Article Text

Matching diabetes treatment and lifestyle
  • Relevant BNF section: 6.1


Strict glycaemic control reduces the likelihood of complications in patients with type 1 or type 2 diabetes mellitus.1,2 However, achieving such control as part of normal daily life can be difficult and may be further complicated by changes to daily routine. Here we highlight key ways of achieving good control without unduly restricting the patient's day-to-day activities. We also discuss the impact exercise, travel, changes in working patterns, and fasting can have on diabetic control and consider what practical advice healthcare professionals can offer to individuals facing such lifestyle changes.

Statistics from

  • Relevant BNF section: 6.1

Balancing control and flexibility

In published randomised controlled trials, the likelihood of microvascular complications in patients with type 1 or type 2 diabetes was reduced by improving control of blood glucose concentrations.1,2 However, in these studies, achieving tight glycaemic control was associated with a 2-3 fold increase in the likelihood of severe hypoglycaemia. Recurrent hypoglycaemia can have a major effect on a patient's life and may discourage him or her from aiming for tight control. It is difficult, therefore, to both optimise glucose concentrations and minimise the likelihood of hypoglycaemia.

Insulin therapy

A range of short-, intermediate- and long-acting insulin preparations are available, for use alone or in combination, to provide a variety of regimens for patients with diabetes. Conventional practice has utilised combinations of intermediate-acting human isophane insulin for basal supply and short-acting soluble insulin before meals. Twice-daily regimens using pre-mixed insulin are relatively inflexible since they require patients to have a regular, frequent and similar food intake at the same time each day in order to prevent hypoglycaemia. Flexibility can be increased with multiple-daily injections of insulin and using the Dose Adjustment For Normal Eating (DAFNE) educational program.3 Continuous subcutaneous infusion of short-acting insulin using a pump is an alternative for selected patients.4 Treatment options have been further increased with the introduction of the rapid-acting insulin analogues, lispro and aspart, which can be injected immediately before meals, and the long-acting analogues, ▼glargine and ▼detemir.5,6

Oral medication for diabetes

Conventional drug treatment for type 2 diabetes includes metformin, or a sulphonylurea such as glibenclamide or gliclazide. In selected patients, one of the thiazolidinediones, such as pioglitazone or rosiglitazone, may be added to such treatment, or used alone when conventional therapies are not tolerated. The meglitinides, nateglinide and repaglinide, have a rapid-onset and short-lasting stimulating effect on insulin secretion.7 On current evidence, they appear no better than older treatments in controlling glucose concentrations and their effects on clinical outcomes in the long term are unknown. However, their quick onset of action means they may have a role for some patients with irregular mealtimes.7

Effects of physical activity

Diabetes is not a barrier to regular physical activity, even for those who wish to compete in sport at international level.8 Indeed, for patients with diabetes, exercise helps to reduce excess weight, improve insulin sensitivity and glycaemic control, and reduce the likelihood of cardiovascular disease.9 However, exercise obviously covers a broad range of activities. Therefore, treatment needs to be tailored to the individual to accommodate the type, intensity, duration and frequency of exercise. This may require the involvement of a specialist in the management of diabetes.

Type 1 diabetes

Maintaining glucose homeostasis in people without diabetes requires complex metabolic adaptations to meet the energy needs of working muscle. Such a precise endocrine response is difficult to replicate in people with type 1 diabetes.10

Few studies have examined the effect of exercise on insulin requirements. In a crossover randomised controlled trial, 8 men with type 1 diabetes controlled on a regimen of ultralente basal insulin plus insulin lispro bolus injections before meals, received reduced pre-meal doses of insulin lispro before postprandial exercise of different intensities and durations.11 The study suggested that the likelihood of hypoglycaemia post exercise can be minimised by reducing the insulin dose. For example, 60 minutes of 25% maximal intensity exercise required a 50% reduction in insulin dose, and 60 minutes of 50% maximal intensity exercise required a 75% reduction in insulin dose. A non-randomised crossover study involving 13 patients with type 1 diabetes, treated with continuous subcutaneous infusion of insulin or multiple-daily injections of insulin, suggested that 50-66% reductions in insulin dose were necessary for exercise of moderate intensity.12 Particular caution should be taken with the use of the analogue insulin lispro because it may produce a more rapid and greater fall in plasma glucose than standard insulin.13

