Where to Start? Physical Assessment, Readiness, and Exercise Recommendations for People With Type 1 or Type 2 Diabetes

Although strong evidence supports the promotion of physical activity in adults with diabetes, the most appropriate way to screen and evaluate people with diabetes ahead of exercise participation has remained somewhat controversial. Although people with diabetes who are middle-aged or older tend to have higher cardiovascular disease (CVD) risk than their peers without diabetes, it is important to put the risks into perspective, as the barriers and risks of inactivity in this population are also considerable. For most people with diabetes, the risks of morbidity and mortality associated with being sedentary are considerably greater than the risks associated with exercise.

Traditionally, several professional organizations have promoted survey questionnaires to screen and identify at-risk individuals who should be medically evaluated before initiating an exercise training program (13–16). However, given the high prevalence of CVD risk factors and use of antihypertensive medications and statins in people with diabetes, this approach has proven to be too inclusive (17), with virtually all people with diabetes required to obtain medical clearance ahead of participating in an exercise program, according to these algorithms. Accordingly, updated American College of Sports Medicine (ACSM) guidelines (4) and consensus statements (3) have slightly relaxed this guidance by discontinuing the need for medical clearance before participation in moderate activity for people with diabetes who are currently active. However, these recommendations still suggest that all individuals with diabetes who are currently sedentary, which is a large portion of people with diabetes of any type, obtain medical clearance ahead of participation in moderate, or even light, exercise training.

These available screening surveys and guideline algorithms result in a high number of medical referrals and can create another barrier for sedentary adults to become more physically active. Programs that require a “pre-participation form” to be completed by a medical HCP place the burden of completing this administrative task, and any associated costs, on participants. If HCPs in turn require a physical examination and/or exercise stress testing before completing these forms, individuals may have to coordinate time off of work, child care, and transportation to a medical visit, at which they may be responsible for a copayment and related costs. In some populations, access to a primary care clinician can also be a significant barrier. Because habitually sedentary individuals with diabetes often perceive numerous barriers to the initiation and maintenance of a physical activity program (18), being judicious regarding the need for medical clearance may help to reduce this barrier.

Controversy also remains regarding the use of pre-participation exercise stress testing for screening purposes. Although exercise stress testing can be valuable in the development of an exercise prescription, measurement of cardiorespiratory fitness, and assessment of long-term prognosis (19), concerns about costs, barriers, follow-up with expensive/invasive tests, and potential low yield suggest that it may be impractical to screen asymptomatic individuals with diabetes who want to begin a moderate-intensity exercise training program. Several trials investigating screening for coronary artery disease (CAD) in asymptomatic people with diabetes found that such screening did not reduce cardiac event rates at follow-up (20–22). Accordingly, guidelines (23,24) do not recommend routine CAD screening in people with diabetes.

However, this evidence does not necessarily apply to individuals who wish to initiate exercise. Although recommended by ACSM (4), the utility of pre-participation exercise stress testing in asymptomatic adults with diabetes is unclear. A study (25) in 1,705 people with type 2 diabetes who were referred for supervised exercise reported that the rate of major cardiovascular outcomes was very low (<1%), with no difference between those who underwent stress testing and those who did not, over a follow-up period of 3.4 years. Adverse events were more likely to occur during the follow-up period in participants who reported symptoms of chest pain, had experienced a previous cardiovascular event, or had known CAD, arrhythmias, congestive heart failure (CHF), and/or microvascular complications. No evidence currently exists on the usefulness of pre-participation screening for resistance exercise training.

We suggest that required medical clearance for exercise participation be reserved for individuals who report signs and symptoms suggestive of CVD (Table 1) and for middle-aged and older individuals with diabetes who are not already exercising regularly and who want to begin a very-high-intensity exercise program. Because people with long-term or complicated diabetes who are not appropriately managed medically may be at an increased risk of silent ischemia, ensuring that participants have been medically evaluated within the past 12 months may be good practice. Furthermore, HCPs may want to consider pre-participation exercise stress testing for individuals who are symptomatic or have known CAD, arrhythmias, CHF, and/or microvascular complications and for those planning very vigorous exercise, such as racing, long-distance running, or high-intensity interval training (HIIT). HCPs may also want to consider a stress test for anyone they suspect may be ignoring symptoms or not giving an accurate history.

