Article Text

Download PDFPDF

Hypoglycaemia: current management and controversies
  1. Pratik Choudhary,
  2. Stephanie A Amiel
  1. King's College London School of Medicine, Denmark Hill Campus, Weston Education Centre, London, UK
  1. Correspondence to Dr Pratik Choudhary, King's College London School of Medicine, 3.39, Weston Education Centre, Cutcombe Road, Denmark Hill Campus, Weston Education Centre, London SE5 9RJ, UK; pratik.choudhary{at}


Hypoglycaemia is a major burden on patients and society and is often a barrier to the achievement of tight glycaemic control. Intact awareness of hypoglycaemia is crucial to recognising and treating hypoglycaemia before it becomes severe enough to impair consciousness. Repeated hypoglycaemia can lead to impaired awareness increasing the risk of severe hypoglycaemia up to sixfold. Hypoglycaemia is much less common in those with type 2 diabetes, the incidence increasing with longer duration of treatment with insulin, associated comorbidities, and in the elderly. Alcohol, advancing age and exercise may predispose to hypoglycaemia. Newer agents acting via the incretin axis are associated with low rates of hypoglycaemia. Intensification of therapy to achieve tight glucose control can increase the risk of hypoglycaemia in the outpatient as well as critical care setting. In some studies this has also been associated with increased mortality, although causality has not been proven. Insulin treated patients are currently restricted from driving heavy goods vehicles or public service vehicles, although it is unclear if those with diabetes have any higher rates of accidents than those without diabetes. Surveys show that professionals are poor at emphasising the Driver and Vehicle Licensing Agency recommendations for drivers with diabetes in the UK. At every visit, patients with diabetes on hypoglycaemic agents should be assessed for frequency, severity, and awareness of hypoglycaemia. The main therapeutic strategies for reducing hypoglycaemia are structured patient education, use of modern insulin analogues, insulin pumps, and continuous glucose monitoring. Transplantation of islets or whole pancreas is indicated in those with recurrent disabling hypoglycaemia.

  • Diabetes & endocrinology
  • general diabetes

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Hypoglycaemia (literally low blood glucose concentration) and the fear it engenders is a major barrier to achieving the glycaemic targets known to minimise risk of complications of diabetes.1 It has important consequences, for the patient, carers and also for society and its healthcare systems. Hypoglycaemia impacts on well-being, productivity, and quality of life of people with diabetes2 and can curtail driving and employment opportunities.3 4 In 2003 a Scottish survey found the total cost of hypoglycaemia across a population of about 370 000 (8655 people with diabetes) was just over £92 000 (€110 000, US$140 000).5 Hospital episode statistics from England show that in 2007–2008 there were over 10 000 admissions with hypoglycaemic emergencies in the UK, the majority in those over 75 years of age, resulting in a total of 50 000 bed days.6

Different studies have used different glucose concentrations to define hypoglycaemia, with most clinicians using a cut-off plasma glucose between 3.5–3.9 mmol/l.7 In clinical practice, hypoglycaemia is often defined by its clinical consequences. In ‘mild’ hypoglycaemia, the patient can recognise the symptoms of falling blood glucose and take appropriate action to restore blood glucose to normal by ingestion of carbohydrate. In ‘severe’ hypoglycaemia, blood glucose concentrations fall to a level where the brain is unable to function properly and is defined as any episode that requires third party intervention or is associated with seizures or coma. It often requires parenteral treatment with intravenous glucose or intramuscular glucagon (table 1). Repeated hypoglycaemia blunts symptomatic and hormonal responses to subsequent episodes leading to impaired awareness of hypoglycaemia, also called hypoglycaemia associated autonomic failure (HAAF). These patients often experience measured glucose concentrations below 2.0 mmol/l without symptoms and rely on others diagnosing their hypoglycaemia and/or on capillary glucose testing to identify hypoglycaemia.

Table 1

Clinical classification of hypoglycaemia

In this paper, we highlight some of the relevant issues around hypoglycaemia that are currently topical.