A single-blind crossover randomised controlled trial assessed acarbose 100mg, given with a meal 90 minutes prior to 30 minutes of 50% maximal aerobic exercise in 7 patients with well-controlled type 1 diabetes on a multiple-dose injection regimen.14 Compared with placebo, acarbose was associated with slower glucose absorption and a lower postprandial glucose rise but the two treatments did not differ in the incidence of hypoglycaemia.14

Type 2 diabetes

For patients with type 2 diabetes on insulin or sulphonylurea therapy, hypoglycaemia during physical activity is less likely to occur and exercise may actually improve insulin sensitivity.10 Metformin acts as an insulin-sensitiser and is rarely associated with hypoglycaemia when used as a single agent. The sulphonylureas stimulate pancreatic insulin release and may increase the likelihood of exercise-associated hypoglycaemia. However, there is a lack of comparative data to indicate whether the risk of hypoglycaemia is greater with any specific sulphonylurea. In a double-blind randomised controlled trial involving 167 patients with type 2 diabetes, glibenclamide 10mg once daily or glimepiride 3mg once daily caused similar degrees of blood glucose-lowering following 1 hour's exercise on a bicycle ergometer.15

In a non-blinded randomised controlled trial involving 37 patients with type 2 diabetes, those treated with Humalog Mix 25, containing 25% insulin lispro, showed a smaller decrease in postprandial plasma glucose concentration after exercise, compared with standard 30/70 mixed insulin.16 However, there was no significant difference between the groups in the incidence of hypoglycaemia (blood glucose concentration below 3.0mmol/L). Of note, both formulations of insulin were given 5 minutes before a standard breakfast, even though standard 30/70 mixed insulin should be given 30 minutes before a meal to minimise problems with early hyperglycaemia and late hypoglycaemia.

Advice for patients

Patients with diabetes, particularly those using insulin, need to know how their glycaemic control can be affected by different forms of physical activity. Patients' confidence can be increased by appropriate advice on checking blood glucose concentrations before and after activity, and carrying rapid- and intermediate-acting carbohydrate whenever they plan to exercise. For patients taking oral medication for diabetes, a reduction in the dose may be required if prolonged exercise is planned. For individuals with type 1 diabetes, guidelines from the American Diabetes Association recommend that, because of the risk of precipitating ketoacidosis, exercise should be avoided if the fasting glucose concentration before physical activity is more than 13.9mmol/L and ketosis is present.10 They also advise caution if glucose concentration is more than 16.7mmol/L and no ketosis is present.

Effects of travel

Overseas travel can pose particular problems for people with diabetes, particularly those on insulin. Confusion over timing of meals and insulin or other drug therapy can lead to hyper- or hypoglycaemia. One survey of insulin-treated patients with diabetes travelling to tropical countries showed that 68% developed metabolic disturbances, 26% had significant difficulties in adjusting their insulin regimen and 16% developed significant febrile illnesses.17 Only 36% of the travellers had increased how often they measured their blood glucose concentration. Advance planning can help to minimise risks, and patients should seek advice from a healthcare professional.

Travel across time zones

Treatment for diabetes might need to be adjusted given that air travel across time zones either abbreviates or extends the day, depending on the direction of travel. However, it is worth noting the results of one survey which suggests that many UK physicians, including diabetologists, may be uncertain about how to adjust doses of insulin for patients flying across several time zones.18 Eastward travel will shorten the day, and generally mean a temporary reduction in insulin doses, whereas westward travel will extend the day, and possibly increase insulin requirements. In general, however, adjustments to insulin doses are rarely necessary if patients are crossing fewer than five time zones.

