The Physician and Sportsmedicine
Menubar Home Journal Personal Health Resource Center CME Advertiser Services About Us

Pumping Insulin During Exercise

What Healthcare Providers and Diabetic Patients Need to Know

Sheri R. Colberg, PhD
John Walsh, PA, CDE


In Brief: Exercise can decrease insulin resistance in most people. Insulin pumps deliver precise insulin adjustments that improve fuel availability and provide glycemic control to help diabetic patients overcome obstacles to exercise. Diabetic patients, their physicians, and their healthcare providers should be familiar with the features and nuances of specific pump models and follow some basic guiding principles for exercise to create the most normal metabolic response possible.

People who have diabetes mellitus, whether they are athletes or not, benefit from exercise that improves the body's ability to use insulin. However, a common concern is that exercise can complicate glycemic control, particularly in patients who use insulin. In nondiabetic individuals, exercise decreases insulin release, which is a normal metabolic response1 when more blood glucose is transported into cells via contraction-initiated glucose transport.2 However, a small amount of insulin is always required during exercise to counterbalance glucose-raising hormones such as catecholamines, glucagon, growth hormone, and cortisol.3

When diabetic patients who use insulin participate in physical activity, they must make frequent insulin adjustments to maintain glycemic control, especially during higher-intensity or longer-duration exercise. The insulin regimen that makes an exercising diabetic individual's response closest to normal is continuous subcutaneous insulin infusion (CSII) therapy, also called insulin pump therapy. Physicians and other healthcare professionals should be well versed in the physiologic effects of pump use and the features and nuances of specific insulin pump models to help their patients effectively manage glycemic control during exercise.

Essentials of Insulin Pump Use

For CSII therapy, a pager-sized pump is worn outside the body in a pouch or on a belt clip. About 10% of individuals with type 1 diabetes wear portable insulin pumps, and this percentage has been increasing each year. Some insulin-requiring type 2 diabetic patients are pumping insulin as well. Scientists attempting to mimic the insulin delivery of a normal, healthy pancreas first developed insulin pumps in the late 1970s; an early prototype was the size of a large backpack.

Pager-sized pumps are designed to provide both preprogrammed basal insulin doses (background insulin delivered every few minutes in small increments to meet insulin requirements while fasting and during sleep) and boluses (larger doses delivered before meals, snacks, or anytime blood sugar is elevated).

Several brands of open-loop systems are made to deliver insulin to a subcutaneous depot (located in the abdomen, buttocks, legs, or upper arms), through a subcutaneous needle or through a Teflon infusion catheter with a metal needle that is removed after insertion. (See "Choosing the Best Insulin Pump for Exercise.") The infusion set is replaced every 2 to 3 days with a new set at an alternate site.

Insulin pumps contain a reservoir or cartridge filled with fast-acting insulin or insulin analogues. The analogues are altered insulin molecules that are absorbed more rapidly than regular insulin. Many individuals use fast-acting insulin analogues in their pumps.4 Fast-acting insulin analogues, compared with regular rapid-acting insulin, have a more rapid onset from subcutaneous infusion sites (5 to 15 minutes vs 20 to 30 minutes), an earlier activity peak (90 minutes vs 150 minutes), and a shorter duration (2 to 4 hours vs 3 to 6 hours). Fast-acting insulin analogues allow a more normal glycemic response following carbohydrate intake and correct hyperglycemia more rapidly.5,6

A study by Tsui et al7 demonstrated that the use of fast-acting insulin analogues in pumps did not lessen hormonal responses to hypoglycemia, but it did improve overall blood glucose control and decreased the frequency of hypoglycemia, perhaps through a heightened hepatic sensitivity to glucagon mediated by fast-acting insulin analogues.8 A concern with fast-acting insulin analogue use, however, is the rapidity with which a hypoglycemic episode can occur if insulin is given and carbohydrates are not absorbed quickly enough. The glycemic index of carbohydrates should also be considered when using analogues.

Benefits of Pump Use for Exercisers

Insulin pump use provides many general benefits (table 1). (For a review of the major benefits of exercise for individuals with diabetes, see "Type 1 Diabetes and Sports Participation: Strategies for Training and Competing Safely.") One of the most important benefits for active diabetic patients is that an insulin pump can produce a metabolic response to exercise that, with experience and blood sugar testing, can be similar to the response of a nondiabetic individual. Another advantage is the precision and speed of insulin adjustments. Other advantages include more reliable insulin action through a constant infusion of short-acting insulin, and precise dosing and timing of insulin doses with meals.