Because these suggestions limit the need for medical clearance, it is important that HCPs and exercise program staff have an understanding of this approach. The reduced need for medical clearance does put the impetus on these professionals to be especially vigilant about screening for and educating participants about the warning signs and symptoms suggestive of CVD, particularly at the time an exercise program is started and/or when the exercise intensity is increased.

As discussed above, “exercise” is planned, structured physical activity. These recommendations are focused on these activities rather than on unstructured physical movement, with the exception of reducing sedentary time.

Aerobic exercise, such as walking, bicycling, swimming, or jogging, involves continuous, rhythmic movements of large muscle groups. When performed regularly, aerobic exercise training increases cardiorespiratory fitness. Resistance exercise training, or strength training, involves brief repetitive exercises with free weights, weight machines, resistance bands, or one’s own body weight. When performed regularly, resistance exercise training increases muscular strength and endurance.

Both aerobic and resistance training decrease insulin resistance, and randomized trials have shown that both training types reduce A1C in individuals with type 2 diabetes (26–28), and their combination is likely to be more effective than either type of exercise alone (29). Large, long-term cohort studies have shown that people with diabetes who exercise more have lower cardiovascular and overall mortality rates than those who exercise less (30–34). For these reasons, both regular aerobic and resistance exercise are recommended for people with diabetes. Several professional organizations, including the American Diabetes Association (2), ACSM (3), and Diabetes Canada (5) recommend the accumulation of a minimum of 150 minutes of aerobic exercise per week, spread over at least 3 days and with no more than 2 days without aerobic exercise, and at least two (but preferably three) sessions per week of resistance exercise on nonconsecutive days.

However, many people with diabetes perform very little physical activity and would find it daunting to attempt to meet these goals right away. For very sedentary middle-aged and older people, it may be advisable to begin with very modest goals, such as walking for 5–10 minutes/day, and to gradually increase these amounts every week over a period of several months. Many people prefer to perform several shorter sessions of exercise each day rather than a single longer session (e.g., three 10-minute sessions rather than a single 30-minute session), and the glycemic benefits of doing either may be similar (35).

Strength training is particularly beneficial for people with diabetes. Muscle strength and muscle mass tend to decline with age, especially after the age of 60 years, and these age-related declines tend to be more marked in older adults with diabetes than in those without diabetes (36). Reductions in muscle strength and bulk can result in reduced function, disability, and difficulty with activities of daily living (37).

Pain from osteoarthritis can be a barrier to physical activity (38). Water-based activities, such as walking in a pool, swimming, or aquatic exercise classes, can be recommended for people with osteoarthritis pain (39,40). In adults with type 2 diabetes, a meta-analysis found that aquatic exercise training reduces A1C by about the same amount as nonaquatic aerobic exercise (41). Using exercise equipment such as a recumbent bike or elliptical machine that reduce impact and shear force can also be helpful.

HIIT involves alternating between periods of higher- and lower-intensity exercise in a single session. This type of training results in larger gains in cardiorespiratory fitness than continuous moderate-intensity training (42,43). Some trials have found that HIIT causes greater reductions in A1C than continuous moderate-intensity aerobic training in people with type 2 diabetes (44,45). In type 1 diabetes, some studies have found that HIIT results in less hypoglycemia during exercise compared with continuous aerobic training (46–48).

