Prevention of Exercise-Associated Dysglycemia: A Case Study–Based Approach

A 26-year-old woman (weight 55 kg) who has had type 1 diabetes for 12 years expresses concern to her health care team about repeated episodes of hypoglycemia during her aerobic workout (cycling and training on an elliptical machine). She is using a multiple daily injection (MDI) insulin regimen, taking insulin glargine at bedtime and insulin aspart at mealtimes. She takes her aspart with every meal and glargine each night. She begins exercising 4 hours after her mealtime injection and still cannot control for hypoglycemia. Her mid-afternoon exercise routine consists of 60 minutes of stationary cycling, followed by 20 minutes of elliptical work at a moderate to high intensity, three to four times per week. The health care team must determine whether hypoglycemia is occurring because of too much on-board (active) insulin remaining from her previous mealtime injection, excessive basal insulin for exercise, or too few extra carbs.

Problem: Physiological insulin does not exist for aerobic exercise, and the patient is exercising with relative hyperinsulinemia.

One simple strategy to reduce the likelihood of hypoglycemia for this patient is to recommend that she consume fast-acting carbohydrates just before and throughout the activity. These extra carbs should be consumed without administering insulin.

Additional carbohydrates are often recommended for activities that last >30 minutes (27,29). The amount of extra carbs to consume is based on the size of the individual and the intensity of exercise. Evidence suggests that adolescents and young adults oxidize carbohydrate at a rate of ∼1 g/kg/hour of exercise (30,31), whereas carbohydrate absorption from the gastrointestinal tract appears to be limited to ∼60 g/hour during exercise (14). Thus, the extra carbs needed may be as much as 60 g/hour of exercise in this patient while she is exercising at a moderate intensity.

A study by Francescato et al. (32) showed that the amount of carbohydrate required before and during exercise to prevent hypoglycemia in individuals with type 1 diabetes is correlated to plasma insulin, but not fitness, level. Therefore, this patient’s training status may not affect her extra carbs requirement, but the timing of her exercise in relation to her last meal might. In one study (32), the amount of extra carbs needed by type 1 diabetes patients to prevent hypoglycemia decreased as the time elapsed from insulin administration (regular insulin) increased. However, in another study of people using insulin isophane (Humulin N) and lispro, there was a similar risk of hypoglycemia during early and late postmeal exercise (33). Based on a large survey of people with diabetes (the type of diabetes was not identified), exercising 1–2 hours after a meal was associated with greater drops in glucose than exercising within 30 minutes before or >3 hours after a meal (34).

We have also found that fewer extra carbs will be needed if there is a low level of on-board insulin from a previous bolus. These extra carbs are meant to be consumed in small portions rather than as one large meal or snack (11). For example, if 55 g of extra carbs are required for 1 hour of exercise for this patient, ∼22 g should be consumed before exercise and the remaining 33 g should be divided into two or three small snacks during the exercise session.

Many patients who are on low-carbohydrate diets or who are interested in weight loss might find this amount of carbohydrate excessive. In this case, the woman could consume a total of 55 g of rapid-acting carbohydrate during the first hour of cycling and an additional few grams if needed before her 20-minute elliptical session. Although many patients find this amount of extra carbs excessive, one study of adolescents with type 1 diabetes found that it prevented hypoglycemia without promoting hyperglycemia (11).

Because excessive carbohydrate intake may promote gastric distress (35) and will add to the total daily calories consumed (220 extra kcals in this example), this option may not be preferred for routine exercise, as in this case. Regular exercise reduces total daily insulin needs in lean individuals with type 1 diabetes by ∼10–20% (36–38). Moreover, this amount of training would be expected to lower her reliance on carbohydrates as a fuel for exercise (39).

