| Sign In to gain access to subscriptions and/or personal tools. |
Interventions for Being Active Among Individuals With DiabetesA Systematic Review of the LiteratureFrom the Department of Pharmacy Care Systems, Harrison School of Pharmacy, Auburn, Alabama (Dr Kavookjian), and the Clinical Pharmacy Department, West Virginia University School of Pharmacy, Morgantown (Dr Elswick, Dr Whetsel). Correspondence to Jan Kavookjian, PhD, MBA, Department of Pharmacy Care Systems, Harrison School of Pharmacy, 128 Miller Hall, Auburn, AL 36849-5501 (kavooja{at}auburn.edu).
Purpose The purpose of this systematic review is to assess and summarize evidence and gaps in the literature regarding the intervention for being active (exercise) among individuals with diabetes. Methods Twelve electronic databases were searched. Publications eligible for inclusion specifically studied learning, behavioral, clinical, and humanistic outcomes for exercise interventions in adult patients with type 1 and type 2 diabetes. Results Seven reviews (2 systematic reviews, 3 meta-analyses, 2 technical reviews) and 34 individual, nonreview studies (18 randomized controlled trials, 16 nonrandomized trials) met inclusion criteria. For type 2 diabetes, findings suggested that exercise had a positive effect on glycemic control and decreased cardiovascular risk, but the impact of exercise on behavioral and humanistic outcomes was unclear; long-term outcomes and adherence to exercise interventions is unknown because most studies were of short duration. The overall impact of varied types of exercise in type 1 diabetes was unclear, especially regarding glycemic control. Potential benefits of exercise in type 1 may include improved cardiovascular health. Conclusion The review did not identify specific successful intervention details because of the heterogeneity of studies, subjects, and research gaps. General findings suggest that physical activity is better than no exercise at all; intensive regimens, if tolerated by patients, achieved better clinical outcomes than less intensive regimens. Reviewed studies using structured exercise regimens exhibited a more significant impact on outcomes. Substantial gaps in the literature include studies measuring direct effects of exercise in the US minority populations most affected by type 2 diabetes and economic evaluations of exercise interventions. Interventions must be tailored to individual patient needs to succeed.
The continued rise in the prevalence and incidence of diabetes complications1-3 has prompted researchers to examine factors contributing to the increase. Studies of issues surrounding diabetes complications indicate that only about 30% to 40% of patients with chronic illnesses are ready to take action toward self-care.4-6 In fact, Boyle and colleagues7 found that diabetes patients were less ready to engage in physical activity and other self-care behaviors than patients with other chronic conditions such as hypertension and dyslipidemia. Of the self-care behaviors affecting glycemic control and prevention of diabetes complications, the lifestyle changes of diet and physical activity remain the most problematic.4,8,9 This is perhaps due in part to the complexity of achieving and maintaining these particular behaviors and that of managing change on multiple behaviors at the same time.4,8-10 Physical activity, in particular, is neglected by much of the US population, as suggested by the general increase in obesity and sedentary lifestyle.11,12 Despite evidence that substantiates the prevention and/or delay in complications for diabetes patients who engage in even mild activity, most diabetes patients report that they are not regularly active.4,8,13 The American College of Sports Medicine position statement on exercise and type 2 diabetes reports that physical activity is an underused therapy in the medical care of diabetes patients.13
In 2003, the American Association of Diabetes Educators (AADE) published a technical review of diabetes self-management education (DSME) core outcomes measures14 and a position statement on standards for measurement of the DSME outcomes identified in the technical review.15 The technical review identified 7 self-care behaviors key to diabetes self-management. Among the AADE 7 Core Measures of outcomes performance, being active is listed first. Being active is referred to as physical activity (exercise) and is defined as the act of expending energy. It is generally categorized into aerobic (requiring oxygen to maintain muscular effort) or anaerobic (not requiring oxygen to maintain muscular effort).13,14,16 To simplify and standardize this report, the term exercise will be used within this text as a general label for being active and physical activity as well as exercise. General target outcomes of self-management behaviors are distinguished by AADE as lying along a continuum, ranging from immediate outcomes (learning and skills acquisition), to intermediate outcomes (behavior change), to postintermediate outcomes (clinical improvement), to long-term outcomes (health status improvement: quality of life [QoL] and economic outcomes). The AADE position statement on outcomes measurement in practice indicates that the optimal timing and frequency should include an initial (baseline) measurement and at least 1 follow-up, with the optimal protocol including remeasurement at 2 to 4 weeks and again every 3 to 6 months. Space limitations prevent this report from providing a comprehensive overview of these outcomes categories and measurements; readers are referred to the AADE position statement and technical review for additional details.14,15 