Although not always feasible or affordable, supervised exercise training is particularly beneficial in people with diabetes. A systematic review and meta-analysis found that, for both aerobic and resistance training in people with type 2 diabetes, supervised exercise programs resulted in reduced A1C regardless of whether there was a dietary cointervention, whereas unsupervised programs reduced A1C only if there was also a dietary intervention (26). A meta-analysis of resistance exercise intervention trials in type 2 diabetes found that programs with less supervision showed less favorable changes in A1C, insulin resistance, and body composition than programs with more supervision (49). A randomized trial compared 12 months of twice-weekly supervised aerobic and resistance exercise plus exercise counseling versus exercise counseling alone in people with type 2 diabetes and metabolic syndrome (50). The group receiving the supervised exercise training had significantly greater reductions in A1C, blood pressure, BMI, and waist circumference and greater increases in cardiorespiratory fitness, strength, and HDL cholesterol compared with the group receiving exercise counseling alone (50). Although expensive, supervised exercise training can be cost-effective in type 2 diabetes (51).

Sedentary behavior involves prolonged sitting or reclining while awake, such as when using a computer, watching TV or videos, or driving. Observational studies have shown associations between larger amounts of sedentary time and higher mortality in people with and without diabetes, even after adjusting for time spent doing physical activity (52–55). Small trials have shown that interrupting sitting time with light walking or light resistance exercise for several minutes at a time over the course of a day can reduce postprandial increases in glucose, insulin, and triglycerides (56–58). Therefore, reducing sedentary time is an important goal, in addition to increasing physical activity. A suggested initial activity plan for a person new to exercise is provided in Table 2 (5).

Three main areas worthy of discussion related to becoming and staying physically active—especially for anyone with type 1 diabetes or anyone who takes mealtime insulin—are hypoglycemia risk and prevention, exercise timing (including prandial status), and differences in glycemic responses based on biological sex. These areas should be addressed when planning where to start with being active and overcoming barriers to participation. Additional strategies to support greater exercise participation in all adults with diabetes are given in Table 3.

Fear of hypoglycemia has long been reported as a barrier to regular physical activity in individuals with type 1 diabetes of various ages (59–62), for good reason. A number of factors increase the risk of hypoglycemia around physical activity, and hypoglycemia risk can vary according to the type, intensity, and duration of the activity (63–65). In many studies, most aerobic exercise appeared to lower blood glucose, whereas intense activity (such as interval training or heavy resistance exercise) raised it, at least temporarily (66–68). However, glycemic responses can also vary based on prandial status (69), time of day, sex, and the total volume of activity (70).

The exogenous insulin on board during physical activities can also affect hypoglycemia risk, with higher insulin levels raising hypoglycemia risk, possibly related in part to the later peaks and longer durations of action of higher doses of rapid-acting insulin types (71). Increases in blood flow related to physical activity can lead to faster insulin absorption from subcutaneous depots, increasing circulating insulin levels and, consequently, raising the risk for exercise-induced hypoglycemia (72), although hypoglycemia frequently can be prevented by lowering insulin doses before activities or insulin delivery via an insulin pump during activities (73–77). Moreover, other factors such as inadequate or restricted carbohydrate intake (78–80), failure to monitor glucose levels, and exercising in the heat may, in some cases, increase the risk of hypoglycemia. Even undertaking new or unfamiliar activities presents a higher risk because of greater blood glucose use until a training effect increases fat utilization as an alternate fuel (81).

To further compound hypoglycemia risk, not all exercise-related events occur during activities. In some cases, the risk has been shown to be biphasic, occurring both immediately after the cessation of activity and again 7–11 hours later and up to 24 hours or more after exercising (82,83). When substantial amounts of muscle glycogen are used during an activity of any type, the risk for later-onset hypoglycemia appears to be greater. In particular, repeated interval workouts or intense resistance training can result in substantial glycogen depletion, thereby increasing hypoglycemia risk (84) and suggesting that post-exercise vigilance and possible insulin reductions and increased carbohydrate intake may be warranted to prevent hypoglycemia (85,86).