TABLE 1. General Benefits of Insulin Pump Use
Better overall blood glucose control (improved hemoglobin A1c values)

More flexibility in eating (or skipping) meals and snacks

Easier body weight management due to optional snacks and flexible meal sizes

Fewer episodes of severe hypoglycemia, especially nocturnally

Teens can sleep late without risking hypoglycemia

Improvements in hypoglycemic awareness

Greater flexibility in doing spontaneous or planned exercise

Precise dosing of insulin (in 0.05- to 0.5-unit increments)

Less chance of lipodystrophy resulting from repeated injections

More flexible and instantaneous basal dose adjustments

Convenience of taking insulin without the social stigma of public injections

Effects of Exercise

Aerobic exercise in nondiabetic individuals usually decreases insulin release by the pancreas1 and increases glucagon release.9 In nondiabetic subjects who have fasted overnight, exercise can stimulate whole-body glucose uptake even when insulin levels are relatively deficient.2 Some circulating insulin is required to maintain glycemic control during exercise, and pumps can help diabetic patients control circulating insulin levels.

Pump vs syringe. Viberti et al10 studied the hormonal and glycemic responses to moderately severe exercise performed 2 hours after breakfast by eight patients who had type 1 diabetes. Conventional insulin therapy (CIT), involving two to three daily injections of short- and intermediate-acting insulin, produced large decreases in glucose levels during activity (from 220 mg/dL to 80 mg/dL or 12.1 mM to 4.4 mM) with a large increase in free insulin levels. When patients repeated this exercise using CSII therapy, glucose levels fell (from 129.1 mg/dL to 65.4 mg/dL or 7.1 mM to 3.6 mM) but stabilized, and users experienced no increase in free insulin levels. Furthermore, the exaggerated rise in growth hormone during CIT was normalized by 3 weeks of CSII therapy. The authors concluded that good metabolic control attained with insulin pump therapy was accompanied by nearly normal hormonal and metabolic responses to exercise.

Time of day. Circulating insulin levels can be affected by the time of day that pump users exercise. Suspending regular insulin delivery via CSII therapy, 30 minutes before and during exercise, lowered free insulin level during the activity but still produced hypoglycemia during fasting exercise, when a relative state of hyperinsulinemia existed.11 The exercise consisted of 45 minutes of moderately intense cycling at 60% VO2max. When subjects repeated this exercise in the afternoon at least 4 hours after the last premeal bolus of insulin, they experienced a more pronounced decline in blood glucose levels compared with the morning's fasting bout. Even with today's fast-acting insulin analogues, these results suggest that to reduce circulating insulin levels at the start of prolonged exercise, basal insulin delivery may need to be reduced for a longer period (perhaps 60 to 90 minutes) before some moderate or strenuous activities that last longer than 30 to 45 minutes.

Schiffrin et al12 demonstrated that moderate cycling performed in the fasting state can produce hypoglycemia during CSII therapy. The subjects' normal fasting, basal insulin infusion rates were maintained during a bout of cycling (45 minutes at 50% VO2max), but subjects experienced a sharp decrease in glycemia after exercise.

Trovati et al13 showed that, when using insulin pumps, type 1 diabetic patients who are tightly controlled can perform 30 minutes of mild and moderate exercise 2 or 3 hours after breakfast without a high risk of hypoglycemia, despite mild hyperinsulinemia. To achieve optimal control, it is clear that appropriate insulin reductions should be made for planned exercise that closely follows an insulin bolus. This precaution will minimize circulating free insulin levels and the risk of hypoglycemia during and after the activity.14

Intense activity. Exercise intensity and elevated starting blood glucose levels are other factors that appear to affect the metabolic response, often resulting in hyperglycemia instead of the more usual hypogly-cemia. Mitchell et al15 investigated eight type 1 diabetic subjects on CSII therapy to determine their response to a short period of exercise performed in the fasting state at 80% VO2max. When these individuals began exercise with normal blood glucose levels (86.9 mg/dL ± 4 mg/dL or 4.78 mM ± 0.22 mM), they experienced mild hyperglycemia (128.4 mg/dL ± 7.1 mg/dL or 7.06 mM ± 0.39 mM) for 2 hours postexercise. Furthermore, elevated mean preexercise blood glucose levels (150.5 mg/dL ± 9.1 or 8.28 mM ± 0.50 mM) rose progressively during the same 2 hours. (231.3 mg/dL ± 28.4 mg/dL or 12.72 mM ± 1.56 mM). The authors concluded that metabolic control can deteriorate with intense exercise. These glycemic effects, demonstrated in both nondiabetic and euglycemic diabetic subjects, were attributed to exaggerated epinephrine and norepinephrine responses.16 Apparently, a relative state of hypoinsulinemia during fasting exercise can also alter the normal metabolic response to exercise.