Exercise interventions aimed at achieving these outcomes vary by type, duration, intensity, and frequency. As with any self-care behaviors, barriers to exercise vary and are individual to particular patients. Among potential considerations for barriers, type of exercise and duration, intensity, and frequency may need to be tailored to severity of disease progression since advanced complications such as neuropathy and retinopathy may limit the exercise capabilities of a diabetes patient.16-18 In addition, patients may consider time, among many other factors, as posing a substantial barrier to integrating regular exercise or physical activity into daily lifestyle. Each of these poses challenges to the initiation and maintenance of exercise for patients and therefore to the measurement, monitoring, and management of exercise intervention outcomes for providers. Exercise is important to both type 1 and type 2 diabetes patients. For type 2 diabetes patients, engaging in regular exercise may improve glycemic control and reduce the risk of microvascular and macrovascular complications, increase insulin sensitivity, reduce stress and stave off depression,19-22 contribute to weight loss/maintenance,23 and contribute to control of lipids and blood pressure, thereby reducing the risk of cardiovascular disease, the leading cause of death in diabetes patients.13,14,16-18,24-30 Patients with type 1 diabetes can also benefit from regularly engaging in exercise or physical activity.31 Outcomes from exercise in type 1 diabetes patients may include potential improvements in glycemic control, although findings on this outcome are mixed30,31; reduction in the risk of cardiovascular disease; improvements in lipid profile and blood pressure; improvement in endothelial function (a marker for risk of cardiovascular events); improvement in insulin sensitivity; and reduction of weight.30-34 Because patients with type 1 diabetes have lost normal hormonal responses needed to maintain euglycemia during exercise, they are at risk for hypoglycemia30,31,35; thus, a key issue for exercise in type 1 diabetes is reduction or prevention of hypoglycemic events. Even though evidence for the target outcomes of exercise for both type 1 and type 2 diabetes patients abounds and credible, diabetes-specific exercise recommendations are available,26,27,36 substantial decisions remain for providers who wish to engage patients in an exercise regimen. Major challenges to decision making about exercise include how to design and implement an efficient and effective exercise regimen, how to tailor it to an individual patient's motivation level and barriers so that exercise will be initiated and maintained, and how to select, measure, and achieve specific, desired outcomes. Given these decision-making challenges, it would be useful to have a comprehensive summary of evidence for and gaps in the literature for exercise interventions and outcomes in diabetes.
The objective of this review is to assess the literature and report evidence and gaps regarding outcomes of interventions addressing the self-management behavior of being active (exercise) among adult diabetes patients.
Primary Research Questions
Secondary Research Question
Types of Studies Intervention studies with a baseline measurement and at least 1 follow-up measurement of target outcomes were collected.14,15 Study designs included were randomized controlled trials, nonrandomized controlled trials, and well-conducted trials that included a premeasurement and postmeasurement in which subjects served as their own controls. In addition, because of the large volume of published literature on the topic of exercise interventions in diabetes, review and meta-analysis articles related to any of the target outcomes were also retrieved and evaluated. Reviews were included if they (1) specifically addressed exercise interventions of any type and outcomes in diabetes; (2) were systematic reviews or meta-analyses that included a methodological description of a targeted search strategy and frame, inclusion criteria, and results description; or (3) were technical reviews that were seminal and/or the most recent on a substantive topic. A summative review of the included reviews was conducted first to condense reporting of results in an efficient manner, given the scope and space limitations of the current report. Studies were included individually for review if they were not used in the included reviews or meta-analyses, if they were published after the included reviews or meta-analyses, or if they were included but warranted a separate mention to highlight a major finding or particular outcome category not addressed specifically elsewhere. Of the individual studies reviewed, study designs excluded were computer-generated decision-modeling studies and any studies that were not intervention studies. Studies for which only an abstract was available were not included; dissertations were also excluded because full text was not always available. Methodological quality of individual studies meeting the inclusion criteria was assessed using the American Diabetes Association (ADA) Evidence Grading System, whereby a quality grade of A, B, or C was assigned according to the quality/validity criteria. There were several topical reviews about diabetes and exercise interventions that proved to be useful resources but were not reviewed because they covered the same topics, were already included in substantive technical reviews, and/or did not meet the inclusion criteria stated above.* Readers are encouraged to refer to these topical reviews for in-depth information beyond the scope of this report.
