Modification of adverse lifestyle factors is an important aspect of the management of both type 1 and type 2 diabetes. In particular, appropriate management of cardiovascular risk factors such as smoking, physical inactivity and poor diet is important for the prevention of macrovascular disease. Microvascular complications may also be affected by adverse lifestyle factors e.g. smoking. However, helping patients to modify certain behaviours must take account of other factors such as the patient's willingness to change, their perception of their diabetes, and other factors which may be related to their diabetes, such as depression and adverse effects on quality of life.

This section of the guideline has been divided into the following areas: delivery of lifestyle interventions; self-monitoring; quality of life and depression; and the specific areas of smoking, physical activity, healthy eating and alcohol. The recommendations in several of these areas are supported by evidence extrapolated from large studies conducted in the general population and these recommendations have been graded accordingly.

3.1 Delivery of lifestyle interventions

3.1.1 WHICH LIFESTYLE INTERVENTIONS HAVE BEEN SHOWN TO WORK IN DIABETES?

• Intensive interventions which include frequent contact with health professionals, telephone contact, multiple injections and self-monitoring have led to improvements in self-management.¹⁸
• Education which is supplemented by additional support/follow-up and behaviour modification may result in improvements in metabolic and psychosocial outcomes.^{49, 50, 51}
• Computer-assisted programmes which provide education and trigger self-management have a proven benefit in terms of both metabolic and psychosocial outcomes.^{52, 53}
• Psychological interventions which are varied and include behaviour modification, motivational interviewing, patient empowerment and activation have a positive impact on outcomes. However, the interventions vary greatly and it is not possible to state which element is of proven value.
• Interventions based on a theoretical model or knowledge base have better outcomes.
  Evidence level 1++, 1+, 2++

Patients with diabetes should be offered lifestyle interventions based on a valid theoretical framework.

Education programmes, computer-assisted packages and telephone prompting should be considered as part of a multidisciplinary lifestyle intervention programme.

There is no evidence of a benefit for interventions based in secondary care over those based in primary care. No evidence was identified which addresses long term follow up in educational interventions.

The evidence of the role health beliefs play in diabetes self management is equivocal.

Telephone or postal reminders prompting people with diabetes to attend clinics or appointments are an effective method of improving attendance.⁵⁴ Evidence level 4

3.1.2 TRAINING HEALTH PROFESSIONALS TO DELIVER LIFESTYLE INTERVENTIONS

Patient satisfaction and knowledge improve when lifestyle interventions are delivered by primary care staff who have been trained to take a patient-centred approach.⁵⁵ Evidence level 1+

One study indicated that primary care nurses in contact with diabetes nurse educators are more knowledgeable about diabetes than nurses with no specific training in diabetes, and provide a higher standard of care.⁵⁶ Evidence level 3

Health care professionals should receive training in patient-centred interventions in diabetes.

3.2 Self-monitoring of glycaemic control

The literature in this area is difficult to assess. Many of the studies cannot be compared as the patient groups were different and glucose monitoring was usually just one part of a multifactorial intervention programme.⁵⁷ However, a comprehensive package of care which includes glucose self-monitoring is usually effective in improving glycaemic control in type 1 diabetes.

No studies have adequately assessed the benefits of glucose monitoring on glycaemic control, or the relative benefits of blood glucose monitoring vs. urine testing. In general, urine testing is less costly than blood testing, however the preferred method of glucose monitoring varies according to type of diabetes. Some patients with type 2 diabetes prefer urine testing while patients with type 1 diabetes appear to favour blood testing.

3.3 Quality of life and depression

Quality of life issues and depression are important factors which may influence how patients are able to manage their diabetes.

3.3.1 DEPRESSION AND DIABETES

Depression is more common in people with diabetes than in the general population.^{58, 59} The presence of microvascular and macrovascular complications are associated with a higher prevalence of depression and lower quality of life.^{60, 61, 62} Remission of depression is often associated with an improvement in glycaemic control.^{58, 62} Evidence level 1+, 2++, 4

Antidepressant therapy with a selective serotonin reuptake inhibitor (SSRI) is a useful treatment in depressed patients with diabetes and may improve glycaemic control,⁶³ however tricyclic antidepressants may adversely affect metabolic control.⁶⁴ Evidence level 1+

Cognitive behavioural therapy (CBT) is a psychological treatment which attempts to find links between the person's feelings and the patterns of thinking which underpin their distress. CBT, psychotherapy programmes and coping skills training are useful in treating depression in patients with diabetes. ^{65, 66, 67} However, cognitive behavioural therapy is less effective in patients with complications.⁶² Evidence level 1++, 1+, 2++

The management of patients with post-natal depression is considered in a forthcoming SIGN guideline.