Microalbuminuria. Exercise can also exaggerate albumin excretion (a critical marker for diabetic renal disease). In one small study, Viberti et al17 analyzed the effects of 3 weeks of metabolic near-normalization with CSII therapy versus CIT in eight men who had type 1 diabetes. In response to 20 minutes of cycling, subjects using CSII significantly reduced their exercise-induced albumin excretion rate and moderately reduced their blood glucose levels compared with subjects using CIT, despite a similar increase in systolic blood pressure during exercise. However, the relationship between exercise-induced microalbuminuria and the development of diabetic neuropathy is not well understood.

Concerns for Diabetic Exercisers

A few drawbacks of insulin pump use with exercise should be noted.

Displacement. Diabetic ketoacidosis (DKA), a life-threatening condition that may require emergency treatment in a hospital, can begin as quickly as 5 hours after the displacement of an infusion set, especially when it contains a rapid-acting insulin analogue. When a patient becomes hypoinsulinemic from any interruption of insulin delivery, exercise can exacerbate the situation and speed the development of DKA.5 Because of the severity of DKA, insulin pump users must be especially vigilant about maintaining the integrity of their infusion sites during exercise.

Sweating. Excessive sweating may dislodge the subcutaneous infusion set, producing elevated blood sugar levels or DKA if the displacement remains unnoticed for some time. To prevent sweat-related displacement, patients can use liquid skin preparations (eg, Skin-Tac, Torbot Group, Cranston, Rhode Island) and stronger adhesives to anchor the set more firmly. Applying antiperspirant to the skin at the infusion site can minimize sweating beneath the set. Users should check the integrity of the infusion site following vigorous exercise, sweating, or water contact.

Irritation. For those using metal needles, movement or contact may irritate the infusion site. This problem, however, is greatly reduced with pumps that use flexible Teflon infusion sets. Patients, especially those who exercise vigorously, must adhere to the recommendation to replace their insulin infusion sets every 2 to 3 days to prevent skin irritation.

Ambient temperature. Insulin is temperature-sensitive. Exercising in hot or cold environments may cause insulin to degrade and lose its effectiveness. If an insulin pump is placed close to the skin during exercise in the heat, the insulin may become overheated as well. Patients should be cautioned to replace the entire insulin-filled cartridge and infusion catheter at the first sign of any unanticipated hyperglycemia.

Guiding Principles

Knowledgeable healthcare providers can instruct active diabetic patients to follow some basic principles for exercise to achieve the most normal metabolic response possible:

Understand metabolism and activity. Some activities, such as resistance or high-intensity interval training, stress mainly anaerobic energy sources and can have a markedly different effect on glycemia compared with prolonged or mild aerobic exercise.

Predict glycemic response. The glycemic effect of the exercise can often be predicted based on the type of activity, time of day, and circulating insulin levels. Paramount are the length and intensity of the exercise and the training level of the participant. Regimen changes may be needed to prevent hypoglycemia during the activity and for up to 24 hours afterward.

Establish glycemic patterns. Each time they do a new activity, exercisers should check blood glucose levels before, during, and after the activity to establish usual response patterns and watch for future variations.

Make appropriate changes. Regimen changes (increased carbohydrate consumption or decreased insulin) should be made according to previous glycemic responses to the activity. Additional alterations can be made depending on whether physical fitness or weight loss is the exercise goal.

Expect a training response. Finally, active patients should expect a glycemic training response to occur after they have consistently done an activity for 2 to 3 weeks. Training increases fat utilization and potentially spares blood glucose. This response may reduce the need for regimen changes during prolonged aerobic activities. In addition, as muscle mass increases in response to training, an individual's overall insulin sensitivity may increase, which may allow for lower basal and bolus insulin doses.18

General Recommendations for Regimen Changes

The American College of Sports Medicine and the American Diabetes Association have established general clinical practice recommendations for exercise and diabetes that apply to patients who use insulin pumps.19 The problem with these recommendations lies in their generality. The wide variety of sports and recreational activities makes it difficult to make all-encompassing recommendations, but a rule of thumb is for pump users to first consider preexercise blood glucose levels.

