Exercise and diabetes


  • Disclosures: No conflicts of interest.

  • Endorsed by the International Conference on ATTD organized by Kenes International.

Howard Zisser,
Sansum Diabetes Research Institute, Santa Barbara, CA, USA
Tel.: +1-805-682-7638
Fax: +1-805-682-3332
Email: hzisser@sansum.org


Diet and exercise form the foundation of a healthy lifestyle. These are especially important for people living with diabetes mellitus, as they are the most practical non-pharmacological means by which patients may significantly improve their blood glucose levels. Exercise increases insulin sensitivity (both short and long term), lowers blood sugar levels, reduces body fat and improves cardiovascular (CV) function. Because of this, exercise offers enormous benefit to patients with diabetes.

Blood glucose levels can significantly drop during and after physical activities, due to the increased utilisation of glucose as a fuel during exercise and the up-regulation of glucose transport into working muscles. Therefore, patients (especially those with type 1 diabetes) must account for the effects of exercise and adjust their medications and nutrition accordingly. Improvements in real-time continuous glucose monitoring and optimisation of basal insulin dosing may offer significant benefit to preventing hypoglycaemia in patients with type 1 diabetes who regularly exercise.

Diverse exercise programmes and devices can also assist patients in monitoring their activities as well as motivating them to achieve their exercise goals. For patients with type 1 diabetes, questions such as how much, how long, how strenuous and what kind of exercise must be addressed in order for healthcare professionals to offer maximum benefit to their patients. Additionally, since patients with type 2 diabetes often have other significant co-morbidities such as obesity and CV disease, care providers must evaluate each patient’s risk factors before designing an exercise programme.

Several publications in the last year have addressed these issues and may serve as a valuable resource to provide safe and effective recommendations to patients and their healthcare providers.

To be included in the Exercise and Diabetes chapter for the 2010 YEARBOOK, we reviewed leading peer-reviewed manuscripts that were published in the period July 2009 to June 2010. PubMed was used in the initial screening of articles.

Plasma glucose and hypoglycaemia following exercise in people with type 1 diabetes: a comparison of three basal insulins

V. Arutchelvam,1T. Heise,2S. Dellweg,2B. Elbroend,3I. Minns,3P. D. Home1

1School of Clinical Medical Sciences – Diabetes, Newcastle University, Newcastle upon Tyne, UK,2Profil Institut für Stoffwechselforschung, Neuss, Germany, and3Novo Nordisk, Crawley, UK

Diabet Med 2009; 26: 1027–32

Background: Exercise changes the absorption and blood-glucose-lowering effectiveness of different types of insulin. This study aims to compare the effects of exercise on plasma glucose (PG) excursions in people with type 1 diabetes when using each of three basal insulins: insulin detemir, neutral protamine Hagedorn (NPH) insulin or insulin glargine.

Methods: Fifty-one people with type 1 diabetes (age ± SD, 39 ± 10 years; 67% men) were enrolled in a multinational, open-label, randomised, three-period, crossover clinical trial sponsored by Novo Nordisk. Subjects were managed with insulin regimens during three periods: one period with mealtime insulin (aspart) plus basal insulin detemir (twice daily), one period with aspart and NPH (twice daily), and the other period with glargine (once daily). Before the start of exercise programmes, insulin doses were optimised over 4 weeks. In each treatment arm, subjects were exercised for 30 min, 5 hours after the last mealtime and basal insulin injection.

Results: At initiation of exercise, PG was numerically but not statistically different for each insulin, with a modestly lower level for glargine by a mean of 0.7 mmol/l. During exercise or 150 min after exercise, subjects did not show significant differences in PG excursions. During the 30 min of exercise, five subjects (11%) on detemir, six (12%) on NPH and 18 (38%) on glargine developed minor hypoglycaemia. From the end of exercise to 150 min, five subjects (11%) on detemir, seven (14%) on NPH and nine (19%) on glargine developed minor hypoglycaemia. In total, from start of exercise to 150 min after exercise, 10 subjects (12%) on detemir developed minor hypoglycaemia compared with 13 (27%) on NPH and 27 (57%) on glargine (p < 0.001 glargine vs. detemir and NPH). Subjects on detemir and NPH had lower maximum plasma cortisol levels than glargine.