Health care professionals should be aware of the effects of depression on diabetes.

All people with diabetes should be screened for depression and offered appropriate therapy.

SSRIs are recommended in preference to tricyclic antidepressants for treatment of depression in patients with diabetes.

There is some evidence that negative life events are associated with poorer diabetic control.⁶⁸ Evidence level 2+

Health care professionals should be aware of the potential effects of life events on stress and self-care behaviour.

3.3.2 DIABETES CONTROL AND QUALITY OF LIFE

Severe hypoglycaemia may adversely affect quality of life in patients treated with insulin, particularly in those newly diagnosed. Improvements in blood glucose control are associated with improvements in quality of life, providing there is no increase in hypoglycaemic symptoms.^{69, 70} Frequency of insulin dose adjustment does not appear to affect quality of life.^{17, 69, 70, 71} Evidence level 1+

Patients and health care professionals should make every effort to avoid severe hypoglycaemia, particularly in those who are newly diagnosed.

3.4 Smoking cessation

Smoking is an established risk factor for cardiovascular and other diseases. However, there is conflicting evidence regarding the effect of smoking on glycaemic control.

3.4.1 ASSESSMENT OF READINESS TO CHANGE SMOKING BEHAVIOUR

Standard models for measuring stages of change (pre-contemplation, contemplation, preparation, action, maintenance and relapse) have been used to assess readiness to quit smoking. There is some evidence that interventions aimed at discussing the benefits of quitting smoking may be most useful for pre-contemplators and contemplators, whereas interventions aimed at improving self-efficacy may be more useful for those preparing to quit.^{72, 73} Evidence level 3

A model using stages of change may help health care professionals understand how ready an individual is to quit smoking.

3.4.2 FIRST LINE TREATMENTS

Simple advice to stop smoking given by a physician, a nurse or a counsellor has a small but significant effect (absolute quitting rate is increased by 2.5-14.7%).^{74, 75, 76} Increasing the intensity of the advice is marginally more effective. Group behaviour therapy is more effective than self-help material but has not been proven to be superior to individual advice.^{74, 77} Evidence level 1++

Health care professionals involved in caring for patients with diabetes should advise them not to smoke.

Nicotine replacement therapy (NRT) is effective in increasing the rate of quitting by 1.5 to 2 times.⁷⁸ All the commercially available forms of replacement (gum, patch, nasal spray, inhaler and sublingual tablets) have broadly similar efficacy. The absolute size of effect will depend on the setting. There is no evidence of the benefit of NRT in those smoking less than 15 cigarettes per day. Highly dependent smokers may benefit more from NRT and may need a higher dose. Eight weeks of patch therapy has been shown to be as effective as longer duration of therapy.⁷⁸ Evidence level 1++

Nicotine replacement therapy should be provided for smokers of more than 15 cigarettes per day who are trying to quit. Therapy in a form acceptable to the patient should be offered for up to eight weeks.

Bupropion increases the rate of smoking cessation.⁷⁹ Combination of this with a nicotine patch is more efficacious than using a patch alone. The two studies which demonstrated these effects gave therapy for one week before quitting and for seven or eight weeks after stopping smoking.^{80, 81} Combination of bupropion with nicotine patch increased blood pressure in some patients. Evidence level 1++, 1+

The above studies were not specifically of people with diabetes. The summary of product characteristics recommends a lower dose of bupropion in patients on oral hypoglycaemic agents or insulin, as there is a greater risk of seizure.

Bupropion therapy (in the absence of contraindications) could be used alone or with nicotine replacement, if blood pressure is monitored.

3.4.3 OTHER TREATMENTS

Clonidine and nortriptyline can increase rates of smoking cessation, however there are potential side effects with clonidine use. ^{79, 82} Evidence level 1++

Other therapies which may be considered include clonidine and nortriptyline, however care should be taken to monitor for adverse effects.