Continuous subcutaneous infusion of insulin can provide a flexible regimen for managing diabetes during air travel. However, for most patients, multiple-daily injections of insulin will also allow flexibility with reduced or additional dosing. By contrast, conventional twice-daily fixed-mixture regimens may cause some problems as timing of meals and injections cannot be varied significantly. No published randomised controlled trial has compared different insulin regimens in travellers. However, given that tight glycaemic control is unnecessary during travel, changing insulin regimens (e.g. from twice-daily to multiple-daily injections) is unlikely to be necessary. Timing of oral medication is not as crucial as that for insulin. Metformin with meals should cause no significant problem. However, a twice-daily regimen with a sulphonylurea may require a dose adjustment to avoid hypoglycaemia.

Advice for the traveller

The traveller should be advised to tell their airline that they have diabetes in case of unexpected hypoglycaemia, and to carry a letter from a doctor confirming the need for insulin and needles on the plane. A request for specific 'diabetic' in-flight meals is unnecessary and, in any case, such meals may actually have insufficient carbohydrate and so might increase the likelihood of hypoglycaemia. A vegetarian meal that contains complex carbohydrate, such as rice or pasta, might be a better option. Travellers should be advised to store insulin in hand luggage and to carry extra carbohydrate and insulin in case of unexpected delays. Patients taking insulin should be advised to test blood glucose concentrations regularly, ideally every 4-6 hours. Advice on what to pack is shown in the box.

Please click here to view a box showing what to pack for air travel

Travellers should also be advised that glycaemic control can be relaxed during air travel and healthcare professionals should keep advice as simple as possible, with adjustments in regimens being restricted to those necessary to avoid hypoglycaemia and ketoacidosis.

Transportation and storage of insulin in hot countries can be problematic. In the UK, insulin remains active when left at normal room temperature for up to 28 days. However, in warmer climates, insulin should be refrigerated or, when carried around, kept in a cool bag. Insulin is absorbed more rapidly in hot weather or following physical activity. This can lead to hypoglycaemia, and travellers should be told about this risk.

If the traveller becomes unwell

Further advice should be given regarding management of diabetes when unwell. Gastroenteritis is common during foreign travel, and simple advice on managing this should be given. This includes advice on monitoring blood glucose concentration very frequently, checking urine for ketones, keeping well hydrated with water, and attempting to consume soups, snacks, milk or biscuits. If food cannot be eaten, sugary drinks should be drunk. Insulin and oral medication for diabetes should not be stopped under any circumstances, although doses may need to be decreased. Patients who are unable to eat or drink anything, and who are becoming ketotic, should be managed in hospital.

Effects of changing work patterns

Shift work can predispose an individual to metabolic syndrome and type 2 diabetes.19 Also, for people with pre-existing diabetes, shift working poses difficulties, particularly when working irregular shifts. In a prospective controlled observational study of 32 people with insulin-treated diabetes, glycaemic control was assessed before and 6 months after a change in shift.20 Overall, diabetic control was not significantly different when people were working shifts from when they worked days only. However, in those people who moved to a more rapidly rotating shift pattern, there was a significant deterioration in control.

Insulin glargine can be used as a basal insulin at any time of day. In theory, therefore, it can be given at the same time each day irrespective of work pattern. Short-acting analogues may also be useful in this circumstance because of the lower risk of postprandial hypoglycaemia, especially when mealtimes are unpredictable.

Patients with diabetes who work shifts should discuss their work patterns with a specialist in the management of diabetes. Frequent self-monitoring, and recording of blood glucose concentration, is of great importance to help determine patterns of blood glucose control, identify problem areas and plan ahead.