Conclusions: Insulin detemir was associated with less hypoglycaemia than insulin glargine but not NPH insulin during and after exercise in relatively well-controlled people with type 1 diabetes.

  • Comment: One drawback associated with taking a basal insulin is the fact that once it is delivered there is no turning it off. It will continue to deliver active insulin until it is fully metabolised. This presents a problem if one wishes to interrupt basal insulin delivery in the event of planned or unplanned exercise or in the event of hypoglycaemia. If one were using a pump to deliver the insulin, one would be able to reduce or suspend insulin delivery as needed. Under the conditions of this study (i.e. exercising 5 h post insulin injection for ≈30 min), it would appear that detemir offers more protection against hypoglycaemia than NPH or glargine insulin in multiple daily injection patients and this may be an important distinction for those considering the choice of basal insulin for active individuals. Variability of insulin absorption and timing of insulin injection relative to when the exercise occurred probably had significant effects on PG concentrations following exercise in these subjects, however. In addition, the manufacturer of insulin detemir, Novo Nordisk, supported this study; two of the authors are also employees of Novo Nordisk. This may be a source of potential bias in generating the conclusion that detemir is associated with less hypoglycaemia provoked by exercise in people with type 1 diabetes.

The ageing athlete: screening prior to vigorous exertion in asymptomatic adults without known cardiovascular disease

J. Freeman, V. Froelicher, E. Ashley

Division of Cardiovascular Medicine, Stanford University School of Medicine, Stanford, CA, USA

Br J Sports Med 2009; 43: 696–701

Background: This paper reviews the current literature on screening for CV conditions in order to determine the best form of screening for asymptomatic adults without known CV disease prior to the initiation of a vigorous exercise programme.

Methods: The study offers a review of various screening studies, including ECG sensitivity testing, ECG screenings in subpopulations (men over 45, women over 55, asymptomatic adults with CV risk factors, asymptomatic adults with diabetes) and the cost effectiveness of various forms of pre-exercise screening tests.

Results: Studies have shown that there are benefits of ECG as a screening method in various subpopulations. Exercise ECG testing is currently considered the best available method for screening asymptomatic adults, despite the fact that there have been no studies that demonstrate a significant impact of exercise ECG on morbidity and mortality of completely asymptomatic patients.

Conclusions: Given its test characteristics, widespread availability and cost, and due to an otherwise lack of conclusive evidence on the subject, the authors recommend exercise ECG screening before beginning a vigorous exercise programme for asymptomatic men with diabetes and asymptomatic men over age 45 with two or more CV risk factors. Consideration should also be given to screening asymptomatic middle aged and elderly patients with fewer than two risk factors and asymptomatic patients younger than 45 with strong risk factor exposure.

• Comments: Although diabetes is not the focus of this study, it is nevertheless a relevant paper given the higher risk of CV complications in people with diabetes relative to the general population. Understanding which form of CV screening is most effective and most cost effective will have an important effect on people with diabetes and could help them to avoid and prevent the serious CV complications that can often arise. Although a number of professional organisations including the American College of Sports Medicine and the American Diabetes Association recommend exercise stress tests with ECG prior to the onset of any new exercise regimen in older individuals with diabetes, clear evidence for these recommendations is rather scarce. This paper helps make the case that these recommendations are indeed reasonable, although one would hope that delays in screening would not be a major barrier to getting patients with diabetes more physically active.

Continuous, non-invasive measurement of the haemodynamic response to submaximal exercise in patients with diabetes mellitus: evidence of impaired cardiac reserve and peripheral vascular response

D. Joshi,1A. Shiwalkar,1M. R. Cross,2S. K. Sharma,3A. Vachhani,1C. Dutt1

1Torrent Research Centre, Village Bhat, Gandhinagar, Gujarat, India,2Veeda Clinical Research, Old Convent of Notre Dame, Derriford, Plymouth, UK, and3Veeda Clinical Research, Ambawadi Ahmedabad, Gujarat, India

Heart 2010; 96: 36–41

Background: Impaired supply of oxygen and nutrients, resulting from limitations in cardiac function and microvascular dysfunction, to exercising muscles is thought to account for reduced exercise capacity in patients with diabetes. The authors studied the changes in cardiac function and microvascular utilisation during exercise in individuals with type 2 diabetes compared with age-matched controls during exercise and recovery.

