Type 2 diabetes is a progressive condition characterised by initial insulin resistance followed by gradual loss of beta-cell insulin secretory ability. The UKPDS (UK Prospective Diabetes Study) demonstrated that no matter how type 2 diabetes is treated, there is a progressive increase in HbA1c (UKPDS Group, 1995). This means that oral antidiabetes drugs (OADs) become less effective over time, and eventually most people with type 2 diabetes need insulin to achieve or maintain their ideal HbA1c level (Turner et al, 1999). The UKPDS confirmed that glycaemic control of a level nearing that of people without diabetes reduces the risk of microvascular and macrovascular complications and mortality (UKPDS Group, 1998; Holman et al, 2008), and insulin therapy will therefore be necessary to achieve this in many cases. Box 1 provides some key facts and practical considerations relevant to insulin therapy in type 2 diabetes.
As insulin therapy is likely to be ultimately required in people with type 2 diabetes, it should be discussed early after diagnosis so that, when it is needed, it is not seen as failure of self-management or a punishment for non-adherence. In the author’s experience, people may be fearful of starting insulin as a result of previous experiences of older members of the family (for example, a grandmother using glass syringes with large needles, who started insulin after amputation), a fear of needles, concern about possible hypoglycaemia and weight gain, or the perception that they now have “serious diabetes” (in contrast to “mild diabetes” controlled by diet and tablets). These concerns need to be addressed early to avoid delay in starting insulin therapy when it is needed.
Unlike in type 1 diabetes, which is characterised by a complete lack of endogenous insulin, insulin therapy in type 2 diabetes does not completely replace, but instead supplements, the insulin still being produced by the beta-cells. How much insulin is required, and how many injections, will depend on a person’s remaining endogenous insulin production capacity and the extent of the progression of the condition. Although people with type 2 diabetes still produce some insulin, compared with people with type 1 diabetes, bigger doses of exogenous insulin are often required, as obesity and insulin resistance are common.
Insulin regimens in type 2 diabetes vary from a single daily injection of insulin in combination with OADs to multiple injection regimens that may involve four or more daily injections. In contrast, in type 1 diabetes, a multiple injection regimen or insulin pump therapy is usually used to mimic the physiological insulin profile of someone without diabetes.
Insulin therapy is supported by a number of outcome studies in type 2 diabetes (Ohkubo et al, 1995; UKPDS Group, 1998) and is the only current blood glucose-lowering therapy for which there is no maximum dose or limit to efficacy (Inzucchi et al, 2012). More recent data from studies such as ACCORD (Action to Control Cardiovascular Disease in Diabetes; Gerstein et al, 2008) and the VADT (Veterans Affairs Diabetes Trial; Duckworth et al, 2009) have raised some concerns among healthcare professionals regarding the possible dangers of intensive glucose lowering in people with type 2 diabetes, and in many areas of the UK, primary care teams will be involved in the initiation and intensification of insulin therapy for people with type 2 diabetes in an effort to tighten glycaemic control.
An understanding of the different types of insulin, the various insulin regimens and whether or not OAD therapy should be adjusted is therefore important.
History
Insulin is the oldest of the currently available medications for glycaemic control, with the most clinical experience. Its discovery in 1921 at the University of Toronto, Canada, led to the award of the Nobel Prize in Physiology or Medicine in 1923 for Frederick Banting and J Macleod, who shared the prize with Charles Best and James Collip. The famous experiment to “cure” diabetes with “isletin” in Marjorie the dog (who had her pancreas removed to induce diabetes) led Collip to comment:
“We have obtained from the pancreas of animals a mysterious something which when injected into totally diabetic dogs completely removes all the cardinal symptoms of the disease […] If the substance works on the human, it will be a great boon to Medicine” (Bliss, 1982).
The therapeutic use of insulin began with the treatment of a 14-year-old boy called Leonard Thompson in January 1922 (Bliss, 1982), and its role in the management of hyperglycaemia in type 1 diabetes is undisputed.
Insulin has also been used in the treatment of type 2 diabetes since the 1930s (Himsworth and Kerr, 1939). Indeed, until the 1940s, insulin was the only treatment available for both types of diabetes, at which point OADs were introduced in the form of the first generation of sulphonylureas. Metformin use followed in the late 1950s. The introduction of other OADs has meant that the role of insulin therapy in type 2 diabetes is less ubiquitous than in type 1 diabetes.