Types of Participants
Types of Interventions
Types of Outcomes
Search Methods for Identification of Studies For the type 2 diabetes search, article titles generated by the tier 1 search were screened by 3 people, 2 people screened relevant abstracts in tier 2, and a third person verified abstracts included and excluded for retrieval of full-text articles. The tier 1 search for type 1 diabetes was conducted by 1 person, and the second tier screening was done by 2 people. The third tier of the process for both type 2 and type 1 diabetes included the review of these articles by 2 people for type 2 diabetes and 2 people for type 1 diabetes and retrieval of additional references from the citations of these articles.
Characteristics of Studies Reviews and meta-analyses. Table 1 reports the characteristics of the 7 review articles included in this report. There were 2 systematic literature reviews of 22 and 72 randomized controlled trials (RCTs) with interventions lasting any duration and looking at outcomes of weight loss or lifestyle behavior/skills development. There were also 2 meta-analyses of 7 and 14 RCTs that included only studies of structured, supervised exercise interventions lasting for 8 or more weeks and including outcomes of cardiorespiratory fitness (VO2max) or hemoglobin A1C (A1C) and BMI. In addition, there was 1 meta-analysis of 89 studies of various designs among obese adult patients looking at the clinical outcome of weight loss. There were also 2 technical reviews. All of these addressed type 2 diabetes,16,46-50 except 1 technical review.31 The review process, search frame, inclusion criteria, targeted interventions and outcomes, and major findings are reported in the Table 1.
Individual studies. The included individual studies are described in Table 2.51-82 There were 34 nonreview studies retained and reviewed within the scope of this report, ranging in publication year from 1994 to 2006, with most studies (30) being published after 2000. Among these, 18 were RCTs, 9 were nonrandomized controlled trials, and 7 had a pre and post design in which subjects served as controls for themselves. Sample sizes ranged from 7 to 340 subjects, with 23 studies addressing type 2 diabetes, 9 addressing type 1 diabetes, and 2 addressing both types.
Regarding participants, 8 studies were conducted with men only, 7 with women only (4 of which studied postmenopausal women), and 19 with both genders; 22 studies were conducted among non-US patients and 12 with US patients (3 of which involved only African American patients). In terms of target outcomes, no studies focused on specific measures of learning or knowledge outcomes, 18 addressed behavioral outcomes, 25 addressed clinical outcomes, 4 addressed humanistic outcomes (QoL), and none of the intervention studies addressed economic outcomes. Methodological quality of individual studies. Methodological quality was assessed using the ADA Evidence Grading System. A quality grade of A, B, or C was assigned according to the quality/validity criteria. Of the 34 individual studies included for review, 9 received a quality rating of A, 11 received a quality rating of B, and 14 received a quality rating of C.
Type 2 Diabetes Outcomes Evidence Learning/behavioral outcomes. While a few of the individual studies did suggest that they assessed educational outcomes, none identified specific measures of knowledge or skills acquisition. This has implications for substantiating validity and reliability of outcomes. Were patients actually engaging in the prescribed frequency, intensity, and duration of the exercise intervention using appropriately standardized and optimal exercise skills necessary to make claims about the association with achieving target outcomes? Several studies reported ranges of duration (a minimum of 30 minutes or working up to a minimum of 30 minutes) or work accomplished for exercise episodes, depending on whether the intervention involved aerobic exercise or resistance training and allowed frequency to meet a threshold criteria of acceptability (80%-90%). Validating that subjects are actually engaging in the exercise intervention prescription as set forth in the study and have acquired appropriate and optimal skills, is an important component of methodological rigor and should be conducted and reported in future studies. On a similar note, some of the included studies that did address exercise behavioral outcomes often relied on self-report and did not include a validation of self-reported level of exercise. However, the study conducted by Bjorgaas and colleagues,55 which used a pedometer to validate self-reported aerobic activity, did not show an increase in activity in an additional supervised (exercise physiologist and cardiologist) aerobic exercise program compared with controls. Future studies using similar or other validation measures would be helpful in determining the true impact of structured, as well as unstructured, exercise programs on exercise behaviors. In addition, the role of personal record keeping in exercise remains unclear. The study conducted by Gleeson-Krieg64 in 2006 concluded that maintaining a personal exercise record had no additional