Prevalence of hypoglycaemia

Type 1 diabetes

The landmark Diabetes Complications and Control Trial of intensified insulin therapy reported a threefold increase in the risk of severe hypoglycaemia between the intensive (>3 injections/day or continuous subcutaneous insulin therapy) and conventional (twice daily insulin) treatment arms (61.2 vs 18.7 episodes/100 patient years).8 In this intervention study, for the same given value of glycosylated haemoglobin (HbA1c), those in the intensive control arm experienced more hypoglycaemia. In contrast, structured education programmes such as the UK's DAFNE (Dose Adjustment For Normal Eating) programme and the ITTP (Teaching and Training in Intensification of Insulin therapy) in Germany showed that improvement in glucose control could be achieved without an increase in severe hypoglycaemia.9 10 Observational studies from the UK and Europe suggest that severe hypoglycaemia occurs in 35–42% of patients and the rate of severe hypoglycaemia is between 90–130 episodes/100 patient years.11–14 The UK Hypoglycaemia Study found that those with <5 years duration of diabetes experienced lower rates of severe hypoglycaemia than those with >15 years of diabetes (22% vs 46%).15 Twenty-five per cent to 30% of patients with type 1 diabetes have impaired awareness of hypoglycaemia which is associated with a 3–6 times higher risk of severe hypoglycaemia.16 17

Type 2 diabetes

Oral therapies

Insulin sensitisers and incretin based therapies have reported a low risk of hypoglycaemia. The ADOPT study compared monotherapy with metformin, rosiglitazone, and glibenclamide and found very low rates of severe hypoglycaemia requiring third party help, highest in the group on sulfonylureas (0.1% for metformin and rosiglitazone and 0.6% for glibenclamide). Rates of self reported symptomatic hypoglycaemia (not confirmed with blood tests) were around 10% for metformin and rosiglitazone and 38.6% for glibenclamide.18 Trials of DPP IV inhibitors have reported very low rates of hypoglycaemia unless combined with sulfonylureas.19 20 Other data confirm the higher rate of hypoglycaemia with insulin secretagogues, with 31% of subjects reporting mild hypoglycaemic symptoms in the first year of use in the UK Prospective Diabetes Study (UKPDS).21 A population based study from Germany looking at over 30 000 patients over 4 years found fewer episodes of severe hypoglycaemia with the newer generation sulfonylurea glimepiride (0.86 vs 5.6 events/1000 patient years).22

Insulin treatment

The UK Hypoglycaemia Study was a recent UK based, multicentre, observational study which found similar incidence of hypoglycaemia between those recently started on insulin (≤3 years) and those treated with sulfonylureas. Those with longer duration of insulin therapy (≥5 years) had higher rates of hypoglycaemia (25%), which were similar to rates seen in patients with type 1 diabetes (22% for those with type 1 diabetes <5 years and 46% in those with type 1 diabetes >15 years).15 A retrospective questionnaire based study from Denmark in 401 patients with insulin treated type 2 diabetes found 16.5% of patients experienced at least one episode of severe hypoglycaemia with an incidence of 44 episodes/100 patient years,23 which is similar to the study by Donnelly et al who reported 35 episodes/100 patient years of severe hypoglycaemia in insulin treated type 2 diabetes.24 Population based studies from the UK and from Germany, which only captured episodes requiring assistance from emergency services or hospitalisation, have reported similar rates of severe hypoglycaemia in insulin treated type 1 and type 2 diabetes; however, both studies noted that, while a high proportion of severe hypoglycaemia in type 1 diabetes was self managed or managed by carers, much of the severe hypoglycaemia in type 2 diabetes occurred in older patients with multiple comorbidities and often required hospital admission.5 25 The incidence of hypoglycaemia and in particular severe hypoglycaemia has been reported to be increasing, based on population based surveys, and this may reflect attempts to intensify glucose management.26 Overall, rates of hypoglycaemia are much lower in those with type 2 diabetes than in type 1 diabetes, except in those with a long duration of insulin treatment.

Effect of different insulin regimens

Studies in patients being initiated on insulin due to failing control on oral agents show that once daily basal insulin provides equivalent glucose control with lower rates of hypoglycaemia compared to twice daily biphasic insulin in patients with type 2 diabetes.27 The Veterans Affairs study compared morning basal insulin with more intensive regimens using twice daily premixed insulin or multiple daily injections and found no difference in rates of severe hypoglycaemia, although mild symptomatic hypoglycaemia was more common with the more intensive regimens.28 The more recent 4-T study using analogue insulins found lowest rates of hypoglycaemia with basal only insulin (Detemir), higher with twice daily biphasic insulin (Novomix30) and highest in those started on prandial only insulin but who had some basal insulin added in if not to target.29 30

New therapies

Newer agents based on the incretin system such as DPP-4 inhibitors and GLP-1 analogues help stimulate glucose dependent insulin release. In laboratory studies, some of the hypoglycaemic actions of GLP-1 are attenuated in the presence of hypoglycaemia (insulin secretion and glucagon suppression), and in clinical studies these agents have been associated with low risk of hypoglycaemia except when combined with sulfonylureas or insulin19 31 32.