Mode of action
Insulin is a 51-amino acid polypeptide hormone that has an extensive and fundamental role in metabolism. It is secreted from pancreatic beta-cells in response to increases in blood glucose levels arising from the ingestion of carbohydrate-containing food, and has a number of effects on glucose homeostasis. A detailed description of all its physiological effects is beyond the scope of this article, but, notably, insulin promotes the uptake of glucose by the liver, muscle and adipose tissue, and stimulates the storage of glucose as glycogen in the liver and muscle.
As insulin is inactivated by gut enzymes, it is not suitable for oral administration, and is given by subcutaneous injection in most circumstances. Since the introduction of insulin therapy in the 1920s, a number of types of insulin preparation with different pharmacodynamic properties have been developed. These are considered in more detail in the “Types of insulin” section.
Indications and licence
While the exact wording of the therapeutic indications of different insulins varies, broadly speaking, the different insulin preparations are indicated for the treatment of diabetes where insulin is required for glucose homeostasis. Some insulins, particularly the newer ones, are indicated for treatment in people above a certain age only.
Contraindications and side effects
Hypoglycaemia is a contraindication for many insulin preparations, and is also an important side effect. Although less common than in people with type 1 diabetes, it is still a problem with insulin therapy in type 2 diabetes, especially in older people, in whom the symptoms may not be recognised. Hypoglycaemia risk increases with the duration of insulin treatment (Zammitt and Frier, 2005), and in the UKPDS, at least one severe hypoglycaemic episode per year occurred in 2.3% of recipients (UKPDS Group, 1998).
Many people gain weight when starting insulin (Inzucchi et al, 2012), which is a significant issue for people with type 2 diabetes as many are already overweight. In the UKPDS, insulin therapy was associated with an average weight gain of 4 kg (UKPDS Group, 1998). This leads to increased cardiovascular risk (Russell-Jones and Khan, 2007) and can reduce adherence with treatment. A care plan and education when initiating insulin are essential to minimise the risks of weight gain and hypoglycaemia.
Other safety-related issues relevant to insulin use in type 2 diabetes that have arisen more recently are covered later in this module in their own section.
Types of insulin
Insulin preparations differ in terms of:
- Their origin. The amino acid sequences of animal insulins, human insulins and human insulin analogues are different. “Insulin analogues” are so called because their amino acid sequences are different from those occurring in nature, yet they retain the ability to interact with the human insulin receptor. Different techniques are also used to produce different insulin preparations. Human insulin, for example, may be generated by recombinant DNA technology using yeast or bacteria, or by enzymatic modification of porcine insulin (BMJ Group and RPS Publishing, 2012).
- Their time–action profiles. In prescribing resources, such as the British National Formulary (BNF) and MIMS, insulin preparations are typically categorised according to their time–action profiles. The categorisation in these resources differs slightly. For example, pre-mixed insulins aside, while the BNF (BMJ Group and RPS Publishing, 2012) considers short-acting and intermediate- and long-acting insulins, MIMS (Haymarket Medical, 2012) divides the preparations into very rapidly, short-, intermediate- and long-acting insulins. This article also categorises the different insulin preparations on the basis of their time–action profiles.
There are four manufacturers supplying insulin in the UK. Eli Lilly and Company Limited (Basingstoke), Novo Nordisk Limited (Crawley) and Sanofi (Guildford) manufacture a variety of genetically engineered human insulins and human insulin analogues. Wockhardt UK Limited (Wrexham) is now the only supplier of animal (pork and beef) insulins.
Short- and rapid-acting insulins
Short- and rapid-acting insulins mimic the short burst of insulin associated with eating carbohydrate-containing meals produced by individuals without diabetes. They are usually injected with meals (and are therefore known also as prandial insulins), but additionally are useful in managing hyperglycaemia during periods of illness. As the name suggests, they are relatively short acting, and are usually used in combination with an intermediate- or long-acting insulin. They can be further sub-divided into short-acting (or soluble) insulins and the more recently available rapid-acting insulin analogues.