Methods: This study consisted of 101 subjects divided into three age-matched groups: HbA1c < 8 (= 31), HbA1c ≥ 8 (= 38) and control subjects without type 2 diabetes (= 32). Subjects were continuously monitored while they exercised on a bicycle ergometer at a constant output of 50 W for 10 min, followed by recovery. Cardiac function was measured by impedance cardiography, microvascular flow by laser Doppler shift and haemoglobin oxygen saturation by white light spectroscopy.

Results: Cardiac reserve, as measured by capacity to maximally increase cardiac output during exercise, was significantly less in diabetic patients than the control group (p < 0.001). Maximal rise in stroke volume, the key parameter limiting elevation of cardiac output, was significantly reduced in patients with diabetes, more so in the group with HbA1c ≥ 8 (p > 0.002). Cardiac output during recovery, while greater than resting baseline in all groups, was also lower in the patients with diabetes and significantly lower when comparing the higher A1c group with normal subjects (p < 0.05). Regional blood flow to exercising muscles was significantly reduced in the group with HbA1c ≥ 8 (p < 0.05). Oxygen desaturation during exercise was greater in the patients with diabetes, and was more pronounced in the group with HbA1c ≥ 8 (p < 0.05). Oxygen debt during recovery was also significantly greater in both groups of patients with diabetes (p < 0.05) and regional flow during recovery was greater compared with the control group.

Conclusions: Cardiac response and skeletal muscle profusion during and after submaximal exercise are reduced in patients with diabetes. Thus, patients with diabetes experience greater oxygen desaturation and oxygen debt. These changes may be responsible for reduced exercise capacity observed in patients with diabetes.

  • Comments: Previous studies have reported reduced exercise capacity in patients with type 2 diabetes without coronary heart disease (1) and several studies have demonstrated a correlation between left ventricular dysfunction and impaired exercise capacity in patients with type 2 diabetes (2). In this study, patients with poorer glycaemic control had more severe cardiac and peripheral dysfunction. This is consistent with previous investigations done by others in which A1c was shown to have an inverse correlation with maximum oxygen uptake (3), work capacity (4) and exercise duration (5). Patients with type 2 diabetes also had an impaired response during the recovery period, which appears to have caused accumulation of a large oxygen debt after exercise. There are numerous proposed mechanisms underlying impaired exercise response in patients with type 2 diabetes. These include both systolic and diastolic abnormalities, endothelial dysfunction, altered sympathetic activity stemming from abnormal insulin signalling and widespread CV dysfunction due to the proinflammatory and procoagulant effects of advanced glycation end-products. This confirms the fact that exercise is a very useful tool to quantify the CV limitations of patients with diabetes.

Exercise training ameliorates the effects of rosiglitazone on traditional and novel cardiovascular risk factors in patients with type 2 diabetes mellitus

N. P. E. Kadoglou,1,2F. Iliadis,3N. Sailer,1Z. Athanasiadou,1I. Vitta,1A. Kapelouzou,4P. E. Karayannacos,4C. D. Liapis,5M. Alevizos,3N. Angelopoulou,6I. S. Vrabas2

1First Department of Internal Medicine, ‘Hippokratio’ General Hospital of Thessaloniki, Greece,2Department of Physical Education and Sports Science (Serres), Aristotle University of Thessaloniki, Greece,3First Propedeutic Department of Internal Medicine, AHEPA University Hospital of Thessaloniki, Greece,4Center of Experimental Surgery, Biomedical Research Foundation, Academy of Athens, Greece,5Department of Vascular Surgery, Medical School, University of Athens, Greece, and6Department of Physical Education and Sports Science, Aristotle University of Thessaloniki, Greece

Metabolism 2010; 59: 599–607

Background: Type 2 diabetes mellitus predisposes one to CV diseases. Recently, rosiglitazone (RSG), a member of a class of insulin-sensitizing drugs, thiazolidinediones, has been suggested to increase the risk of ischaemic heart events and CV mortality (6). The present study investigated the effects of RSG and/or exercise training on novel CV risk factors in patients with type 2 diabetes.