Adjusting mealtimes and treatment

Some modification in drug treatment and meal pattern may be necessary, as may a dietetic review. Such adjustments are easiest to make with a slowly rotating shift pattern. Meal planning is important, with meals evenly spaced at around 4-5 hour intervals and snacks if required, with consistent types and amounts of food. When working the afternoon shift, the only change to the meal plan that may be required is to include a larger evening snack to match increased physical activity, particularly if the person is on insulin. Working the overnight shift requires a significant adjustment in meal planning, allowing for the greater amount of food to be eaten during the shift, with an adequate amount of food for the periods of sleep during the day. Insulin adjustments are usually necessary when working shifts, although insulin can be delayed 1-2 hours without significantly affecting diabetic control. This may be all that is required to manage the afternoon shift. If extra food is consumed later in the day, a slight increase in the late-evening insulin dose may be needed. Other key considerations include ensuring availability of food, that meal breaks are at set times and that rapid-acting carbohydrate is available for treatment of hypoglycaemia. Also, increased physical activity can result in fluctuations in blood glucose control. If the work is physically demanding, particularly at unexpected times, there is a risk of hypoglycemia.

Effects of fasting

Fasting is undertaken by people of many religions and cultures. In people with diabetes, such a change in eating pattern without adjustment of therapy may worsen glycaemic control and result in complications, including hypoglycaemia, hyperglycaemia and ketoacidosis.

The effect of fasting on patients with diabetes is most often discussed with regard to Ramadan, the Islamic holy month in which many Muslims above the age of puberty abstain from eating, drinking, taking oral medicines or smoking between dawn and dusk. As the Muslim calendar is lunar, Ramadan comes forward by approximately 10 days per year and lasts for 29 or 30 days. During Ramadan, two meals are usually taken daily, one after sunset (Iftar) and one before dawn (Suhur). The duration of fasting depends on the season and geographical location; in the UK, fasting lasts for around 7-8 hours during winter and up to 18 hours in summer.

Advice for patients

People with diabetes should be advised not to fast. Muslims can be reassured that the Koran specifically exempts those with chronic illness from fasting, especially if fasting could lead to harmful consequences.21 Even so, many Muslims with diabetes do fast. In the Epidemiology of Diabetes and Ramadan 1422/2001 (EPIDIAR) survey of 12,243 people with diabetes living in 13 predominantly Muslim countries, 42.8% of those with type 1 diabetes and 78.7% with type 2 diabetes fasted for at least 15 days; the average number of days of fasting was 23 and 27 days, respectively.22 The overall rate of complications was low. However, the average number of episodes of severe hypoglycaemia per month was higher during Ramadan than in the preceding months for both patients with type 1 (0.14 vs. 0.03, p=0.0174) and type 2 diabetes (0.03 vs. 0.004, p<.0001). Severe hyperglycaemia or ketoacidosis was also more likely during Ramadan than in the preceding months in patients with type 2 diabetes (0.05 vs. 0.01 episodes/month, p<.0001).

Fasting patients with diabetes should monitor blood glucose concentrations frequently, particularly if they take insulin.21 The fast should be broken immediately if the fasting individual becomes hypoglycaemic (blood glucose concentration below 3.3mmol/L, or below 3.9mmol/L during the first hours of fasting). If blood glucose concentration exceeds 16.7mmol/L, the fast should also be broken to allow the concentration to be brought under control.

Patients with diabetes who fast should eat a healthy, balanced diet at all other times.21 In particular, they should avoid eating large amounts of food at the break of the fast. Instead, patients with diabetes should take two or three smaller meals in the non-fasting period. During Ramadan, the pre-dawn meal should be eaten as close to the start of the fast as possible. At this time, foods rich in complex carbohydrate are preferable as they are digested over a number of hours, whereas sugary foods are preferable at the break of fast.

Treatment changes

Evidence to support the use of any one specific treatment for patients with diabetes during fasting is limited. Of the few studies that have been published, most have either included small numbers of patients, or been non-blinded or uncontrolled.