Short-acting (soluble) insulins
Soluble insulins are clear solutions that are injected between approximately 15 and 30 minutes before meals, have a rapid onset of action (approximately 30–60 minutes), have a peak action between approximately 2 and 4 hours and can last for up to approximately 8 hours. Figure 1 lists the soluble insulin preparations currently available in the UK.
Rapid-acting insulin analogues
Rapid-acting insulin analogues have been developed using genetic and protein engineering techniques, with the aim of changing the amino acid sequence of the human insulin molecule to reduce its tendency to self-associate (Williams and Pickup, 2004). Such changes give these preparations a faster onset of action and a shorter duration, allowing them to be injected immediately before or even after a meal, which may be more convenient for users.
There is evidence that, compared with soluble insulins, they are associated with a lower risk of hypoglycaemia (Zammitt and Frier, 2005) and can lower 2-hour postprandial blood glucose levels, lower the risk of late postprandial hypoglycaemia, and give a better quality of life through greater flexibility in timing and dosing of insulin (Rossetti et al, 2008). The currently available rapid-acting insulin analogues are listed in Figure 1.
Intermediate- and long-acting insulins
Intermediate- and long-acting insulins are also called basal insulins as their function is to provide a relatively steady supply of insulin to maintain blood glucose levels overnight and between meals, mimicking the background insulin produced by individuals without diabetes. Collectively, they have an onset of action within approximately 1–2 hours and a duration of between around 16 and 35 hours (BMJ Group and RPS Publishing, 2012). A number of different methods of prolonging the effect of insulin after injection have been developed over the years, including suspending human insulin with protamine or zinc and altering the amino acid sequence of human insulin.
Depending on the insulin used, they are usually given: once or twice daily; before breakfast, at bedtime or both; and often in combination with OADs or short- or rapid-acting insulins. There are a number of types of intermediate- and long-acting insulin.
NPH (isophane) insulins
Isophane insulins are the “traditional” cloudy insulins, which comprise a suspension of insulin with protamine. They are commonly classified as intermediate-acting insulins and are also known as neutral protamine Hagedorn (NPH) insulin. NPH insulin must be re-suspended before use, it has quite a marked peak in its time–action profile and there may be large day-to-day variability in absorption after injection (Yki-Jarvinen, 2004), which, compared with long-acting insulin analogues, may result in variability in blood glucose levels and a higher risk of hypoglycaemia (Rossetti et al, 2008).
Information on the five NPH insulins available in the UK (Humulin I, Hypurin Bovine Isophane, Hypurin Porcine Isophane, Insulatard and Insuman Basal) is provided in Figure 1.
Long-acting insulin analogues
The long-acting insulin analogues are formed by alteration of the amino acid sequence of human insulin to give the desired prolonged duration of action. These preparations are clear and do not require re-suspension before use. HbA1c attainment is similar to that achieved with NPH insulins, but long-acting insulin analogues may have some advantages in that their use can result in comparatively reduced fasting blood glucose levels with a lower risk of nocturnal hypoglycaemia and lower variability of blood glucose levels (Rossetti et al, 2008). However, they are more expensive than NPH.
There is evidence to suggest that treatment with insulin detemir is associated with slightly less weight gain than insulin glargine or NPH insulin (Haak et al, 2005; Dornhorst et al, 2007; Rosenstock et al, 2008), but otherwise use of the long-acting insulin analogues results in similar HbA1c levels and risk of hypoglycaemia (Rosenstock et al, 2008).
They are often injected at bedtime but can be given first thing in the morning (Standl et al, 2006), and where required, insulin detemir can be given in two daily doses depending on the person’s needs. There is some evidence that insulin glargine given in the morning may be more effective in reducing HbA1c than that administered at bedtime (Fritsche et al, 2003).
Other preparations
Long-acting suspensions of animal insulins with zinc, or protamine and zinc, are also in use. Currently, two bovine preparations, Hypurin Bovine Lente and Hypurin Bovine PZI, are available in the UK (Figure 1).
Pre-mixed (biphasic) insulins
As the name suggests, pre-mixed (biphasic) insulins are a mixture of a short-acting insulin or rapid-acting insulin analogue with a longer-acting protaminated version of the same insulin in a fixed ratio. The number in the brand name denotes the proportion of short-acting insulin in the mixture (Figure 1). These insulins are designed to provide a peak of activity to address postprandial hyperglycaemia, as well as a basal component to address hyperglycaemia between meals or overnight in a single injection.