Risk factors for hypoglycaemia

Factors such as antecedent hypoglycaemia, recent glycaemic control, alcohol, exercise, and age affect symptomatic and hormonal responses to subsequent hypoglycaemia.33 Female sex,34 35 sleep,36 37 and older age38 are associated with lower hormone and symptom responses to hypoglycaemia. Older patients also report a different profile of symptoms and respond to hypoglycaemia with fewer autonomic symptoms and more prominent neuroglycopenic symptoms.39–41 In this group, hypoglycaemia can be misdiagnosed as dementia or neurological events.42 In patients with tablet treated type 2 diabetes, counter-regulatory hormonal and possibly symptom responses to hypoglycaemia may be triggered at higher glucose concentrations than non-diabetic controls,43 and symptoms develop at higher plasma glucose concentrations in poorly controlled type 2 patients than in those with type 1 diabetes or people with type 2 diabetes in tighter glucose control.43 44 Those treated with sulfonylureas may develop symptoms at higher glucose concentrations than insulin treated patients45 45; however, perhaps because of frequency of exposure, asymptomatic and severe hypoglycaemia can complicate sulfonylurea therapy. When severe hypoglycaemia does occur in someone on sulfonylureas, it is likely to be prolonged and may recur after emergency treatment.

With increasing age, the ‘protective window’ between generation and recognition of symptoms and onset of cognitive dysfunction is lost, making the elderly more vulnerable to the morbidity related to hypoglycaemia.

Intervention studies of intensive insulin therapy such as the DCCT and ITTP found that duration of insulin therapy, loss of endogenous insulin (reflected by undetectable or very low C-peptide concentrations), and past history of severe hypoglycaemia are the main predictors of an increased risk of future severe hypoglycaemia.9 46 As mentioned above, duration of diabetes and impaired awareness of hypoglycaemia (itself usually a result of recurrent exposure to hypoglycaemia earlier) also predispose to severe hypoglycaemia.15 16 Some studies have found that high concentrations of angiotensin converting enzyme (ACE) were associated with an increased risk of severe hypoglycaemia,14 47 but an observational study in type 2 diabetes did not detect any relationship between ACE genotype and severe hypoglycaemia rates. However, 50% of patients in this study were taking drugs which block the renin–angiotensin system which may have influenced the results.48

Is hypoglycaemia a risk factor for cardiovascular death?

The ACCORD study was a randomised study of over 10 000 type 2 diabetic patients randomised to intensive therapy or standard care that was terminated early due to an increased risk of mortality in the intensive treatment arm.49 In this study, HbA1c was rapidly reduced from 8.3% to under 6.5% within a few months, with over 77% of patients receiving insulin in the intervention group versus 55% in the control group. Those randomised to the intensive therapy arm had a higher incidence of severe hypoglycaemia (16.2% vs 5.1%; intensive vs conventional therapy), and although there was no causal linkage, it is possible that hypoglycaemia may have contributed to some of these deaths. The mechanism may be similar to that implicated in the ‘dead in bed’ syndrome which occurs very rarely in young adults with type 1 diabetes.50 51 It is thought that these sudden deaths occur as a result of sudden cardiac arrhythmias triggered by hypoglycaemia. In the experimental setting, hypoglycaemia has been shown to cause QT prolongation, increased QT dispersion, atrial and/or ventricular ectopics and sudden bradycardia, all of which can predispose to sudden cardiac death.51 It is important to recognise that other studies of intensive control in type 2 diabetes published recently such as the ADVANCE and VA-DT studies did not show any increase in mortality with randomly allocated intensive glycaemic control. Although rates of hypoglycaemia were higher in the intensively controlled groups (2.7% vs 1.5% in the ADVANCE study; 9 vs 3/100 patient years in the VADT study), overall rates were lower than those in the ACCORD trial.52 53 The different outcomes may reflect differences in the patients or the speed, vigour, and method of the glucose lowering strategies in the intensive programmes. The clinical lesson from these data is that treatment targets need to be defined individually. Based on evidence from the follow-up studies of type 1 and type 2 diabetes, intensified diabetes therapy should be applied early in the course of the disease to obtain reduced risk of diabetes complications. Those with advanced age or established cardiovascular disease may require a more tailored approach that takes into account the gain to be expected and the individual risk of a severe hypoglycaemic episode.