benefits compared with patients who kept no formalized record. However, it should be noted that participants self-selected themselves into the study and may have had an overall higher interest in exercise, even in the control group, than the general population did. In addition, the study was only 6 weeks in duration, and the authors reported that patients expressed discouragement that they made the effort to keep the record but their health care providers paid no attention to it at an office visit. Regarding patient characteristics, Agurs-Collins et al51 suggested that there may be a difference in the way genders initiate exercise behaviors; at 3 months, women reported to be significantly more likely to increase physical activity behaviors compared with men, but baseline differences in physical activity among genders were not reported. However, as in several studies, follow-up assessment at 6 months revealed no significant increase from baseline. Other studies that included both men and women did not emphasize gender comparisons. Another study suggested that a patient's physician may play a large role in the likelihood that a patient will achieve his or her intended exercise goals.59 This study found that type 2 patients were more likely to achieve their exercise goals when they received specific counseling from their physician, who used a structured checklist. Clinical outcomes. Overall, physical activity seems to have a positive effect on glycemic control. Numerous studies demonstrate that exercise can decrease A1C levels and often by an amount that should decrease the risk of complications.47 Consequently, the decrease in A1C level may be independent of weight loss.16 Intensive exercise regimens may result in even greater reductions in A1C levels.16 Concerning cardiovascular risk, those with low aerobic fitness and low physical activity may have greater cardiovascular mortality.16 Decreases in cardiovascular risk could result from the effects of exercise on decreasing blood pressure, BMI, and improving lipid profiles.53,62 Additional benefits from physical activity, specifically programs that combine resistance and aerobic exercise, may include decreased abdominal and visceral fat and obesity. However, the effects of exercise on BMI are still somewhat unclear, as some studies report no significant resulting effect.71,72 This finding and others may be affected by varying levels of intensity and duration of exercise interventions. In addition, other research suggests that weight loss and reduction in BMI are more significantly affected by dietary strategies.47-49 Part of the reason for this may be that BMI does not reflect changes in body composition. BMI will change less than expected when body fat (less dense) is lost and muscle mass (more dense) is concurrently gained. Using BMI or scale weight alone may lead to erroneous interpretations of weight loss outcomes. Regarding specific types of exercise, recent research on the effects of resistance exercise on type 2 diabetes suggests a positive impact on clinical outcomes. Trials involving resistance exercise have continued to demonstrate corresponding improvements in A1C levels,53,57 especially in a structured gym-based setting over a duration of 6 months.60 These results unfortunately seem to be lost in at least 1 study, when participants moved to a home-based program over a course of 1 year.60 This could perhaps be due to a lack of necessary equipment and support in a home environment as compared with the study setting in a gym. Strategies for maintaining adherence should also be addressed; creating accountability and an environment conducive to exercise (eg, structured gym-based setting) could also help support favorable long-term clinical outcomes. Combination aerobic plus resistance training exercise programs have been shown to create a significant improvement in insulin sensitivity compared with aerobic exercise alone.58 Resistance exercise alone has also been correlated with improved insulin sensitivity.69 Clinically, improved insulin sensitivity is a desirable outcome in type 2 patients who often require 1 or more insulin-sensitizing medications (eg, metformin, rosiglitazone, pioglitazone, pramlintide, exenatide). While exercise may not completely eliminate a patient's need for insulin-sensitizing agents or even insulin injections, it may help reduce the required doses of such agents, thereby improving the side effect profiles of these medications. However, more research on this important topic is required before any substantial conclusions can be made. Low cardiorespiratory fitness, an independent and substantial marker of long-term mortality in patients with diabetes, is a modifiable risk factor that must be addressed.27,46 Physical activity is the only intervention that directly affects cardiorespiratory fitness and is essential to the management of type 2 diabetes.46 The gold standard for measuring cardiovascular fitness and exercise capacity is VO2max, the maximal amount of oxygen consumed during exercise. A meta-analysis of 7 studies, presenting data for 9 randomized trials comparing exercise and control groups, was conducted to review and quantify the