The main change from previous recommendations20 is that patients who have preexercise blood glucose levels less than 100 mg/dL (5.5 mM) may not require a carbohydrate snack; pump users can simply reduce or suspend basal insulin during an activity.4 The insulin reductions and the carbohydrate intake necessary for aerobic activity will depend on its intensity and duration. A change in either insulin (basal or bolus doses) or carbohydrate intake can often compensate for shorter, less intense activities.

Exercise is advised if blood glucose levels are less than 250 mg/dL (13.75 mM) and no ketones are detected. If blood glucose is more than 250 mg/dL, exercise is still advised if ketones are not present; however, a small insulin bolus may be needed. If ketones are found and the preexercise blood glucose exceeds 250 mg/dL, exercise is not advised. Exercise may be possible if blood glucose is more than 300 mg/dL (16.5 mM) and ketones are not present, but extra caution is advised, and an insulin bolus may be necessary.

Short, intense activities, such as weight training, may not require any immediate regimen changes, but delayed-onset hypoglycemia should be anticipated. Longer, more intense exercise generally requires a combination of carbohydrate intake and insulin reduction to maintain normal glycemia. For sport- and activity-specific recommendations, more information about initiating pump use, and guidance about insulin dose adjustments, please consult recent texts.4,21

Pumping Up Exercise Benefits

Active people who have diabetes and use insulin pumps can benefit from more normal physiologic responses to exercise and blood glucose levels that remain more stable during and following exercise. Choosing an insulin pump is mainly a matter of patient preference.

Normal exercise concerns for diabetic individuals apply to pump users, but some of these concerns are abated by pump use. To counsel patients on safe and effective insulin pump use during exercise, physicians should be aware of current insulin pump features, basic guiding principles for exercise, and general recommendations for insulin regimen changes.


  1. Martin MJ, Horwitz DL, Nattrass M, et al: Effects of mild hyperinsulinemia on the metabolic response to exercise. Metabolism 120211;30(7):688-694
  2. Lavoie C, Ducros F, Bourque J, et al: Glucose metabolism during exercise in man: the role of insulin in the regulation of glucose utilization. Can J Physiol Pharmacol 1997;75(1):36-43
  3. Hirsch IB, Marker JC, Smith LJ, et al: Insulin and glucagon in the prevention of hypoglycemia during exercise in humans. Am J Physiol 1991;260(5 pt 1):E695-E704
  4. Colberg S: The Diabetic Athlete. Champaign, IL, Human Kinetics, 2021
  5. Draznin MB: Type 1 diabetes and sports participation. Phys Sportsmed 2021;28(12):49-56
  6. Attia N, Jones TW, Holcombe J, et al: Comparison of human regular and lispro insulins after interruption of continuous subcutaneous insulin infusion and in the treatment of acutely decompensated IDDM. Diabetes Care 192021;21(5):817-821
  7. Tsui EY, Chiasson JL, Tildesley H, et al: Counterregulatory hormone responses after long-term continuous subcutaneous insulin infusion with lispro insulin. Diabetes Care 192021;21(1):93-96
  8. Launay B, Zinman B, Tildesley HD, et al: Effect of continuous subcutaneous insulin infusion with lispro on hepatic responsiveness to glucagon in type 1 diabetes. Diabetes Care 192021;21(10):1627-1631
  9. Shilo S, Sotsky M, Shamoon H: Islet hormonal regulation of glucose turnover during exercise in type 1 diabetes. J Clin Endocrinol Metab 1990;70(1):162-172
  10. Viberti GC, Home PD, Bilous RW, et al: Metabolic effects of physical exercise in insulin-dependent diabetics controlled by continuous subcutaneous insulin infusion or conventional injection therapy. Acta Endocrinol (Copenh) 120214;105(4):515-520
  11. Edelmann E, Staudner V, Bachmann W, et al: Exercise-induced hypoglycaemia and subcutaneous insulin infusion. Diabet Med 120216;3(6):526-531
  12. Schiffrin A, Parikh S, Marliss EB, et al: Metabolic response to fasting exercise in adolescent insulin-dependent diabetic subjects treated with continuous subcutaneous insulin infusion and intensive conventional therapy. Diabetes Care 120214;7(3):255-260
  13. Trovati M, Carta Q, Cavalot F, et al: Continuous subcutaneous insulin infusion and postprandial exercise in tightly controlled type 1 (insulin-dependent) diabetic patients. Diabetes Care 120214;7(4):327-330
  14. Sonnenberg GE, Kemmer FW, Berger M: Exercise in type 1 (insulin-dependent) diabetic patients treated with continuous subcutaneous insulin infusion: prevention of exercise induced hypoglycaemia. Diabetologia 1990;33(11):696-703
  15. Mitchell TH, Abraham G, Schiffrin A, et al: Hyperglycemia after intense exercise in IDDM subjects during continuous subcutaneous insulin infusion. Diabetes Care 120218;11(4):311-317
  16. Sigal RJ, Fisher SJ, Halter JB, et al: Glucoregulation during and after intense exercise: effects of beta-adrenergic blockade in subjects with type 1 diabetes mellitus. J Clin Endocrinol Metab 1999;84(11):3961-3971
  17. Viberti G, Pickup JC, Bilous RW, et al: Correction of exercise-induced microalbuminuria in insulin-dependent diabetics after 3 weeks of subcutaneous insulin infusion. Diabetes 120211;30(10):818-823
  18. Devlin JT: Effects of exercise on insulin sensitivity in humans. Diabetes Care 1992;15(11):1690-1693
  19. American Diabetes Association: Position statement: diabetes mellitus and exercise. Diabetes Care 2021;25(suppl 1):S64-S68
  20. Colberg SR, Swain DP: Exercise and diabetes control: a winning combination. Phys Sportsmed 2021;28(4):63-81
  21. Walsh J, Roberts R: Pumping Insulin: Everything You Need to Know to Use an Insulin Pump Successfully, ed 3. San Diego, Torrey Pines Press, 2021