Oral medications for diabetes

In patients taking metformin who fast, the likelihood of hypoglycaemia is low. However, some specialists suggest taking two-thirds of the total daily dose immediately before the main meal after sunset and one-third before the pre-dawn meal.

Sulphonylureas are more likely than metformin to cause hypoglycaemia, so care should be taken with their use during fasting. In a non-blinded non-randomised study involving 332 patients with type 2 diabetes fasting during Ramadan, glycaemic control was not adversely affected by changing the once-daily dose from morning to immediately before the sunset meal.23 However, some specialists recommend that as well as changing the timing of treatment, the dose should be halved.24 They also suggest that, for patients on a sulphonylurea twice daily, half the normal dose should be taken at the pre-dawn meal and the full dose taken at the main meal after sunset.24

A randomised controlled trial compared repaglinide with glibenclamide in 235 patients with type 2 diabetes who were fasting during Ramadan.25 A statistically significant reduction in mean serum fructosamine concentration from baseline was observed in the repaglinide group (-16.9?mol/L, p<.05) but not the glibenclamide group (-6.9?mol/L). There was no difference between groups in the number of hypoglycaemic episodes. Therefore, the evidence to support the use of repaglinide instead of a sulphonylurea during fasting is weak.

Insulin therapy

Analogue insulins have been compared to standard human insulins in fasting patients with diabetes. In a non-blinded cross-over randomised controlled trial involving 151 people, insulin lispro Mix 25 (Humalog Mix 25) improved fasting (7.1 vs. 7.5mmol/L, p=0.034) and post-prandial glycaemia (3.4 vs. 4.0mmol/L, p=0.007) compared with the human insulin 30/70.26 No significant difference in incidence of hypoglycaemia was noted between the two groups. Insulin lispro was compared with human insulin in 64 patients with type 1 diabetes in a non-blinded crossover randomised controlled trial.27 The 2-hour blood glucose excursion after the post-sunset meal was significantly lower with insulin lispro than with human insulin (p=0.026). The incidence of hypoglycaemia was lower with insulin lispro than with human insulin (23.4% vs. 48.4%, p=0.004), as was frequency of hypoglycaemic episodes (0.70 vs. 2.25 episodes/patient/30 days, p<.001). In a further non-blinded crossover randomised controlled trial of insulin lispro with soluble human insulin involving 70 patients with type 2 diabetes fasting during Ramadan, no significant difference in fasting glucose concentrations was seen between the two treatments.28 Also, there were no episodes of severe hypoglycaemia with either treatment, but postprandial glucose excursions were less with insulin lispro as were hypoglycaemic episodes (1.3 vs. 2.6 with soluble insulin, p<.002).28 Hypoglycaemia was defined as symptoms or signs of hypoglycaemia or a blood glucose concentration below 3.5mmol/L. However, blood glucose measurements were not routinely performed during the fasting period and the study design means that bias in frequency of testing for hypoglycaemia cannot be excluded.

While insulin lispro has theoretical advantages over standard insulin therapy, these data do not suggest a significant clinical advantage for its use during fasting.


Tight glycaemic control reduces the likelihood of microvascular complications in patients with type 1 or type 2 diabetes mellitus. However, it can be difficult to achieve, especially when normal daily routine is disrupted by, for example, exercise, travel, changing work patterns or fasting. Healthcare professionals have an important role in advising patients on coping with such changes. A flexible approach tailored to each individual is the most important aspect of care for such patients. Although, in theory, newer treatments, such as insulin analogues and postprandial glucose regulators might help, published evidence to support their use is lacking.


[M=meta-analysis; R=randomised controlled trial]

  1. R 1.
  2. R 2.
  3. R 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. R 11.
  12. 12.
  13. 13.
  14. R 14.
  15. R 15.
  16. R 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. R 25.
  26. R 26.
  27. R 27.
  28. R 28.
View Abstract

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.