effects of exercise on cardiorespiratory fitness in patients with type 2 diabetes.46 Upon analysis across study samples data, an 11.8% increase in VO2max in the exercise group and a 1.0% decrease in the control group (postintervention standardized mean difference = 0.53, P < .003) was found. Higher exercise intensity was associated with larger improvements in VO2max and predicted postintervention weighted mean differences in A1C (r = –0.91, P = .002) to a larger degree than did exercise volume (r = –0.46, P = .26).47 Despite these positive results, the researchers noted that high-intensity exercise might prove too difficult or even hazardous for many individuals who have previously been sedentary and that these results would not be sufficient to advocate high-intensity exercise for all patients with diabetes. However, for patients who are already exercising at a moderate level of intensity, providers may wish to consider encouraging these individuals to increase their level of exercise intensity because of the potential additional benefits on both metabolic control and cardiorespiratory fitness.46 It is important to note that there have been few studies or reports that addressed the risks associated with the effects of exercise on diabetes complications17,18 and what types of pre-exercise assessments are warranted.13,16,18,25,26,35,37 Patients presenting with nephropathy, neuropathy, or retinopathy need special considerations when designing an exercise regimen. Maintaining blood glucose monitoring, foot care, and cardiovascular assessment are keys to minimizing the impact of exercise risks, and further research is warranted to review the implications of diabetes complications on exercise interventions. While the overall clinical benefits of exercise seem to outweigh the risks in type 2 diabetes, further research involving long-term maintenance of exercise studies and programs (greater than 1-year duration) is needed to adequately distinguish the effects of sustained exercise patterns on a person's future clinical outcomes, cardiovascular risk, and mortality. Humanistic outcomes. The impact of exercise on a patient's QoL is somewhat unclear. Some studies report that exercise results in no significant changes in a patient's QoL,67,68 whereas other studies suggest a positive exercise effect on humanistic outcomes in diabetes patients.78 Among the 3 type 2 diabetes studies that assessed QoL, methodological differences were apparent. Of the studies not showing exercise impact on QoL, the 6-month RCT conducted by Hendricks and Hendricks67 included a small (n = 30) convenience sample ($100 recruitment incentive) of sedentary African American men whose A1C at baseline was less than 8.3%. The intervention was educational and did not include a structured, supervised exercise component; the QoL measure was the SF-36. The nonrandomized controlled trial conducted by Holton and colleagues68 also included a small sample size (n = 31) with 3 arms, and the exercise intervention arm received a structured, supervised moderate exercise training for 10 weeks. Quality of life, measured by only 1 general SF-36 item, increased in all groups from baseline, but there were no significant differences between the intervention and control groups. The third study conducted in type 2 patients by Sykes and colleagues78 was an RCT among 36 Hong Kong Chinese patients (mean baseline A1C = 8.1%). The structured, 12-week exercise program resulted in a significant increase in QoL (SF-36) in the intervention group. The structured interventions produced improved QoL, but the increase was not significant in the Holton study, and the QoL measure included only 1 general well-being item from the SF-36. The studies included diverse, small samples who self-selected into the studies. Perhaps those choosing to participate were already exercising (except in the Hendricks and Hendricks study) or had somewhat controlled diabetes (see the mean baseline A1Cs noted above) and were not yet experiencing the deleterious effects of diabetes complications, suggesting that they may have started the studies with moderate to high QoL. None of these studies used diabetes-specific measures of QoL and may not have fully captured the impact of diabetes and exercise on general QoL. Among other literature in the general population, a cross-sectional study assessing QoL found that participants who performed light physical exercise reported the highest QoL.83 Interestingly, this study found that only higher intensity exercises were significantly associated with physical/health domains of QoL measures, while only low-intensity regimens were associated with mental/self-esteem domains. The authors suggest that this may reflect the possibility that as participants increase their intensity of exercise beyond what they are able to realistically achieve, feelings of failure and low self-esteem occur. Thus, the concept of setting patients up to succeed by prescribing smaller, incremental exercise goals is seemingly appropriate for use by diabetes educators today. Exercise goals should be realistic and achievable so that patients see improvements in not only clinical but also humanistic outcomes.