The insulin appears cloudy and needs to be thoroughly re-suspended before each injection. They are usually given twice a day, before breakfast and before the evening meal, but can be given once or three times daily, with a meal (Kilo et al, 2003).
Mixtures containing soluble insulin should ideally be given 15–30 minutes before the meal. In contrast, pre-mixed insulins containing a rapid-acting insulin analogue can be given just before a meal and so may be more convenient to use than human mixtures (Garber et al, 2007). Pre-mixed insulin analogues are, however, more expensive than their human or animal counterparts.
The aim of insulin therapy in type 2 diabetes
The philosophy of insulin therapy for people with type 1 diabetes, who do not produce any insulin, is to mimic as closely as possible with exogenous insulin the insulin secretion pattern of someone who does not have diabetes. This includes a continuous, steady flow of insulin (basal) with rapid bursts of insulin (bolus) following carbohydrate consumption. Multiple injection regimens (one or two injections of intermediate- or long-acting insulin, and short-or rapid-acting insulin with meals) or insulin pumps are used to achieve this.
However, insulin therapy in type 2 diabetes is not as straightforward and there are a variety of insulin regimens in use in clinical practice. Adding insulin to type 2 diabetes treatment can significantly improve glycaemic control (Wright et al, 2002), but when and how to do so is the subject of considerable debate.
When to initiate insulin therapy in type 2 diabetes?
Evidence of the long-term benefits of achieving tight glycaemic control in the early stages of type 2 diabetes (the “legacy effect”) may encourage early use of insulin (Holman et al, 2008). This is endorsed in the American Diabetes Association (ADA) and European Association of the Study of Diabetes (EASD) consensus guidelines, where insulin is considered as an option for second-line add-on therapy after metformin has failed (Inzucchi et al, 2012).
However, NICE, in its revised type 2 diabetes guidelines (NICE, 2009), positioned it as a third-line treatment option for most people. Similarly, the subsequently published update on newer therapies for blood glucose lowering in type 2 diabetes positioned insulin as a third-line therapy option (NICE, 2009).
In practice, the degree of hyperglycaemic symptoms, especially unintentional weight loss, and level of HbA1c will influence how quickly insulin is introduced in a person with type 2 diabetes.
Types of insulin regimen: Evidence and clinical guidance
Evidence
Table 2 describes a number of the insulin regimens currently used by people with type 2 diabetes in the UK. There still remain some gaps in the evidence directly comparing the different insulin regimens in people with type 2 diabetes. NICE has provided guidance on how to initiate insulin therapy and some direction on subsequent intensification (NICE, 2009).
The previous version of this module presented a non-exhaustive description of some of the notable trials examining insulin therapy regimens in type 2 diabetes, covering: direct comparisons of different insulin initiation regimens; direct comparisons of human insulin and insulin analogue regimens; and comparisons of titration method. The reader is directed to the previous version for those details (Hill, 2009).
Subsequent developments in research into the analogues include: publication of 3-year results for the 4T (Treating to Target in Type 2 Diabetes) study, which was conducted to directly compare the effects of different insulin analogue regimens in type 2 diabetes (Holman et al, 2009); presentation of results from the 24-month maintenance phase of DURABLE (Durability of Basal Versus Lispro Mix 75/25 Insulin Efficacy; Buse et al, 2011); and an upload onto ClinicalTrials.gov of results from the “All to Target” trial (ClinicalTrials.gov, 2011).
What does NICE recommend?
Recommendations for OADs when used in combination with insulin
It seems sensible to continue OADs that are insulin sparing (i.e. those that reduce insulin requirements) when initiating insulin therapy, and the evidence demonstrates achievement of a significantly lower HbA1c, a lower injection dose requirement and less weight gain in regimens where insulin is combined with OADs compared with insulin monotherapy (Goudswaard et al, 2004), although a contrasting view as presented in a recent meta-analysis (Hemmingsen et al, 2012).