Methods: One hundred overweight/obese individuals with type 2 diabetes, with HbA1c > 7% despite combined treatment with gliclazide and metformin, were randomised into four groups (each = 25): (1) control group: maintenance of habitual activities; (2) RSG group: add-on RSG therapy (8 mg/day); (3) exercise group: adjunctive exercise training; and (4) RSG + exercise group: RSG therapy (8 mg/day) plus exercise training. No participant had diabetic vascular complications or was receiving lipid-lowering therapy. Anthropometric parameters, cardiorespiratory capacity, glycaemic and lipid profile, apolipoprotein (apo) A-I, apo B, interleukin (IL) 10, IL-18, insulin resistance and blood pressure were measured before and after 12 months of intervention.

Results: Compared with the control group, both the RSG and exercise groups experienced significantly (p < 0.05) reduced glycaemic indexes, insulin resistance, blood pressure and IL-18. RSG and exercise groups showed significantly (p < 0.05) increased high-density lipoprotein, cardiorespiratory capacity and IL-10. Exercise-treated participants presented a notable down-regulation in the rest of the lipid parameters (total cholesterol, low-density lipoprotein, cholesterol, triglycerides, apo B) and body fat content (p < 0.05) compared with the control group. In contrast, the RSG group in comparison with the control group showed remarkably increased apo A-I levels and body mass index (p < 0.05). Furthermore, the RSG + exercise group (vs. control group) yielded prominent beneficial changes in glycaemic indexes, lipid profile, insulin resistance, blood pressure, IL-10, IL-18, apo A-I and apo B (p < 0.05). Last, but not least, exercise in the RSG + exercise group counteracted the negative effects of RSG on body weight, low-density lipoprotein and total cholesterol.

Conclusions: RSG was best taken in combination with exercise, as they both elicited complementary benefits on body composition, glycaemic control and novel CV risk factors in type 2 diabetes mellitus patients.

  • Comments: RSG is reported in a recent study on elderly diabetic patients to be associated with an increased risk of such serious CV events as stroke and heart failure compared to pioglitazone, the other thiazolidinedione marketed in the USA (7). On the other hand, regular aerobic exercise is known to enhance insulin sensitivity, promote cardiorespiratory functions and reduce CV risks in patients with pre-diabetes and diabetes (8–10). The study of exercise combined with RSG in the treatment for type 2 diabetes provides a new prospective in evaluating the advantages and disadvantages of current RSG therapies. Unfortunately, probably because of the short nature of the study and the small sample size, cardiac events during the study were not reported. Importantly, however, using RSG alone was shown to increase weight gain, which could have a detrimental effect on an individual’s health and self-esteem, but exercise countered this effect in addition to raising markers associated with high-density lipoprotein (apo A-I) and decreasing those associated with low-density lipoprotein (apo B). It is interesting to note that exercise alone induced a greater decrease in total cholesterol and low-density lipoprotein than did the combined therapy of RSG and exercise.

Pedometer use among adults at high risk of type 2 diabetes, Finland, 2007–2008

E. E. Korkiakangas, M. A. Alahuhta, P. M. Husman, S. Keinänen-Kiukaanniemi, A. M. Taanila, J. H. Laitinen

Finnish Institute of Occupational Health, Oulu, Finland

Prev Chronic Dis 2010; 7: A37

Background: This study was conducted in order to qualitatively describe the experiences of persons at high risk of type 2 diabetes using pedometers to promote exercise.

Methods: Seventy-four people at high risk of type 2 diabetes were selected to participate in 6 months of group counselling and pedometer use. The group sessions were used to collect data through questionnaires, theme interviews and video recordings. Data collection took place from April 2007 to April 2008.

Results: Most participants agreed that pedometers were useful for observing levels of physical activity, setting personal goals and evaluating whether or not those goals were met. Problems associated with pedometers included equipment failure, unsuitability for exercise other than walking, and the standard goal of 10,000 steps per day, which many found to be too high.

Conclusion: The pedometer can be used as a motivational tool to increase the exercise level of sedentary adults, provided that they are informed of the experiences of other regular users and that they set realistic goals for walking.

  • Comments: In last year’s yearbook, we highlighted a study illustrating that pedometers do not necessarily increase walking behaviour or metabolic outcomes in people with type 2 diabetes (11). In contrast, this study effectively shows the positive motivational effects that these inexpensive simple devices may have in persons at high risk for the disease. Again, pedometers are limited to just walking exercise, however, which might not be motivating enough for some people and regular counselling and goal setting are critical for effective behaviour change (see paper by Shenoy et al., below).