Driving and hypoglycaemia

Current UK legislation prevents people with insulin treated diabetes holding group 2 licences required to drive LGV (large goods vehicles or those >7.5 tonnes) or PCVs (passenger carrying vehicles or those with >9 seats used vocationally.3 These restrictions were based on the risk of hypoglycaemia in insulin treated individuals and the hazard of damage, which is greater in a large vehicle compared with a smaller one. Based on police notifications to the Driver and Vehicle Licensing Agency (DVLA) in the UK, hypoglycaemia is implicated in approximately 30 serious road accidents per month and up to five fatalities per year.54 Driving safely places a significant metabolic demand on the brain, subjects utilising 20% more glucose when driving a simulator compared to just watching a film. Driving performance deteriorated even during modest hypoglycaemia (3.4 mmol/l)55 and there was a delay of about 20–30 min after the blood glucose was corrected before driving performance improved.56

Cox et al found that patients with type 1 diabetes but not insulin treated type 2 patients were more likely than their spouses to have road crashes or traffic violations.57 However, population based studies from Wisconsin and from the UK showed no difference in accident rates between patients with diabetes and controls.58–60 Most people at high risk of road accidents are probably prevented from driving by current regulation around fitness to drive, and it is possible that those with diabetes may be more careful while driving due to concerns about losing their licence. Surveys show that diabetes professionals are poor at emphasising the recommendations regarding driving and hypoglycaemia.61 Advice for those with diabetes includes:

  • inform the DVLA and their insurance companies of their condition.

  • check their glucose every time they are about to drive and if driving for more than 2 h they must stop and recheck glucose.

  • keep a source of carbohydrate and glucose checking equipment in the vehicle.

  • do not drive if blood glucose is <5 mmol/l if always aware of hypoglycaemia (a higher glucose concentration may be appropriate for someone with impaired awareness).

  • inform the DVLA of any severe hypoglycaemia that occurs during waking hours.3

If hypoglycaemia is experienced during driving, the driver should stop the vehicle, remove the keys from the ignition, test the blood glucose if possible, treat by ingesting appropriate carbohydrate, and wait for at least 30 min before restarting their journey. This is because following hypoglycaemia, cognitive function remains impaired for up to 30 min after restoration of euglycaemia.62 It is important for the practitioner certifying for safety to drive to provide the correct information to the DVLA and correct advice to the patient.

At present in the UK, private driving licences may be held by people with insulin treated diabetes who are otherwise fit to drive. However, day time severe hypoglycaemia must be reported by the patient to the DVLA and may lead to suspension of the licence. A medical review will then be required before reinstatement. Patients who experience severe hypoglycaemia during waking hours must not drive. Hypoglycaemia unawareness is also a risk for serious accident, as reaction times slow before the patient is aware of the situation. Current EU legislation is considering making unawareness alone a disqualification for driving, but at present, patients who are unaware but have never experienced severe hypoglycaemia in the day may continue to drive. It is sensible to modify the blanket advice about checking blood glucose before driving to include a higher glucose concentration before and during driving (we use 7 mmol/l rather than the 5 mmol/l permissible in someone with good hypoglycaemia awareness); and to avoid long drives and check at least hourly if necessary. There is evidence that driving itself causes blood glucose to fall.

Hypoglycaemia in critical care

In 2001 van den Berghe et al published a randomised controlled trial of critically ill surgical patients showing that tight glucose control reduced hospital mortality by one third.63 However, subsequent larger multicentre studies of tight glucose control in the critically ill64 65 were unable to reproduce these benefits and a large study of over 10 000 patients showed a trend to increased mortality in the intensive treatment arm.66 All these studies reported higher rates of hypoglycaemia in those randomised to intensive therapy, some as high as 40% compared to the 5% rate in the initial trial. It remains unclear whether intensive insulin therapy induced hypoglycaemia per se is responsible for this adverse outcome, or if hypoglycaemia is just a marker of those who are more likely not to do well. One contributor to the discrepant results of various studies may be the protocols used to apply and, most importantly, monitor blood glucose control. Although more work is required in this field, it seems sensible to suggest that in the setting of intensive care, blood glucose should be controlled as tightly as possible without inducing hypoglycaemia.