Specifically, NICE made the following recommendations for OADs in combination with insulin in its 2009 guidance on type 2 diabetes (NICE, 2009):
- When starting basal insulin therapy, continue with metformin and sulphonylureas (and acarbose, if used) but review the use of sulphonylurea if hypoglycaemia occurs.
- When starting pre-mixed insulin therapy (or basal–bolus regimens), continue with metformin and sulphonylurea initially, but discontinue the latter if hypoglycaemia occurs.
- Consider combining pioglitazone with insulin therapy in a person who has previously had a marked glucose-lowering response to a thiazolidinedione, and in people on high-dose insulin therapy whose blood glucose is inadequately controlled.
Recommendations for insulin therapy
Box 2 provides an abbreviated version of NICE recommendations for insulin therapy in type 2 diabetes (NICE, 2009).
Regimens summary
In summary, no single regimen is the best. In practice, there is often a compromise to be struck between the achievement of HbA1c targets and minimising the risk of hypoglycaemia and weight gain, and the frequency of daily injections a person is willing to accept. The chosen regimen needs to be individualised to take account of personal choice, lifestyle, job and work shifts, travel, eating habits, dependence on others for injections, age, life expectancy, visual or manual dexterity issues, the HbA1c target to be achieved, complications, other comorbidities, cognitive function, weight and hypoglycaemia risk (especially in older people).
A once-daily basal insulin regimen added to OADs is a simple starting point but, with the progression of type 2 diabetes, is unlikely to be sufficient in the long term. A basal insulin regimen addresses fasting hyperglycaemia in particular, but the lower the HbA1c level to be achieved, the more significant the management of postprandial hyperglycaemia becomes if the target is to be reached. For example, at an HbA1c level of 56 mmol/mol (7.3%), postprandial hyperglycaemia accounts for about 70% of overall hyperglycaemia and fasting hyperglycaemia accounts for around 30% (Monnier et al, 2003). In contrast, if HbA1c is >88 mmol/mol (>10.2%), these percentages are reversed. As a result, postprandial hyperglycaemia will need to be addressed – either in the form of short-acting or biphasic insulin with one or more meals – to achieve tighter HbA1c targets, and particularly over time as endogenous insulin secretion diminishes (Barnett et al, 2003).
Given that initial insulin regimens will require intensification, the author recommends that practitioners “think ahead” when insulin therapy is initiated. For example, some older people may struggle to self-care if prescribed a basal–bolus regimen, and in such cases intensifying a basal-only regimen by adding short-acting insulin at mealtimes would not be the most sensible long-term strategy. A consensus statement from 2008 offers some sensible and practical suggestions in this area (Barnett et al, 2008). Boxes 3 and 4 provide case examples that highlight some of the practical considerations related to insulin therapy in type 2 diabetes.
Starting and adjusting insulin
Start doses
Guidance from the Royal College of Nursing (2006) on “Starting insulin treatment in adults with Type 2 diabetes” makes the following comment concerning start doses:
“Once-daily regimens often start with 10 units. Most twice-daily regimens start with 6–10 units twice daily, depending upon the person’s build. There is no such thing as a ‘correct’ dose: starting low and working up will build the person’s confidence and your own.”
Adjustment
In the author’s view people with type 2 diabetes should ideally be encouraged to self-titrate their insulin dose to achieve target blood glucose levels without unacceptable hypoglycaemia. Indeed, the results of the ATLANTUS study demonstrated that self-titration can be more effective than titration advised by healthcare professionals (Davies et al, 2005).
People with diabetes should be encouraged to look for patterns in their blood glucose readings, and to not alter insulin doses on the basis of a single result. They should be able to identify what the problem is (i.e. readings are above or below target), when the problem is occurring (e.g. during the night) and which insulin or insulins are active when it occurs (e.g. basal or prandial). Before making an adjustment to the insulin dose, however, other potential causes should be excluded, such as poor injection technique, use of inappropriate injection sites, lipohypertrophy, exercise, dietary indiscretions and inaccurate blood glucose monitoring.
Blood glucose monitoring readings generally inform on the effect of the last insulin injection, and therefore it is this injection that should be adjusted. The paragraphs that follow give some simple advice, which is adapted from the Tayside Diabetes Handbook (NHS Tayside Diabetes MCN, 2009):
- Once-daily basal regimen. Increase or decrease the dose if pre-breakfast readings are above or below target, respectively.