Acupuncture and silver acetate treatment are ineffective interventions in smoking cessation.^{83, 84} Evidence level 1++

Acupuncture or silver acetate should not be used as part of a smoking cessation strategy.

The evidence on hypnotherapy, nurse counselling plus support, and rapid smoking aversive therapy in smoking cessation in patients with diabetes is of too poor quality to support recommendations.

3.4.4 MONITORING

Relapse to smoking remains a problem even in those patients who have successfully quit at one year. The relapse rate has been recorded as 23-40%.^{85, 86} Evidence level 1+

Health care professionals should continue to monitor smoking status in all patient groups.

3.5 Exercise and physical activity

3.5.1 DEFINITIONS

Physical activity is defined as any skeletal muscle movement which expends energy beyond resting level (e.g. walking, gardening, stair climbing).

Exercise is a subset of physical activity which is done with the goal of enhancing or maintaining an aspect of fitness (e.g. aerobic, strength, flexibility, balance body mass index). It is often supervised (e.g. in a class), systematic and regular (e.g. jogging, swimming, attending exercise classes).

3.5.2 EFFECTS OF PHYSICAL ACTIVITY ON THE PREVENTION OF DIABETES

Regular physical activity is associated with a reduced risk of development of type 2 diabetes. This risk reduction is consistent over a range of intensity and frequency of activity, with a dose-related effect. Greater frequency of activity confers greater protection from development of type 2 diabetes and this is valid for both vigorous and moderate intensity activity. The length of time to confer the effect is greater than one year and, on current evidence, requires a minimum of four years.^{87, 88, 89, 90, 91, 92, 93} Evidence level 2++, 2+

The Diabetes Prevention Program (DPP) is a major study currently in progress to determine whether intensive lifestyle intervention or treatment with metformin delays or prevents the onset of diabetes. Preliminary results indicate a substantial decrease in progression from IGT to diabetes in patients who follow a programme of intensive lifestyle management.⁹⁴

All people should be advised to maintain at least moderate levels of physical activity (e.g. daily walking) as a lifelong lifestyle modification.

3.5.3 ASSESSMENT OF PHYSICAL ACTIVITY

Physical activity is a very difficult behaviour to measure since it incorporates mode of activity, duration, frequency and intensity. There is no gold standard and techniques range from heart rate monitoring to motion counters and self-reports. Self-report is the easiest format but there is often an over reporting of minutes spent in activity. The Scottish Physical Activity Questionnaire is an example of one self-report format that has known validity and reliability for assessing moderate activity.⁹⁵ As with smoking cessation (see section 3.4), it is important in assessing what kind of support a patient needs for increasing or maintaining physical activity to know their stage of change. A rate of perceived exertion scale is useful for estimating exercise intensity, particularly in people with autonomic neuropathy who have reduced maximal heart rate.⁹⁶ Evidence level 4

3.5.4 PHYSICAL ACTIVITY AND EXERCISE FOR PEOPLE WITH DIABETES

Various guidelines exist for physical activity and exercise in the general population. For example, for aerobic fitness a minimum of 20 minutes of continuous aerobic exercise at least 50% of maximal aerobic capacity (which would equate to brisk walking for people with low fitness levels) on three days each week is recommended.⁹⁷ The Health Education Board for Scotland (http://www.hebs.co.uk) recommends a two-stage approach. The first stage is to encourage sedentary people to accumulate moderate physical activity for 30 minutes on most days of the week. The second stage is to encourage those who are interested, motivated and already active to engage in more vigorous activity at least three days of the week. Evidence level 4

In people with type 2 diabetes physical activity or exercise should be performed at least every second or third day to maintain improvements in glycaemic control. In view of insulin adjustments etc. it may be easier for people with type 1 diabetes to perform physical activity or exercise every day. ⁹⁸ Evidence level 4

Aerobic, endurance exercise is usually recommended, however resistance training with low weights and high repetitions is also beneficial.⁹⁹ Evidence level 1+

Exercise and physical activity (involving aerobic and/or resistance training) should be performed on a regular basis.