Management of hypoglycaemia

In diabetes clinics, clinicians are often focused on reducing HbA1c and cardiovascular risk to achieve targets set out by various bodies67 68; it is often easy to overlook the importance of hypoglycaemia, which for patients is often more important and immediate than the potential risks of years of poor control (please see box 1).

Box 1

Assessment and management of a patient with frequent hypoglycaemia

  • Ask about:

    • Frequency and timing of:

      • Severe hypoglycaemia

      • Mild hypoglycaemia

      • Hypoglycaemia detected by others

      • Hypoglycaemia detected only because of monitoring

    • Patient awareness of hypoglycaemia

    • Relatives' comments about hypoglycaemia

  • Check:

    • Daily routine for insulin administration and eating patterns

    • Exercise routine

    • Alcohol habits

  • Examine:

    • Home glucose diary for frequency and distribution of hypoglycaemia

  • Check:

    • 9 am cortisol or Synacthen test

    • Thyroid function

    • Growth hormone secretion

    • Insulin binding antibodies

    • Renal function

    • Liver function

    • Markers of coeliac disease

    • Indications of malabsorption

    • Drug interactions (especially with sulfonylureas, eg, warfarin, aspirin, trimethoprim, sulfonamides, fibrates, monamine oxidase inhibitors, probenecid, allopurinol)

  • Identify undetected hypoglycaemia—consider:

    • 3 am blood glucose testing

    • Continuous glucose monitoring

  • Treatment options include sequential use of:

    • 1. Treatment of associated comorbidities

    • 2. Structured education for flexible insulin therapy

    • 3. Structured education for hypoglycaemia avoidance including management strategies for intermittent exercise or alcohol intake

    • 4. Appropriate use of insulin analogues

    • 5. Use of insulin pump therapy

    • 6. Transplantation


Patients should be asked whether (and how often):

  1. they experience hypoglycaemia

  2. other people notice their hypoglycaemia first, and

  3. home testing shows results of <3 mmol/l during a routine test when they had no suspicion of being low.

Positive responses to the last two define hypoglycaemia unawareness. It should be realised that with current insulin delivery tools some hypoglycaemia is inevitable, even in very well managed patients, but it should not be more than two or three mild, symptomatic episodes a week. If hypoglycaemia (even just mild) is more frequent than that, further enquiry is required.

In those with problematic or frequent hypoglycaemia, contributors to hypoglycaemia risk should be excluded such as:

  1. malabsorption (eg, due to coexistent undiagnosed coeliac disease or exocrine pancreatic insufficiency) and eating disorders.

  2. stress (hyperglycaemic) hormonal insufficiency such as Addison's disease, growth hormone deficiency, hypopituitarism.

  3. failure of insulin clearance (liver disease, renal failure, hypothyroidism, high concentrations of insulin binding antibodies).

  4. impaired endogenous glucose release or stores (acute alcohol excess or chronic alcoholism; liver failure).

However, in the vast majority of patients, problematic hypoglycaemia will be due to mismatch in insulin requirement and delivery. Detailed information should be obtained about the timing and severity of hypoglycaemia. It is worth documenting patterns (timing and quantity) of food intake, exercise and alcohol, particularly if any are variable in terms of frequency and intensity. Common patterns we see in those with frequent hypoglycaemia are large correction doses, frequent doses of quick acting given close together (stacking of insulin), and injection of insulin after the meal leading to high post-prandial values and then late hypoglycaemia. Many patients are not aware of the propensity for exercise and alcohol to cause delayed hypoglycaemia—sometimes 12 to 18 h later. Unusual (for that patient) duration or degree of exercise may not only lower blood glucose at the time but will exhaust liver and muscle glycogen, increasing risk of further hypoglycaemia as they are restored. Alcohol suppresses gluconeogenesis, which patients need to maintain blood glucose after 6–8 h of fasting—for example, overnight.

The patient's level of hypoglycaemia awareness can be assessed using simple questionnaires such as the Clarke or Gold scores,17 69 which have been proven to predict increased rates of severe hypoglycaemia. The Gold score consists of a seven point visual analogue scale which rates 1 as always aware and 7 as no awareness of hypoglycaemic episodes. Scores of 1 or 2 are associated with normal awareness, scores of 3 are intermediate or indeterminate, and scores of 4 or above are associated with impaired awareness of hypoglycaemia and reduced symptomatic and hormonal responses to hypoglycaemia,17 with a sixfold increase in risk of severe episodes.