Because of the timing of her exercise, the health care team has noted that this patient has little to no active on-board prandial insulin at the start of her afternoon workout. During exercise, glucose is taken up into skeletal muscle and then oxidized via a noninsulin-mediated process (40). However, some insulin is still required in the blood during exercise to prevent hyperglycemia caused by excessive hepatic glucose production and impaired glucose uptake into working muscle (41,42). Thus, one strategy might be to lower her bedtime insulin glargine by 20% on the evening before exercise or to split her long-acting insulin into two equal doses (morning and night) and reduce the morning dose on the days she exercises. A reduction in bedtime insulin could also then be made if nocturnal hypoglycemia continued to occur after exercise.

However, for MDI patients, this strategy may increase the risk of hyperglycemia throughout the day before the physical activity. In any case, this patient should measure glucose levels during the night after exercise at 3:00 a.m. to determine how much her glucose drops between bedtime to 3:00 a.m. Anyone whose glucose drops >40 mg/dL overnight should be considered to be at high risk for nocturnal hypoglycemia (43).

It is usually recommended that at least two pre-exercise glucose measurements be taken and that monitoring be done intermittently during exercise (e.g., every 30 minutes). Frequent glucose monitoring after exercise should help protect against post-exercise hypoglycemia. Research has shown that the greatest susceptibility to hypoglycemia after exercise occurs at ∼2:00–3:00 a.m. (43,44). However, nocturnal hypoglycemia can occur any time after exercise (45).

If hypoglycemia cannot be managed through any of these recommendations, the health care team should ask this patient to consider continuous subcutaneous insulin infusion (CSII, or insulin pump therapy). Since the late 1970s, many studies have shown that CSII leads to significant improvements in glycemic control, with resulting improvements in A1C (46). Increased flexibility with regard to exercise has also been reported with CSII (19). For this patient, CGM and pump therapy should provide added flexibility with regard to basal insulin adjustments and carbohydrate intake based on directional trend arrows from the CGM. It should be acknowledged, however, that insulin pumps and CGM have not yet been widely adopted, perhaps because of accessibility issues, costs, device complexity, and attitudes about ease of use (47).

A 17-year-old boy who has had type 1 diabetes for 6 years has been experiencing nocturnal hypoglycemia after high-intensity interval training with sprints. He has been using an insulin pump for 3 years and has been participating in contact sports for numerous years. He has a low body fat percentage, weighs 84 kg, and is 6 feet, 2 inches tall. Because of the nature of his exercise, he removes his insulin pump during training, usually for ∼1 hour. When he reconnects his pump, he always administers a half correction bolus of insulin based on his elevated glycemia in early recovery. He has told his health care team that he knows from experience that without the half correction bolus after reconnecting his pump, he experiences severe rebound hyperglycemia after exercise. He has been noticing that, in the early morning hours after exercise, his blood glucose levels are low (50–70 mg/dL).

Problems: Prolonged disconnection of the insulin pump for high-intensity exercise results in relative hypoinsulinemia by the end of exercise, and heightened insulin sensitivity in recovery (for ∼12 hours) predisposes the patient to nocturnal hypoglycemia.

Post-exercise hyperglycemia can result from intense exercise (16), excessive carbohydrate intake (11,19), or large insulin dose reductions (18). Treating post-exercise hyperglycemia must be done cautiously, particularly at bedtime, because severe hypoglycemia may ensue (48). Although no standard guidelines exist for treating post-exercise hypoglycemia, a conservative insulin correction (50% correction dose), along with frequent glucose monitoring, seems prudent.

Episodes of nocturnal hypoglycemia are a concern for active people with type 1 diabetes (49). Strategies for preventing nocturnal hypoglycemia differ for patients who are on MDI versus those on CSII. Kalergis et al. (50) suggest that patients on MDI should consume a bedtime snack that consists of protein and complex carbohydrates. Those on CSII can do this as well, but they also have the capacity to adjust their basal insulin infusion rate both during exercise and in recovery. Small bolus corrections are also more easily calculated and performed with pump therapy. In addition to consuming a bedtime snack, insulin adjustments may be required as discussed in option 2 below. Although low–glycemic index meals and bedtime snacks often help to limit postprandial hyperglycemia in MDI or CSII patients, it appears that this strategy does not entirely protect against post-exercise, late-onset hypoglycemia (21).