No trial-based evidence was identified which described how to promote physical activity for patients with diabetes. Expert opinion suggests using social-cognitive models and making advice person-centred and diabetes specific.¹⁰⁰ Evidence level 4

Advice about exercise and physical activity should be individually tailored and diabetes-specific and should include implications for glucose management.

Evidence from the non-diabetic population suggests that teaching CBT skills, tailoring advice to stage of exercise behaviour change and providing on going support (for all stages) will enhance long term adherence.¹⁰¹ The most appropriate mode of activity for adherence is home based, individual, lifestyle exercise of moderate intensity (i.e. activity that is incorporated into daily life such as walking, gardening or stair climbing).^{102, 103} Continual support appears to be required to maintain adherence, however the intensity of support required is as yet unknown.¹⁰⁴ Evidence level 1+ and 4

To maximise adherence, exercise programmes should be home-based and should be accompanied by ongoing support which includes education in cognitive behaviour skills and advice tailored to the individual's stage of change.

3.5.5 ADVICE FOR PATIENTS TAKING INSULIN OR ORAL ANTIDIABETIC DRUGS

Exercise with normal insulin dose and no additional carbohydrate significantly increases the risk of hypoglycaemia during and after exercise. If exercise can be anticipated, a reduction of the normal insulin dose (by up to 65% for vigorous exercise for up to 45 minutes) will significantly reduce the risk of hypoglycaemia and delayed hypoglycaemia.¹⁰⁵ Evidence level 2+

The amount of reduction in insulin dose will depend on duration and intensity of exercise being performed, insulin and glycaemic level before exercise, and the time of day. If exercise can not be anticipated and insulin dose has already been taken, extra carbohydrate before exercise will reduce the risk of hypoglycaemia.

Injection of insulin into exercising areas increases the absorption of insulin and the risk of hypoglycaemia and should therefore be avoided.^{105, 106, 107} Evidence level 2+

Individualised advice on avoiding hypoglycaemia when exercising by adjustment of carbohydrate intake, reduction of insulin dose, and choice of injection site, should be given to patients taking insulin.

High temperatures can also increase insulin absorption. This should be taken into consideration when exercising in hot climates. A further reduction in insulin dose may be required.¹⁰⁵ Evidence level 4

Patients using oral antidiabetic drugs, such as sulphonylureas, may also be at risk of hypoglycaemia during exercise.

3.5.6 DIABETIC COMPLICATIONS AND EXERCISE

There is no known association between exercise participation and development or exacerbation of diabetic complications, however exercise during insulin deficiency can cause hyperglycaemia.¹⁰⁵ Evidence level 4

Research demonstrates that high intenisty exercise may transiently increase albumin excretion rate in people with or without diabetes. No evidence of more rapid progression of nephropathy or retinopathy was identified in subjects with diabetes who exercise more.^{108, 109} However, in theory, haemodynamic changes which accompany high intensity exercise could have an adverse effect on microvascular disease. Evidence level 2+

A joint position statement from the American Diabetes Association and American College of Sports Medicine^{97, 98} recommends that patients with diabetes planning moderate to high intensity exercise should undergo graded exercise testing if one or more of the following criteria apply:

• age >35 years
• type 2 diabetes >10 years duration
• type 1 diabetes >15 years duration
• presence of any additional risk factor for cardiovascular disease (CVD)
• presence of microvascular disease (retinopathy or nephropathy, including microalbuminuria)
• peripheral vascular disease
• autonomic neuropathy.
  Evidence level 4

Graded exercise testing is not standard clinical practice in the UK, however, it can provide useful information if time and resources allow.

There is higher risk of myocardial infarction (MI) after heavy exertion in sedentary compared with non-sedentary people with type 1 diabetes.¹¹⁰ Evidence level 2+

Patients with existing complications of diabetes should seek medical review before embarking on exercise programmes.

A gradual introduction and initial low intensity of physical activity should be recommended for sedentary people with diabetes.