3 am readings or continuous glucose monitoring often identify nocturnal hypoglycaemia, which is known to be suppress responses to hypoglycaemia the next day.70 Download of glucose meters or of pumps in those on continuous subcutaneous insulin infusion often provides additional insight into insulin doses and timing of hypoglycaemia. Treatment of acute hypoglycaemia is discussed in table 2, and common correctable errors of insulin usage are listed in box 2.

Table 2

Treatment of acute hypoglycaemia

Box 2

Common errors of insulin usage predisposing to hypoglycaemia

  • Timing of meal insulin doses (eg, delayed post-prandial administration predisposes to late hypoglycaemia and overestimate of requirements; immediate post-prandial dosing can be useful in management of small children with unpredictable food intake but is otherwise not recommended).

  • Failure to reduce insulin (most usefully the overnight background insulin) after relative increase in exercise or alcohol intake.

  • Over-correction of high glucose values, particularly post-prandial, when the measured high glucose is likely to be falling already, or repeated correction doses taken because the glucose concentration is deemed not be falling fast enough.

  • Failure to snack between meals on insulin regimens that require this for good control (eg, twice daily mixture insulin).

  • Build up of insulin from frequent small doses taken too close together (insulin ‘stacking’).

  • Treatment of post-hypoglycaemic hyperglycaemia, especially when the cause of the initial hypoglycaemia may still be active.

Therapeutic strategies

The following discussion relates primarily to the use of insulin in insulin deficient diabetes. Type 2 diabetic patients with residual insulin secretion are at lower risk for severe hypoglycaemia, but problems do occur with insulin secretagogue therapies, particularly during other illnesses or changing lifestyles, and patients using such therapies do need to be aware of when to look for problems and how to avoid them.


The cornerstone of diabetes self management is education. Structured education programmes for patients with type 1 diabetes provide patients with the skills necessary to use flexible insulin therapy. The Düsseldorf programmes of Assal and Berger have been shown to improve glucose control with reduction in hypoglycaemia,71 72 and audit data from the UK's DAFNE programme, a translation of the Dusseldorf programme, have confirmed a reduction in the rate of severe hypoglycaemia with improving HbA1c. Restoration of hypoglycaemia awareness was found 1 year after training in nearly 50% of the patients entering the programme with unawareness.73 74 Other educational programmes such as Blood Glucose Awareness Training75 and Hypoglycaemia Awareness And Avoidance (HAAT),76 focusing mainly on hypoglycaemia detection and avoidance, can be effective. In a clinic setting, Smith et al reported that patients with hypoglycaemia unawareness were less likely to adhere to recommendations for adjusting treatment regimens than those with intact awareness.77 Neuroimaging data suggest that patients with hypoglycaemia unawareness show reduced activation in neuronal networks associated with perception of pleasure and reward, and that psychological and behavioural strategies may be beneficial.

Insulin regimens

The aim is to avoid exposure to blood glucose concentrations <3 mmol/l, as this seems to be the level most likely to be associated with unawareness. Note that this does not necessarily mean increasing the upper end of the glucose target range and, when properly managed, hypoglycaemia avoidance can improve rather than deteriorate HbA1c. The most important issue to address is how the insulin taken may best approach both physiology and the patient's habits in terms of energy ingested and expended. Rigid regimens do not support a flexible lifestyle. Patients need to be able to adjust insulin around meals, exercise, alcohol ingestion, etc. A regimen that replaces background insulin (to control endogenous glucose production) effectively and separately from meal insulin replacement gives most flexibility in terms of meal time and content. Twice daily mixed insulin injection regimens and regimens which do not provide daytime background insulin independent of the tail-off of meal related insulin injections should be avoided.

Insulin analogues

A meta-analysis of studies comparing the human insulin analogue lispro with regular soluble insulin found reduced frequency of severe hypoglycaemia with lispro (3.1% vs 4.4%; p=0.024), with event rates of 14.2/100 patient years in the lispro group and 18.2/100 patient years in the regular insulin group.78 There is also evidence to support lower incidence of hypoglycaemia (mainly nocturnal) with peak-less basal insulins such as Glargine and Detemir in patients with type 1 diabetes.79–81 Although meta-analyses fail to establish significant benefit for use of analogue basal insulins across the board,82 83 they can be valuable in insulin deficient patients with high risk of nocturnal hypoglycaemia.