Type 1 Diabetes Outcomes Evidence Clinical/health status (humanistic) outcomes. In patients with type 1 diabetes, some studies have failed to show an improvement in glycemic control with exercise.26,30,32,33 Despite a decrease in insulin requirements in patients engaged in physical activity, nearly all studies suggest that overall glycemic control does not improve. Researchers speculated that this is likely due to patients' consuming more carbohydrates to prevent and/or treat hypoglycemia associated with exercise. Patients with type 1 diabetes do benefit from exercise in improvement in cardiovascular disease risk factors. Exercise improves lipid parameters,32,33 decreases blood pressure,32 and improves endothelial function in patients with type 1 diabetes.34 (Endothelial dysfunction is associated with arteriosclerosis and is regarded as a risk factor for cardiovascular events.) Exercise also leads to improved insulin sensitivity, weight loss,32 and improved QoL.83 The Wiesinger and colleagues82 study, in particular, was a nonrandomized controlled trial among 23 German men and women with a mean A1C at baseline of less than 8.5%. This study reported a significant increase in QoL (German adaptation of the SF-36) in the study group after 4 months of a structured, supervised (by a physician), aerobic (cycling) exercise regimen. In other literature, a cross-sectional study among type 1 patients suggested an association between engaging in exercise and improved sense of well-being.84 It is important to consider special precautions when prescribing exercise for patients with type 1 diabetes. Patients with type 1 diabetes have lost the normal hormonal mechanisms for maintaining euglycemia during exercise and thus are at risk of hypoglycemia and hyperglycemia with exercise.31 Patients with type 1 diabetes need instruction on how to safely exercise and avoid hypoglycemia associated with exercise.26,65 A number of strategies to reduce the risk of exercise-induced hypoglycemia have been evaluated, including type and amount of carbohydrate ingested prior to exercise,85 insulin dose adjustments,76,86 exercise intensity,56,87 and time of day exercise occurs.88 Self-monitoring of blood glucose is essential for patients with type 1 diabetes to individualize 1 or a combination of these strategies to reduce the risk of hypoglycemia.26
Implications for Research A systematic review of self-care interventions in older African American or Latino adults was searched in the literature up through December 2000.90 Only 1 of the 12 included studies directly assessed the impact of exercise in African American men.51 Within the scope of the current review, which searched up through April 2006, only 2 additional studies were identified that addressed African American men and women,67,74 and no studies meeting the inclusion criteria directly assessed exercise outcomes in Hispanic populations. In addition, there were no studies meeting the inclusion criteria of this review that addressed exercise outcomes in rural versus urban populations. Future research should develop and test culturally sensitive and environmentally relevant exercise interventions for these underserved populations. Another concern is that many of the reviewed studies include very small sample sizes and study durations too brief to assess long-term effects of exercise interventions on outcomes. Clearly, the nature of behavioral intervention research presents many challenges, including access to, recruitment of, and retention of patients. In the United States, the 2004 initiation of the Health Insurance Portability and Accountability Act compliance requirements in health care and research settings makes this even more challenging. Randomized controlled trials that include blinding patients to the study arm are ideal for research purposes but are not always purely practical in research conducted in health care settings. Future research efforts should attempt to adhere to rigorous methods while striving for larger sample sizes. One strategy may include collaboration with additional health care providers to gain access to additional patients. Another population issue to consider is the fact that many of the studies involved well-controlled diabetes patients who self-selected into studies. While these patients are willing and are considered lower risk patients, the measured impact of an intervention on outcomes may not be as evident in well-controlled as in poorly controlled patients who have a more significant capacity, or range, for improvement. In addition, many type 2 diabetes studies included participants with a mean age of more than 45 years; studies are needed to address the younger adult population and adolescents, among whom type 2 diabetes is becoming more common. It is often the adherent and healthier patient who chooses to participate in a study or program. Greater efforts should be made to recruit poorly controlled patients into studies. While using financial incentives to recruit and retain patients into studies is controversial, 1 pilot study reviewed in this report used a modest financial incentive ($25), and nearly two-thirds of the subjects recruited had an A1C level greater than 7.0%.75 In addition, another reviewed study used a $100 incentive, with $50 given at each of the before and after measurements; the baseline mean A1C level of subjects was 8.1%.67 Outside of the scope of this review topic, others have used a variety of means to motivate patients to participate in their studies, and this is a topic that warrants further research for strategies to include more poorly controlled diabetes patients in studies. Study setting and design. It is first important to note that the methodological quality of included individual studies was mixed. Only 9 of 34 studies achieved a quality rating of A, suggesting room for improvement in the rigor of future study designs. The most prevalent methodological flaw among studies was a small sample size. The active engagement nature of well-conducted exercise intervention programs places a burden on the participant in terms of time, commitment, and energy. It is often the patients who are already engaging in the target behavior or who are well-controlled who are interested in participating, and yet they are not the