3.6 Healthy eating

3.6.1 RECOMMENDED DIET FOR PEOPLE WITH DIABETES

Healthy eating is of fundamental importance as part of diabetes health care behaviour and has beneficial effects on weight, metabolic control and general well-being. In particular, weight control in overweight subjects with diabetes is associated with improved glycaemic control.^{111, 112} Salt restriction in the general population is discussed in the SIGN guideline on lipids and the primary prevention of coronary heart disease (SIGN 40).¹¹³ Dietary recommendations, including dietary constituents for healthy eating and weight control in patients with diabetes, are summarised elsewhere.^{114, 115} Evidence level 1+, 4

3.6.2 DIETARY INTERVENTIONS TO PREVENT THE ONSET OF DIABETES

There is conflicting evidence for the role of specific dietary intervention programmes. Studies either show a beneficial effect or no effect, but there is no evidence of a harmful effect. Most recently, one large trial from Finland demonstrated a short term reduction in the development of type 2 diabetes in high risk subjects (overweight and impaired glucose tolerance) by encouraging lifestyle change, including diet and exercise advice. It is not possible to determine which aspects of the programme were successful.¹¹⁶ However, other studies have demonstrated that if people who are overweight lose weight, by whatever method, their risk of developing diabetes is reduced.^{89, 112, 117, 118, 119} Evidence level 1+

Overweight individuals and those at high risk of developing diabetes should be encouraged to reduce their risk by lifestyle change.

3.6.3 ASSESSMENT OF DIET IN CLINICAL PRACTICE

There are few studies which have examined the validity of dietary assessments in clinical practice, particularly in patients with diabetes. Self-report questionnaires have been developed and are currently being validated.^{120, 121, 122, 123, 124, 125} The most accurate form of dietary assessment is the seven day weighed food record, although this is impractical in the clinical setting. Assessment of diet over shorter periods is less accurate.¹²⁶ Evidence level 4

3.6.4 ASSESSING READINESS TO CHANGE DIETARY BEHAVIOUR

The stages of change (transtheoretical) model is valid when assessing dietary behaviour.¹²⁷ Questionnaires to assess the stage of change of a patient are easily administered in clinical practice. ^{128, 129} Evidence level 3

Before giving dietary advice to patients with diabetes, assessment of readiness to change diet behaviour should be undertaken.

3.6.5 ENCOURAGING DIETARY CHANGE IN CLINICAL PRACTICE

The use of a behavioural approach to dietary interventions in patients with diabetes shows clinically significant benefit in terms of weight loss, HbA1c, lipids, and self-care behaviour for up to two years after the initial intervention^{130, 131, 132, 133, 134, 135, 136} However, it is not always possible to identify if the benefit is wholly attributable to the intervention, or dependent on how or where the care is delivered. Evidence level 1+

Intensive therapy or contact in patients with diabetes shows clinically beneficial effects on weight and glycaemic control during the period of intervention. More education and contact appears to improve outcomes.^{137, 138, 139, 140, 141} Pre-packaged meal programmes show significant clinical benefit in terms of weight, blood pressure, glycaemic control and lipids during the study period but are impractical outside the trial setting.^{142, 143, 144, 145} Evidence level 1+

Clinical interventions aimed at dietary change are more likely to be successful if a psychological approach based on a theoretical model is included.

3.7 Alcohol

In people with type 1 diabetes, drinking 2-3 glasses of wine at one time in the rested state has no significant effect on blood glucose up to 10 hours after consumption.^{146, 147, 148} Evidence level 1+

In people with type 2 diabetes, drinking 2-3 glasses of wine or an equivalent quantity of beer may result in a non-significant decrease in blood glucose, but no increased risk of hypoglycaemia ^{149, 150, 151} However, if patients with type 2 diabetes exercise after drinking alcohol blood glucose may be lowered by up to 27% from baseline, but in the laboratory situation there was no increased risk of developing hypoglycaemia.¹⁵⁰ Evidence level 1+

All patients with diabetes should be aware of the high calorific value of alcohol and the implications of excess consumption on body weight.

Patients with diabetes should be advised that they may drink up to 3 units of alcohol with a minimal effect on blood glucose. Patients should be advised that if execise and consumption of alcohol are combined there may be a greater lowering of blood glucose. *

* The grade of recommendation has been adjusted due to setting and sample sizes of trials

Excess alcohol consumption is associated with a worsening in general health and can lead to weight gain, reduced fertility and memory loss. As in the general population, alcohol consumption should be limited to 3-4 units per day in men and 2-3 units per day in women.