Either way, some examples of appropriate snacks in the recovery period once post-exercise hyperglycemia has been resolved include fruit smoothies (dairy-based), yogurt drinks, or fruit mixed with yogurt. Studies have shown that dairy (e.g., chocolate milk) consists of a 4:1 ratio of carbohydrate to protein, is beneficial for muscle glycogen resynthesis, and aids in rehydration (51). For individuals who are lactose intolerant, protein options include nuts and seeds (e.g., almonds, peanuts, or pumpkin seeds), quinoa, and soy milk. Lactose-free yogurt or dairy options are also available. Because protein requirements are greater in athletes than in nonathletes (∼1.2–1.7 g • kg • day^–1 in endurance-trained and strength-trained athletes) (52), this patient would require ∼100–140 g of protein daily. This could include a snack containing 7–10 g of protein and 30–40 g of carbohydrates.

Another recommendation would be to reduce the basal insulin infusion rate at bedtime on the evening after interval training. For young patients using CSII, Taplin et al. (53) demonstrated that reducing the basal rate by ∼20% between 9:00 p.m. and 3:00 a.m. largely prevented nocturnal hypoglycemia caused by afternoon aerobic exercise. On days when this patient is active, the health care team might consider setting a new pre-programmed basal insulin pattern on his pump (i.e., a 20% basal rate reduction starting at bedtime and continuing for 6 hours). Once again, the post-exercise hyperglycemia would need to be resolved before initiating an overnight basal rate reduction.

It is common for individuals with type 1 diabetes to experience rebound hyperglycemia with intense exercise, particularly after an extended period of time (>1–2 hours) without basal insulin infusion. Competition stress may also promote hyperglycemia. The health care team might suggest that the patient maintain his insulin pump usage during interval training to help limit the hyperglycemia that occurs after exercise. If the insulin pump is to be worn, a basal rate reduction of 50–80% starting 1 hour before training would be recommended to help limit the risk of hyperglycemia but still offer some protection against hypoglycemia. This strategy should help to minimize the need for a post-exercise correction bolus that could be contributing to late-onset hypoglycemia.

Aggressive post-exercise insulin corrections near bedtime may have contributed to severe hypoglycemia and death in at least one patient with type 1 diabetes (48). One key recommendation would be to perform more frequent glucose monitoring at bedtime. It is also important to stress that hyperglycemia during and after exercise can be caused by a number of factors, including stress, overconsumption of carbohydrates before exercise, or even miscalculating insulin dose adjustments (26). Going to bed with a slightly elevated blood glucose concentration after a bedtime snack or a 20% basal rate reduction would be expected to minimize hypoglycemia risk. Cautious (conservative) correction of post-exercise hyperglycemia close to bedtime is also warranted.

A 46-year-old, lean, active man with type 1 diabetes incorporates both aerobic and resistance training into his exercise regimen 4–5 days/week. He likes to run and cycle but also feels it is important to do some resistance activity during every workout to maintain strength and lean mass. The aerobic portion of exercise usually lasts from 30–45 minutes followed by 20–30 minutes of weight-lifting.

The patient has been wearing an insulin pump for 12 years and recently added CGM to provide real-time information about changes in his glucose levels during exercise. He tells his health care team that his blood glucose control is better on an insulin pump than it was on MDI therapy. However, he sometimes still struggles with hypoglycemia during exercise despite a basal rate reduction at the onset of exercise. He also mentions that his glucose sensor appears to be “off” or at least delayed compared to his actual blood glucose levels based on the frequent monitoring he performs with a glucose meter during exercise. Hypoglycemia occurs within 30–40 minutes after the start of his aerobic workout, and this often delays his resistance workout or makes him too weak for weight training.

Problem: Performing steady-state aerobic exercise before resistance exercise causes rapid hypoglycemia, leaving minimal energy for strength training.