Effectiveness of an aerobic walking programme using heart rate monitor and pedometer on the parameters of diabetes control in Asian Indians with type 2 diabetes

S. Shenoy, R. Guglani, J. S. Sandhu

Department of Sports Medicine and Physiotherapy, Guru Nanak Dev University, Amritsar, India

Prim Care Diabetes 2010; 4: 41–5

Background: India is often termed the ‘diabetes capital of the world’ because of the overwhelming number of people with diabetes on the subcontinent. Prevention and management of diabetes is now a high priority in this geographic region. Walking is a low impact, easy to perform, and therefore acceptable physical activity for people with type 2 diabetes. This study analyses the effects of 8 weeks of aerobic walking, using a heart rate monitor (HRM) and pedometer for monitoring exercise intensity, on glycaemic outcomes, fasting blood glucose, CV fitness and well-being in patients with type 2 diabetes.

Methods: Forty adults with type 2 diabetes, not treated with insulin, were randomly divided into either an 8-week supervised walking programme using an HRM and pedometer (group A, = 20) or the control group (group B, = 20). All outcomes (glycaemic and metabolic outcomes, anthropometric and body composition, CV indices and general well-being) were determined at baseline and after the 8-week training period.

Results: Overall, the exercise training programme generated an improvement in group A with a significant decrease in fasting blood glucose level by 37% (p < 0.05). Body mass index decreased by 3.9% in the exercising group (A) whereas it increased by 2.2% in the control group (B). General well-being of group A improved by 28.8% (p < 0.05). All parameters associated with CV health of the exercising group (A) were also improved.

Conclusions: Monitoring an exercise programme in patients with type 2 diabetes using an HRM and pedometer was effective, and decreased the HbA1c level, fasting blood glucose level and body mass index and improved general well-being. Additionally, using an HRM helped investigators to attribute all improvements to the exercise intensity used in the study.

  • Comment: This relatively small-scale trial shows nicely the motivational effects of tracking the intensity of physical activity in people with type 2 diabetes. The study also shows that exercise improves quality of life, since the intervention group scored far higher on their test of general well-being. However, in order to truly measure the effectiveness of using a pedometer and HRM on monitoring physical activity, a second study should be done using a control group that still exercises, but without anything to keep track of their intensity. As it is, the study shows that exercising with a pedometer and HRM is better for treating hyperglycaemia than remaining sedentary – which is a good thing.

Exercise and glucose metabolism in persons with diabetes mellitus: perspectives on the role for continuous glucose monitoring – a review article

M. Riddell,1B. A. Perkins2

1School of Kinesiology and Health Science, Muscle Research Centre, Faculty of Health, York University, Toronto, ON, Canada, and2Division of Endocrinology and Metabolism, University of Toronto and University Health Network, Toronto, ON, Canada

J Diabetes Sci Technol 2009; 3: 914–23

Background: To help the reader use a continuous glucose monitor (CGM) to its fullest potential for active patients with type 1 or type 2 diabetes mellitus. The paper explains the technical features, performance characteristics and clinical utility of current CGMs whilst also warning against limitations that may be exacerbated during physical activity.

Methods: The paper is a teaching manual for CGM use as it relates to exercise. It explains the challenges involved with monitoring blood glucose during aerobic and anaerobic activity, how CGMs work, how they can help improve glucose control during and after exercise, and the technological strengths and limitations of the devices.

Results: Advancements include up to the minute glycaemic information when sedentary and during moderately intense exercise, helping reduce the risk of late-onset post exercise hypoglycaemia. ‘Alerts’ can warn patients not just when they are already hypoglycaemic but when their glucose level is dropping quickly and hypoglycaemia may be imminent. The CGM reduces the need to constantly record self-monitored blood glucose levels, an advance that is much more convenient during exercise and for some forms of activity may be the only practical way of monitoring glucose levels.