patients in most need of change. This is clearly a limitation of some of the included studies and should be addressed in future research. Self-management with exercise is an elusive behavior in research and in the reality of a patient's life. The concept of self-management was conceived to allow patients to manage their diabetes on their own on a daily basis. Translation of exercise outcomes research into practice is challenged by the fact that most well-conducted exercise outcomes studies take place in a controlled, artificial setting (eg, in a gym or laboratory setup for the temporary purpose of the study). The setting and protocol are often not representative of what is feasible in the patient's daily life and/or what will be available after the study ends. Certainly, these structured settings and protocols are important to generating valid and reliable outcomes data. They also appear to be more effective at generating improved outcomes than studies involving education alone and sending patients to unstructured exercise on their own in the reality of their daily lives. Initiatives are being undertaken through the National Institutes of Health to address the paucity of studies regarding the effective translation of intervention research into the lifestyles of patients, particularly at the community level.91 In seeking feasible means of translating research into a patient's daily practice, perhaps exercise outcomes should be directly studied within other contexts of reality in a patient's daily life. One such context for which research is evolving includes employer-based disease management initiatives. Such settings should be formally researched in greater depth than the minimal current literature reveals. The setting for such a study has the potential to represent something fairly realistic and ongoing in a working diabetes patient's daily life, particularly if his or her employer offers and supports the initiative. Such initiatives are on the rise and have been able to show not only better patient outcomes but also improvements to the company's bottom line. One example is a pharmacist-delivered diabetes disease management program for employees of the city of Asheville, North Carolina. The disease management program resulted in not only improved clinical outcomes for employee patients but also significantly increased productivity, decreased sick days, and decreased direct medical costs.91-93 The program is currently being formally reproduced in several US cities with similar findings; preliminary results for the exercise education component suggest an increase in patient-reported exercise goal achievement.94 Longer studies are needed to determine whether clinical outcomes such as BMI would be affected by exercise over time. Postintervention follow-ups of at least 6 to 12 months are indicated in the conclusions of some of the included review studies. QoL studies. Within the literature, for both type 1 and type 2 diabetes, minimal empirical research assessed the direct, independent impact of exercise interventions on outcomes related to QoL. Among the few studies in this report that included QoL or well-being as a measured outcome, the findings were mixed. This can perhaps be explained by the fact that these studies included only a generic health-related QoL measure and did not also include a disease-specific measure. Generic measures such as the SF-36 or SF-12 cannot always be sensitive to the unique experience of managing diabetes and its health consequences. Since diabetes is highly comorbid with conditions such as depression and obesity, both of which are confounding variables that affect general QoL separately from the impact of diabetes, it is important to also include in QoL studies a disease-specific measure along with the generic measure to separate out the differences in impact.95 More studies evaluating the impact of structured and unstructured exercise programs, as well as resistance exercise programs, are needed to fully distinguish the influence of activity frequency, intensity, volume, and motives on QoL outcomes. Economic studies. Direct measures of the cost-effectiveness of exercise interventions in adults with diabetes have not been reported in a published prospective clinical trial. A technical review of studies addressing the economic analysis of several types of interventions for diabetes, including exercise, indicated that there were no studies directly addressing the topic and concluded that cost-effectiveness of exercise interventions in diabetes is unclear.43 Future research studies should be conducted to assess the direct cost-effectiveness of exercise interventions. Type 1 diabetes research. For type 1 diabetes, much research exists on physiological associations with exercise but very little on learning/behavioral outcomes or on clinical outcomes other than interventions to reduce the risk of hypoglycemic events associated with exercise. Only recently have studies been reported that assessed the same clinical outcomes (eg, cardiovascular risk reduction and others) typically associated with type 2 diabetes,32-34 and more work is needed to substantiate findings in long-term trials. In addition, recent mainstream adoption of the use of insulin pump therapy in adult diabetes patients suggests that studies assessing exercise outcomes need to be conducted in this population. An important gap in the literature suggests a need for outcomes studies among type 1 diabetes patients. One of the limitations of the current review is lack of space to give adequate attention to exercise outcomes in type 1 diabetes that have been reported thus far. In addition, the last technical review for exercise in type 1 diabetes was published by Wasserman and Zinman in 1994.31 Many advances in science, insulin types, and dosing methods and regimens have occurred over the past 12 years, warranting a full technical review in the near future, as also suggested by Sigal and colleagues in the 2004 technical review for exercise in type 2 diabetes.16