Insulin pump therapy

Continuous subcutaneous insulin infusion (CSII) via an insulin pump has been proposed as the ‘gold standard’ of therapy in type 1 diabetes and is recommended by the National Institute for Health and Clinical Excellence (NICE) if HbA1c <8.5% cannot be attained despite optimal therapy.84 A recent meta-analysis found that patients with CSII had a 2–4 times lower rate of severe hypoglycaemia than those treated with multiple daily injections and pooled severe hypoglycaemia rate in those treated with multiple daily injections was 62 events/100 patient years, similar to the rate in the DCCT.85 However, most studies in the meta-analysis used older basal insulin regimens. Studies comparing CSII with basal bolus regimens using modern analogue basal insulins do not show any differences in hypoglycaemia.86–88

Novel therapies

A glucose responsive insulin delivery system should allow tight glycaemic control with much less risk of hypoglycaemia. Continuous glucose monitors can provide patients with almost ‘real-time’ information of changes in glucose concentrations, particularly speed and direction of change. They can be programmed with alarms which warn of impending hypoglycaemia.89 The latest devices, combined with insulin pumps, can even suspend insulin delivery for up to 2 h if the patient fails to respond to an hypoglycaemia alarm. In some, but not all, clinical studies, these systems have been shown to reduce time spent hypoglycaemic.90 A recently published study of continuous glucose monitoring in patients already using intensified therapy (mostly CSII) showed that continuous monitoring in adults is associated with a reduction in HbA1c, but was unable to show a significant reduction in hypoglycaemia.91 Those with HbA1c <7% were able to maintain tight control with reduction of severe hypoglycaemia.92


Another option might be replacement of the pancreatic islet. Breakthroughs in islet isolation and anti-rejection therapy by the Edmonton group led to resurgent interest in islet cell transplantation as a treatment for type 1 diabetes in the early part of this decade. In the initial studies, 70% of patients also became insulin independent at 1 year. This rate fell to 15% at 5 years, although hypoglycaemia protection was retained longer than insulin independence. Despite the inefficiencies of the procedure and the side effects of life long immunosuppression, transplantation is currently a useful option for those with intractable hypoglycaemia or those already immunosuppressed for other reasons. Whole organ pancreas transplantation, either alone or in combination with renal transplantation, offers a greater chance of insulin independence when done in the best centres, but remains a very major operation with significant operative morbidity.

Where should we aim?

The NICE guidance for adults with type 1 diabetes recommends pre-prandial glucose targets of 4.0–7.0 mmol/l and post-prandial concentrations <9.0 mmol/l, although it is recognised that these are expert committee recommendations rather than category 1 evidence based. The guidelines also remind us to be aware of and look for unrecognised or disabling hypoglycaemia and hypoglycaemia unawareness in those with HbA1c values <6.5%.93 It is also important to remember that many patients with impaired awareness of hypoglycaemia actually have quite high HbA1c as their blood glucose values swing between over-correction of hyperglycaemia to overcorrection of the resultant hypoglycaemia. Relaxing glucose control is no guarantee of reduced hypoglycaemia, and in our clinical practice we have often found that adopting strategies to minimise hypoglycaemia results in improved control as we reduce the swings. In fact, no HbA1c is a guarantee of immunity from severe hypoglycaemia. For those with type 2 diabetes, NICE recommend against pursuing HbA1c values <6.5%.68 The American Diabetes Association (ADA) recommend a goal of <7% HbA1c in both type 1 and 2 diabetes and suggest that less stringent goals may be appropriate in those with extensive comorbidities, limited life expectancy or in whom tight control is not possible without unacceptable morbidity due to hypoglycaemia.67 It is important to remember that target concentrations should be accompanied by both upper and lower limits as good control is not defined just by being below a target; it also requires the absence of unacceptable hypoglycaemia.