Finally, it has long been reported that individuals with type 1 diabetes may have multiple impairments in their counterregulatory hormone response to hypoglycemia, placing them at high risk for severe hypoglycemia, especially during subsequent hypoglycemia or exercise (87). Most individuals with type 1 diabetes eventually develop a diminished glucagon response to hypoglycemia (88), and their epinephrine release may also be blunted during both exercise and hypoglycemia. Exercise itself may also reduce their subsequent counterregulatory responses to hypoglycemia on the same day or the next day (89,90).

For individuals with type 1 diabetes, it may be that exercise timing and prandial status (e.g., exercising after an overnight fast and before eating breakfast) has a bigger impact on keeping blood glucose levels stable or higher during activity than the type of exercise undertaken (69). Performing exercise training first thing in the morning usually results in different glycemic outcomes than doing the same activity in a fed state later in the day. In 35 adults with type 1 diabetes using sensor-augmented insulin pumps, 60 minutes of moderate aerobic exercise on a treadmill before breakfast resulted in less hypoglycemia and more time in the target glycemic range than the same exercise performed in the afternoon on another day (91).

Similarly, in a small study involving 12 active participants with type 1 diabetes who undertook sprint interval training before breakfast or before dinner, the morning activity raised their blood glucose during and after the activity, whereas activity later in the day caused a decline both during the activity and the hour following it (92). Likewise, pre-breakfast resistance exercise increased blood glucose in eight participants with type 1 diabetes (93), but afternoon resistance exercise decreased blood glucose in 12 others (94). Conversely, in another group of 12 habitually active adult participants with type 1 diabetes using continuous glucose monitoring, overall glycemia, time in range, and other glycemic variables were largely unchanged during the 24-hour period after either aerobic or resistance exercise performed in the late afternoon, suggesting that both types of activity undertaken at that time of day had a similar glycemic impact (95). Thus, undertaking fasted exercise in the morning may help to prevent hypoglycemia.

Biological sex has long been investigated as a basis for differences in counterregulatory hormonal responses to exercise (96–98), with reduced responses in males with type 1 diabetes likely leading to a higher risk of hypoglycemia after recent exercise (99) and a blunted hormonal response to exercise after hypoglycemia (100) compared with females with type 1 diabetes. In adults without diabetes, sex-related differences in hormonal responses to exercise may result in altered fuel metabolism during exercise (101,102). In individuals with type 1 diabetes, the advent of attempted closed-loop insulin delivery systems has raised this question again because hormonal fluxes could potentially impact insulin dosing (103). In females with type 1 diabetes, at least one study has suggested that they may experience smaller declines in blood glucose during exercise and a lower risk of late-onset hypoglycemia after acute resistance exercise compared with males (104), perhaps due to higher estrogen levels after resistance exercise, resulting in higher lipid and lower blood glucose use. Also, in females with type 1 diabetes, blood glucose levels are generally higher during the luteal phase of the menstrual cycle, when estradiol and other hormones surge (105). In a cross-sectional study, more males than females with type 1 diabetes reported consuming extra carbohydrates, both before and during physical activity, as a strategy to prevent hypoglycemia (106). More investigation into the potential impact of these differences in exercise responses and hypoglycemia risk related to biological sex is warranted (107).

When recommending exercise for, or even discussing activity with, people with any type of diabetes, it is important to know where to start. In most cases, requiring pre-participation screening, exercise stress testing, or medical clearance creates a barrier to increased exercise that poses more of a risk for individuals than starting activity would. Use such requirements judiciously and mainly for people with signs and symptoms of CVD or those who want to undertake very vigorous exercise. Most individuals will benefit from regular aerobic and resistance exercise training, although starting levels may need to be much lower than recommended daily and weekly target amounts. Reducing sedentary time should be recommended to all. Finally, individuals with type 1 diabetes and other insulin users require consideration of hypoglycemia risk and prevention, exercise timing, and differences in glycemic responses when planning where to start with being active and overcoming barriers to participation.