Implications for Practice In addition to the results reported in this review, 2 well-conducted review articles were published since the submission of this review,96,97 including a Cochrane systematic review of 14 RCTs of structured exercise interventions in type 2 diabetes. The conclusions of these 2 reviews confirm and/or complement those of the articles included within this review. Even though the weight of the participants did not change significantly in the studies, other clinical outcomes did improve with exercise. This should be stressed to patients who may just focus on weight. Because various exercise types were found to be beneficial, exercise regimens should be individualized, allowing patients to set priorities regarding activities they most enjoy or would find most suitable. This may improve the sustainability of exercise over long periods. Recommendations regarding the evaluation of the patient with diabetes prior to recommending an exercise program are clearly defined. Patients with diabetes beginning an exercise program more vigorous than brisk walking should be evaluated for conditions that are associated with an increased likelihood of cardiovascular disease or that may predispose a patient to injury (eg, severe autonomic neuropathy, severe peripheral neuropathy, and preproliferative or proliferative retinopathy) and for contraindications for certain types of exercise.16,97 Cardiac testing should also be conducted in patients with diabetic autonomic neuropathy before beginning physical activity more intense than that to which they are accustomed. An exercise electrocardiogram stress test should be performed in previously sedentary individuals with microalbuminuria and proteinuria. It may be best to encourage non-weight-bearing activity such as swimming, biking, or arm exercises in patients with severe peripheral neuropathy. Vigorous aerobic or resistance exercise may be contraindicated in the presence of proliferative or severe nonproliferative diabetic retinopathy because of the possible risk of vitreous hemorrhage or retinal detachment. There are no studies indicating the time interval to wait between successful photocoagulation and initiation or resumption of resistance exercise; however, a waiting period of 3 to 6 months may be advisable. Age and previous activity level should also be considered. Previous ADA recommendations to withhold exercise in patients with type 2 diabetes with blood glucose levels >300 mg/dL (16.5 mmol/L) is probably more cautious than necessary, especially postprandially.97 In summary, in the absence of complications, general recommendations to improve glycemic control, assist with weight management, and reduce cardiovascular disease risk suggest at least 150 minutes per week of moderate-intensity aerobic physical activity (40%-60% of VO2max or 50%-70% of maximum heart rate) and/or at least 90 minutes per week of vigorous aerobic exercise (>60% of VO2max or >70% of maximum heart rate) over at least 3 days per week, with no more than 2 consecutive days without physical activity. Patients with type 2 diabetes without contraindications should also be encouraged to perform resistance exercise 3 times a week, targeting all major muscle groups, working up to 3 sets of 8 to 10 repetitions at a weight that cannot be lifted more than 8 to 10 times with initial supervision and periodic reassessments by a qualified exercise specialist. Patients taking insulin or secretagogues should check their blood glucose levels before, after, and several hours after completing physical activity or at least until they know what their usual glycemic response to exercise is. It is possible to derive only these general guidelines for exercise intervention and not a specific exercise regimen that is optimal for all patients with diabetes. The heterogeneous nature of even the RCTs, including their participants and types of exercise at varying frequencies, durations, and intensities, suggests that there is no discrete 1 best exercise intervention that can be generalized to a population attempting to achieve particular outcomes outside of a controlled experimental setting. The state of the research still suggests that to be effective, an exercise regimen needs to be individually tailored to a person's needs, priorities, circumstances, and capabilities while also relying on the extensive, evidence-based practice guidelines that are available to aid in decision making.13,16,26,27,97
It is important to note that including the review of reviews along with the review of individual studies deviated from the typical process for a systematic review. Given the scope of the exercise/being active target behavior and the hundreds of articles that could have been included for individual review, and given that there already exist several well-conducted, landmark reviews or meta-analyses covering portions of the objectives of this review, it was decided to revise the systematic review process. The most efficient means of giving full coverage to the topic of exercise/being active in 1 review of both type 1 and type 2 diabetes, across multiple types of exercise and the multiple outcomes categories, meant a deviation from the normal process and structure for a systematic review, which is best suited for addressing a narrow research question regarding 1 specific type of patient, intervention, and outcome. There is still much to learn about how to develop interventions that will sustain exercise behavior and outcomes for a period of time longer than that examined within the intervention studies included here. In addition, while the outcomes findings within this review are important, the practitioner is cautioned in his or her decision making to consider all aspects of an intervention program, so that it may be best tailored to the individual patient's needs and priorities. Long-term feasibility must also weigh evidence that is lacking in the literature: availability of resources (economic outcomes), reinforcement of sustained exercise behavior over time (behavioral outcomes), and an empathic understanding of what diabetes and its treatments mean to an individual patient who must live with the disease, manage it on many levels, and endure the consequences of action or inaction on a daily basis (humanistic outcomes).98,99
* References 13, 17, 18, 25, 27, 30, 35, 37-44. 
The Diabetes Educator, Vol. 33, No. 6,
962-988 (2007)
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
Sign In to gain access to subscriptions and/or personal tools.
Top
Abstract
Background