001;358:1709–1716.
Drucker D.J., Nauck M.A. The incretin system: glucagon-like peptide-1 receptor agonists and dipeptidyl peptidase-4 inhibitors in type 2 diabetes. Lancet. 2006;368:1696–1705.
Eckel R.H., Grundy S.M., Zimmet P.Z. The metabolic syndrome. Lancet. 2005;365:1415–1428.
Gerstein H.C., Miller M.E., Genuth S., et al. Long-term effects of intensive glucose lowering on cardiovascular outcomes. N. Engl. J. Med.. 2011;364:818–828.
Ginsberg H.N., Elam M.B., Lovato L.C., et al. Effects of combination lipid therapy in type 2 diabetes mellitus. N. Engl. J. Med.. 2010;362:1563–1574.
Haslam D.W., James W.P.T. Obesity. Lancet. 2005;366:1197–1209.
Hirsch I.B. Insulin analogues. N. Engl. J. Med.. 2005;352:174–183.
Jellinger P.S. Focus on incretin-based therapies: targeting the core defects of type 2 diabetes. Postgrad. Med.. 2011;123:53–65.
Marshall S.M., Flyvbjerg A. Prevention and early detection of vascular complications of diabetes. Br. Med. J.. 2006;333:475–480.
Nathan D.M., Buse J.B., Davidson M.B., et al. Medical management of hyperglycemia in type 2 diabetes: a consensus algorithm for the initiation and adjustment of therapy: a consensus statement of the American Diabetes Association and the European Association for the Study of Diabetes. Diabetes Care. 2009;32:193.
Sherifali D., Nerenberg K., Pullenayegum E. The effect of oral antidiabetic agents on A1C levels: a systematic review and meta-analysis.. 2010;33:1859–1864.
Stumvoll M., Goldstein B.J., van Haeften T.W., et al. Type 2 diabetes: principles of pathogenesis and treatment. Lancet. 2005;365:1333–1346.
Wright J.R. From ugly fish to conquer death: J J R Macleod's fish insulin research, 1922–24. Lancet. 2002;359:1238–1242.
1 The Extant Works of Aretaeus, trans. Francis Adams (London 1856) p. 338 (quoted by Ackerknecht E H 1982 A short history of medicine. Johns Hopkins, Baltimore, pp. 71–72).
2 F G Banting and C H Best of Toronto, Canada (see also Journal of Laboratory and Clinical Medicine 1922; 7:251).
3 Abbreviated from Lawrence R D 1961 King's College Hospital Gazette 40:220. Transcript from a recorded after dinner talk to students’ Historical Society.
4 The three forms of human insulin have the same amino acid sequence, but are separately designated as insulin emp (Enzyme Modified Porcine), prb (Pro-insulin Recombinant in Bacteria) and pyr (Precursor insulin Yeast Recombinant). Although one of the incentives for introducing human insulin was avoidance of insulin antibody production, the allergies to older insulins were caused largely by impurities in the preparations, and are avoided equally well by using the highly purified, monocomponent porcine and bovine insulins.
5 The discovery of the AMPK response, and of other players in the pathway, has enabled experiments to be performed in which the hepatic response to metformin is selectively knocked out. In the mouse, at least, these experiments show that actions of metformin at other sites are of little importance.
6 UK Prospective Diabetes Study (UKPDS) Group 1998 Effect of intensive blood-glucose control with metformin on complications in overweight patients with Type 2 diabetes (UKPDS 34). Lancet 352:854–865.
7 UK Prospective Diabetes Study (UKPDS) Group 1998 Tight blood pressure control and risk of macrovascular and microvascular complications in Type 2 diabetes. British Medical Journal 317:703–713.
8 Three trials compared an angiotensin blocker with other blood pressure lowering drugs and found a 20% reduction in the proportion of patients in whom proteinuria worsened or serum creatinine concentration doubled during follow-up: (1) Parving H H, Lehnert H, Brochner-Mortensen J et al 2001 The effect of irbesartan on the development of diabetic nephropathy in patients with Type 2 diabetes. New England Journal of Medicine 345:870–878; (2) Brenner B M, Cooper M E, de Zeeuw D et al 2001 Effects of losartan on renal and cardiovascular outcomes in patients with Type 2 diabetes and nephropathy. New England Journal of Medicine 345:861–869; (3) Lewis E J, Hunsicker L G, Clarke W R et al 2001 Renoprotective effect of the angiotensin-receptor antagonist irbesartan in patients with nephropathy due to Type 2 diabetes. New England Journal of Medicine 345:851–860.
9 The HOPE study included patients with diabetes as one of its high-risk group of cardiovascular patients, in whom ramipril reduced further coronary heart disease endpoints by about 30%. Yusuf S, Sleight P, Pogue J et al 2000 Effects of an angiotensin converting enzyme inhibitor, ramipril, on cardiovascular events in high-risk patients. The Heart Outcomes Prevention Evaluation Study Investigators. New England Journal of Medicine 342:145–153.
10 The weight in kilograms divided by the square of the height in metres (kg/m2).
11 http://www.nice.org.uk/nicemedia/live/11000/30365/30365.pdf
12 Weight loss surgery: the procedures include reducing the size of the stomach by resection, gastric banding and gastric bypass.