Conclusions: CGMs are a great way for physically active individuals with diabetes to maintain their active lifestyles with greater security and reassurance due to real-time information regarding their trends towards hyperglycaemia or hypoglycaemia. They allow people with type 1 diabetes to be more in control of their blood glucose levels and along with lifestyle interventions they can motivate people with type 2 diabetes to exercise and improve their glycaemic control. These devices allow a greater degree of freedom and ease of use during exercise and in general, but they are still not accurate enough to replace capillary blood glucose testing.

  • Comment: CGM technology may be considered one of the best recent inventions for people with type 1 diabetes who struggle to maintain good glycaemic control. Exercise itself is just one more variable that can make good control challenging as it can be associated with more glycaemic variation. This paper attempts to provide some of the useful CGM tools available that allow the user to be much more proactive in maintaining good glycaemic levels before, during and after exercise. Future study on the usefulness of these devices to improve overall glycaemic management in active patients is urgently needed, however.

Resistance exercise training lowers HbA1c more than aerobic training in adults with type 2 diabetes

S. Bweir,1M. Al-Jarrah,2A. M. Almalty,1M. Maayah,2I. V. Smirnova,3L. Novikova,3L. Stehno-Bittel3

1Department of Physiotherapy, Allied Medical Sciences, Hashemite University, Zarqa, Jordan,2Department of Physiotherapy, Applied Medical Sciences, Jordan University of Science and Technology, Irbid, Jordan, and3Department of Physical Therapy and Rehabilitation Science, University of Kansas Medical Center, Kansas City, KS, USA

Diabetol Metab Syndr 2009; 10: 27

Background: Both aerobic and resistance training are recommended to control blood glucose in people with diabetes. However, only a few studies have attempted to compare the benefits from the two forms of exercise. This study directly compared the effects of 10 weeks of resistance or treadmill exercises on blood sugar levels before and after exercise and also HbA1c.

Methods: The study included 20 inactive subjects (mean age 53.5 years) with type 2 diabetes. After obtaining baseline measurements of each subject’s HbA1c, blood glucose levels, heart rate and blood pressure, subjects were matched to age, waist circumference and sex and assigned to either resistance or treadmill exercise groups. Both intervention groups met three times per week for 10 weeks under supervision of an exercise therapist. In both groups, exercise intensity progressively increased over the course of the study. Special care was taken to ensure that total energy expenditure, perceived exertion and heart rate were equivalent between treatments. Blood glucose was measured immediately prior to and immediately after each exercise session. HbA1c was measured at the beginning and end of the trial.

Results: Pre- and post-exercise blood glucose levels as well as HbA1c values were improved in both groups. However, the resistance training group clearly had greater benefit in achieving glycaemic control. After the 10-week resistance programme, 80% of the subjects had post-exercise blood glucose levels within the normal range, while only 20% of the aerobic group reached this goal. Additionally the resistance training group had significantly greater reductions in HbA1c compared with the aerobic group (p < 0.006). Although aerobic training did result in statistically significant reduction in HbA1c (p < 0.05), none of the subjects in the aerobic group reached the target HbA1c < 7.0% while 40% of the resistance exercise group achieved this goal. Indeed, resistance training reduced the value of HbA1c by an average of 18% compared with an 8% reduction in the aerobic group. Blood pressure and resting heart rate did not change over the 10-week course in either group.

Conclusion: Ten weeks of resistance training resulted in significantly better improvements in glycaemic control compared to isocaloric and equally difficult aerobic exercise.

• Comments: Few studies have examined the potential benefits of resistance training in patients with type 2 diabetes and even fewer have directly compared it with aerobic exercise. Evidence is mounting, however, that resistance exercise may be as good as or even superior to endurance exercise for people with type 2 diabetes for a number of reasons (12,13). In line with this, the authors in this study point out that the HbA1c value was reduced by 18% in the resistance training vs. 8% in the aerobic exercise group. Major strengths of this study in establishing clinical importance are that the two groups were equal in not only calorie expenditure but also perceived exertion and were very similar in total workout time. Thus, from a patient’s point of view, the results of this study indicate that greater results may be obtained through resistance training with equal effort compared to aerobic training. This study was performed on inactive patients and therefore does not address the potential benefits to adding an exercise programme to more active individuals with diabetes (e.g. who perhaps have more physically demanding jobs). As devotion to an exercise plan should be a long-term commitment to aid in glycaemic control, larger and longer studies should be performed comparing the benefits after many months or years of aerobic and resistance exercise.