Choosing the Best Insulin Pump for Exercise

Insulin pumps differ in delivery options and features, so physicians need to know the nuances of current models to help their patients select the best model for safe and effective use during exercise.

Current insulin pumps provide a variety of features, such as different basal increments and duration, basal profiles, frequency of basal insulin delivery, temporary basal rate settings, bolus increments, bolus delivery, size, cost, and use in water during swimming or bathing. For exercise, the most important model features are basal rate and bolus increments, basal profiles, and temporary basal settings.

Certain pump models, such as the ones listed below, are clearly superior in their ability to adjust for extended exercise, either planned or spontaneous, through refinements in basal rate reduction.

Animas R-1000 (Animas Corp, Frazer, Pennsylvania) has four basal profiles that are designed to accommodate varying workouts on subsequent days. It has smaller basal increments (0.05 U/hr) that allow insulin-sensitive individuals, such as those on total daily insulin doses of 30 units or less, the flexibility of making minute changes. It also has a temporary basal setting (±10% for 0.5 to 12 hour) that allows slightly more fine-tuning than, but is similar to, the D-TRON discussed below. It is waterproof for surface water activities.
D-TRON (Disetronic Medical Systems, St Paul) has a second basal rate profile that can be set for exercise days. The D-TRON can also deliver meal boluses over an extended period from 15 minutes to 4 hours. This feature could be used to delay or reduce meal boluses for exercise after meals. This pump can also combine extended and normal boluses. The D-TRON offers flexibility in setting a temporary basal rate in 10% increments or decrements over 1 to 24 hours. This model is waterproof during surface water activities.
Model 508 (Medtronic MiniMed, Northridge, California) has three separate basal profiles that can be programmed for regular activity, and small basal (0.1 U/hr) and bolus increments (0.1 unit) similar to the D-TRON. The 508 also offers extended boluses or a combination of extended and normal boluses. One mild drawback of the 508 is that it is only water resistant, not waterproof, and cannot be worn during extended water surface activities. For shorter water workouts, most infusion sets can be disconnected near the infusion site. A bolus (0.5 to 1 unit) may be needed when the unit is reconnected.
Medtronic also makes the Paradigm pump, a newer, smaller model that features a simplified menu system. The Paradigm is watertight up to 8 feet for 30 minutes, so users can shower or swim.

For a detailed comparison chart of common models, go to More specific information is available online at

Photos: Courtesy of Animas Corp; Disetronic Medical Systems; Medtronic MiniMed

Dr Colberg is an assistant professor of exercise science at Old Dominion University in Norfolk, Virginia, and Mr Walsh is a certified diabetes educator at North County Endocrine Medical Group in Escondido, California. Address correspondence to Sheri R. Colberg, PhD, Old Dominion University, ESPER Dept, Norfolk, VA 23529; e-mail to [email protected].

Disclosure information: Dr Colberg and Mr Walsh disclose no significant relationship with any manufacturer of any commercial product mentioned in this article. No drug is mentioned in this article for an unlabeled use.