- Twice-daily pre-mixed insulin regimen. Increase or decrease the morning dose if the pre-lunch and pre-evening meal readings are above or below target, respectively. Increase or decrease the evening dose if the pre-bedtime and pre-breakfast readings are above or below target, respectively.
- Basal–bolus regimen. Increase or decrease the basal insulin dose if pre-breakfast readings are above or below target, respectively. Increase or decrease the breakfast bolus dose if pre-lunch readings are above or below target, respectively. Increase or decrease the lunch bolus dose if pre-evening meal readings are above or below target, respectively. Increase or decrease the evening bolus dose if pre-bedtime readings are above or below target, respectively.
The amount by which the dose is adjusted can vary. The experience of the clinician, symptoms, concern about hypoglycaemia, the level of involvement of the person with diabetes and existence of complications are some of the factors that will determine how quickly and by how much insulin doses are adjusted.
Insulin delivery devices
Most insulins are available in a 10 mL vial for use with a syringe (especially useful if insulin is being given by someone else to reduce risk of stick injury), 3 mL cartridges for use in durable pens, or in 3 mL disposable pens. Insulin manufacturers generally produce insulin pen devices compatible with cartridges containing their own insulin, which are not interchangeable. Needles are available in a variety of lengths, from 4 mm to 12.7 mm, and should be used once only and disposed of according to local sharps policy.
Although insulin pump therapy is used in the USA for people with type 2 diabetes, this is not recommended by NICE (2008).
Care planning and education
All people starting insulin therapy should have a care plan agreed with their healthcare professional and a programme of education to enable them to build up their skills in self-management. The latter was recommended by NICE in its 2009 guidance on type 2 diabetes (NICE, 2009). Such a programme will include advice on injection technique, suitable injection sites, disposal of needles, identification, treatment and avoidance of hypoglycaemia, management of insulin during illness, diet (weight management and carbohydrate load) and insulin adjustment to achieve target HbA1c levels. Knowledge should be checked at the annual diabetes review, including inspection of injection sites for signs of lipohypertrophy.
Safety-related issues
In addition to the well-established issues of hypoglycaemia and weight gain, both of which were discussed earlier in the module, themes concerning other aspects of insulin’s safety and the need to minimise errors have risen to prominence in the past few years.
“Safe Use of Insulin” e-learning module
A report by the NHS’s National Patient Safety Agency (NPSA) revealed that there had been over 16000 medication incident reports involving insulin between 1 November 2003 and 1 November 2009 (NPSA, 2010; 2011). These included errors resulting from abbreviating “units” to “u” or “iu”, as well as many mistakes relating to people being given the wrong insulin owing to the names of several insulins being similar. As part of the response, an e-learning module on the safe use of insulin was released by NHS Diabetes (2010).
Insulin passports
A subsequent report from the NPSA (2011) has made it compulsory for adults with type 2 diabetes on insulin therapy to receive, by the end of August 2012, an insulin passport and a patient information booklet, in order to “help provide accurate identification of their current insulin products and provide essential information across healthcare sector.” The NPSA has produced a double-sided A4 template for the insulin passport, which is designed to fold down to the size of a credit card (NPSA, 2011).
Insulin safety cards have been produced by the insulin companies and a short patient information booklet is also available on the websites for the NPSA (http://www.npsa.nhs.uk/) and NHS Diabetes (http://www.diabetes.nhs.uk/), which may be more helpful for some people with diabetes. If using these, instead of the NPSA tools, this should be recorded on the organisation’s risk register to comply with the NPSA alert action.
Driving legislation
Recent changes to the Driver and Vehicle Licensing Agency (DVLA) medical standards potentially have a significant bearing on drivers with diabetes who are treated with insulin (Diabetes UK, 2012). People using insulin have been advised to test their blood glucose level before driving Group 1 vehicles (cars and motorcycles). People with insulin-treated diabetes can now apply to hold a licence to drive Group 2 vehicles (large goods vehicles and passenger-carrying vehicles) as long as they meet specific safety criteria, which include:
- Monitoring blood glucose at least twice daily, and at times relevant to driving, using a meter with a memory function.
- Attending an annual examination by a diabetologist, at which th