Multiple choice questions (true (T)/false (F); answers after the references)

1. Incidence of hypoglycaemia

  1. Frequency of hypoglycaemia is related to duration of insulin treatment

  2. Some agents such as DPP-4 inhibitors and GLP-1 agonists have a low risk of hypoglycaemia

  3. The lowest risk of hypoglycaemia in insulin treated type 2 diabetes is in those treated with twice daily mixed insulins

  4. The incidence of hypoglycaemia is decreasing

  5. It is not possible to improve glucose control without a higher risk of hypoglycaemia

2. Driving and hypoglycaemia—the advice for those with diabetes who want to drive includes:

  1. They must not drive at night

  2. They must check blood glucose before they start and can only drive if blood glucose is >5 mmol/l

  3. They cannot drive if they have ever had any mild hypoglycaemia

  4. They must inform the DVLA every time they have any hypoglycaemia

  5. Those with diabetes have a significant increase in the risk of road traffic accidents

3. Assessment of hypoglycaemia

  1. Renal and/or liver failure can increase the risk of hypoglycaemia

  2. Exercise can lead to delayed hypoglycaemia 12–18 h after the exercise

  3. Using the GOLD score, a score of 6 or 7 indicates good awareness of hypoglycaemia

  4. Acute hypoglycaemia should be treated with 500 ml of Lucozade

  5. Patients with hypoglycaemia awareness were found to be very adherent to treatment changes

4. Management of hypoglycaemia

  1. Continuous glucose monitoring has been shown to reduce rates of hypoglycaemia

  2. Peak-less basal insulins are as good as insulin pump therapy in preventing nocturnal hypoglycaemia

  3. Warning signs of hypoglycaemia cannot be restored in those who have lost them

  4. Islet transplantation can offer life long insulin independence

  5. Islet transplantation offers significant reduction in hypoglycaemia not controlled by usual therapies

5. Risks of hypoglycaemia

  1. Hypoglycaemia can cause QT prolongation which may increase risk of arrhythmias

  2. All studies of intensive glucose control have shown an improvement in mortality

  3. Tighter glucose control in intensive care may be associated with an increased risk of mortality

  4. Symptoms and hormonal responses to hypoglycaemia are stronger during the night

  5. Cognitive function is impaired for 24 h after restoration of blood glucose

Main messages

  • Hypoglycaemia is a major barrier to the achievement of glycaemic targets known to prevent complications.

  • Hypoglycaemia carries significant personal and societal burden.

  • Up to 30% of patients with type 1 diabetes have impaired or absent awareness of hypoglycaemia which significantly increases the risk of severe hypoglycaemia.

  • Hypoglycaemia is also an important problem in type 2 diabetes, particularly in the elderly or those with other comorbidities.

  • An assessment of hypoglycaemic exposure and degree of awareness is an essential part of the diabetes consultation.

  • Structured education, modern methods of insulin delivery and glucose monitoring can all be used to identify hypoglycaemia and reduce exposure.

Current research questions

  • Is hypoglycaemia responsible for the increased mortality seen in some studies of tight glycaemic control?

  • What are the mechanisms of hypoglycaemia unawareness and can they be reversed?

  • What is the best strategy to restore awareness of hypoglycaemia? What are the relative roles of technology, education and behavioural interventions?

Key references

▶ Leese GP, Wang J, Broomhall J, et al. Frequency of severe hypoglycemia requiring emergency treatment in type 1 and type 2 diabetes: a population based study of health service resource use. Diabetes Care 2003;26:1176–80.

▶ UK Hypoglycemia Study Group. Risk of hypoglycaemia in types 1 and 2 diabetes: effects of treatment modalities and their duration. Diabetologia 2007;50:1140–7.

▶ DCCT. Epidemiology of severe hypoglycemia in the diabetes control and complications trial. The DCCT Research Group. Am J Med 1991;90:450–9.

▶ Pickup JC, Sutton AJ. Severe hypoglycaemia and glycaemic control in type 1 diabetes: meta-analysis of multiple daily insulin injections compared with continuous subcutaneous insulin infusion. Diabet Med 2008;25:765–74.

▶ Donnelly LA, Morris AD, Frier BM, et al. Frequency and predictors of hypoglycaemia in type 1 and insulin-treated type 2 diabetes: a population based study. Diabet Med 2005;22:749–55.


  1. A (T); B (T); C (F); D (F); E (F)

  2. A (F); B (T); C (F); D (F); E (F)

  3. A (T); B (T); C (F); D (F); E (F)

  4. A (F); B (T); C (F); D (F); E (T)

  5. A (T); B (F); C (T); D (F); E (F)



  • Competing interests None.

  • Provenance and peer review Commissioned; externally peer reviewed.