The main concern for this patient is that he experiences hypoglycemia during the aerobic portion of his workout, and this deteriorates his strength for subsequent weight training. As mentioned above, hypoglycemia is a barrier to physical activity (54), and exercise training does not appear to minimize its risk (55). Real-time glucose sensing with a CGM device can help alleviate fear, increase glucose awareness, and improve glucose control during exercise. However, some limitations of CGM need to be acknowledged.

Although exercise does not appear to affect sensor accuracy (55), the delay in equilibrium between interstitial fluid and capillary blood can be troublesome during rapid drops in glycemia, which tend to occur during aerobic exercise. Indeed, CGM readings have been reported to have lag times of anywhere from 5 to 28 minutes compared to capillary or venous glucose measurements in humans, depending on the experimental conditions (56–59). Exercise-mediated acidosis has also been reported to reduce sensor accuracy (60). If blood glucose is increasing or decreasing at a rapid rate (220 mg/dL/hour^–1), there may be a more pronounced mismatch between sensor values and actual blood glucose values because of an intrinsic sensor lag (61). However, when calibration conditions are optimized, sensor lag time has been reported to be as short as 1.5 or 8.9 minutes when glucose levels are falling or rising, respectively, in people with type 1 diabetes who are at rest (62). During aerobic (63,64) and resistance (64) exercise, CGM has been shown to track glucose changes accurately with a reduced lag time compared to during rest, perhaps because of increased blood flow–mediated equilibrium.

Although CGM can help to alert patients to drops in glycemia and thus help to prevent hypoglycemia, patients need to develop strategies for carbohydrate intake if downward trend arrows are observed. In one small pilot study of adolescents with type 1 diabetes (15), an intake algorithm for rapid-acting carbohydrate helped to eliminate hypoglycemia in a sports camp setting. Table 2 provides recommendations based on sensor glucose readings and trending arrows.

Also, the starting blood glucose concentration before exercise is extremely important in determining when carbohydrate intake should be initiated. The health care team could suggest the strategies described in Table 1 for the aerobic portion of the exercise session. They should also recommend a basal insulin infusion rate reduction to start well before the start of exercise (Table 4).

The order of this patient’s exercise routine needs consideration. Yardley et al. (65) recently published evidence that performing resistance exercise before aerobic exercise reduces the likelihood of developing hypoglycemia in individuals with type 1 diabetes. In this case, the health care team could recommend doing the resistance training before the aerobic exercise, with the caveat to alternate daily the muscle groups used during strength training to help promote muscle recovery. Resistance training before aerobic exercise also has been shown to decrease reliance of fast-acting carbohydrates during exercise (66).

Because resistance exercise may not be done every day for recovery reasons, sprinting either at the start (67) or at the end (68) of aerobic workouts may help boost glucose levels in recovery. It should be noted that for individuals who often experience pre-exercise hyperglycemia, performing an aerobic activity first might be preferable to help lower glucose levels to target before performing any anaerobic or resistance-based activities.

Numerous possible strategies exist for managing blood glucose levels during and after exercise for individuals with type 1 diabetes. These include increasing carbohydrate intake before, during, and after exercise; lowering pre-exercise insulin levels by reducing prandial or basal insulin or both; and changing the type or order of the exercise performed (anaerobic vs. aerobic). Because no strategy can guarantee stable blood glucose levels during and after exercise, using CGM may improve control; CGM can help to facilitate minute-by-minute changes in insulin delivery or nutrient intake.

Patients and caregivers should be made aware of the multiple factors that can affect glucose levels during and after exercise, including the amount of active, or on-board, insulin at the time of exercise; the intensity, duration, and type of activity performed; and the type and amount of carbohydrates consumed before exercise. Although the fear of hypoglycemia during and after exercise may remain, implementing these reasonable management strategies can help to reduce this fear and enhance the overall well-being of active people with type 1 diabetes.

D.P.Z. is a speaker for Medtronic Canada. M.C.R. is a speaker for Medtronic Canada and Eli Lilly and Co., and an advisory board member for Sanofi Global. No other potential conflicts of interest relevant to this article were reported.