Comparison of Combined Aerobic and High-Force Eccentric Resistance Exercise With Aerobic Exercise Only for People With Type 2 Diabetes Mellitus

Robin L Marcus, Sheldon Smith, Glen Morrell, Odessa Addison, Leland E Dibble, Donna Wahoff-Stice, Paul C LaStayo


Background and Purpose: The purpose of this study was to compare the outcomes between a diabetes exercise training program using combined aerobic and high-force eccentric resistance exercise and a program of aerobic exercise only.

Subjects and Methods: Fifteen participants with type 2 diabetes mellitus (T2DM) participated in a 16-week supervised exercise training program: 7 (mean age=50.7 years, SD=6.9) in a combined aerobic and eccentric resistance exercise program (AE/RE group) and 8 (mean age=58.5 years, SD=6.2) in a program of aerobic exercise only (AE group). Outcome measures included thigh lean tissue and intramuscular fat (IMF), glycosylated hemoglobin, body mass index (BMI), and 6-minute walk distance.

Results: Both groups experienced decreases in mean glycosylated hemoglobin after training (AE/RE group: −0.59% [95% confidence interval (CI)=−1.5 to 0.28]; AE group: −0.31% [95% CI=−0.60 to −0.03]), with no significant between-group differences. There was an interaction between group and time with respect to change in thigh lean tissue cross-sectional area, with the AE/RE group gaining more lean tissue (AE/RE group: 15.1 cm2 [95% CI=7.6 to 22.5]; AE group: −5.6 cm2 [95% CI=−10.4 to 0.76]). Both groups experienced decreases in mean thigh IMF cross-sectional area (AE/RE group: −1.2 cm2 [95% CI=−2.6 to 0.26]; AE group: −2.2 cm2 [95% CI=−3.5 to −0.84]) and increases in 6-minute walk distance (AE/RE group: 45.5 m [95% CI=7.5 to 83.6]; AE group: 29.9 m [95% CI=−7.7 to 67.5]) after training, with no between-group differences. There was an interaction between group and time with respect to change in BMI, with the AE/RE group experiencing a greater decrease in BMI.

Discussion and Conclusion: Significant improvements in long-term glycemic control, thigh composition, and physical performance were demonstrated in both groups after participating in a 16-week exercise program. Subjects in the AE/RE group demonstrated additional improvements in thigh lean tissue and BMI. Improvements in thigh lean tissue may be important in this population as a means to increase resting metabolic rate, protein reserve, exercise tolerance, and functional mobility.

Type 2 diabetes mellitus (T2DM) is a chronic illness marked by decreased insulin sensitivity and overall poor glucose control. A universally accepted component of the nonpharmacologic treatment for T2DM is exercise.1 Historically, aerobic exercise has been advocated for improving glucose control, although there is an increasing body of literature supporting the merits of resistance exercise as a beneficial component in the management of T2DM.24 Current consensus statements from several professional associations now regularly recommend both aerobic and resistance exercise for the management of T2DM.57 Moreover, when combining resistance exercise with aerobic exercise, the effect of improving glucose control has been reported to be greater than for aerobic exercise alone.8,9 For greater discussion of the role of exercise in glycemic control, see the review articles by Turcotte and Fisher10 and Gulve11 in this issue.

The benefits of resistance exercise are not limited to enhanced glucose control but also include the maintenance and improvement of muscular strength (force-generating capacity), endurance, and power. In addition, increases in lean tissue mass are possible if the resistance exercise induces high muscle forces.12 Improved lean tissue mass contributes to the maintenance of basal metabolic rate,13 promotes functional independence,14 and helps to reduce fall risk.15 To maximize lean tissue improvements from resistance exercise, high-intensity resistance exercise has been advocated.12 The greatest force production stimulus to increase muscle size and strength is possible when an external force exceeds that of the muscle and the muscle lengthens eccentrically. Eccentric muscle contractions can result in 2 to 3 times greater force production than more-traditional isometric or concentric muscle contractions.16 Additionally, this high force production occurs at the lowest metabolic cost, making eccentric-induced, negative work resistance exercise less difficult for individuals to perform.16 This is especially advantageous for individuals whose aerobic abilities are limited due to complications (eg, cardiovascular disease) associated with diabetes. Although some studies indicate an increase in insulin resistance after exhaustive eccentric work that may be related to inflammation due to muscle injury,17,18 the impact of repeated exposures of eccentric exercise on insulin resistance has not been studied in individuals with T2DM.

Considering that 80% to 90% of glucose in the body is disposed in skeletal muscle,19 it would not be surprising that an increase in lean tissue mass alone would result in enhanced glucose control. Additionally, resistance exercise may result in changes in muscle composition that may alter metabolism. Intramuscular fat (IMF), or the fat deposition within and between muscle groups and beneath the muscle fascia,20 has recently gained attention as a potential contributor to glucose homeostasis. In a mechanism that has yet to be identified, IMF has been shown to have a strong negative association with insulin sensitivity in individuals with T2DM.21 Taken together, this strongly suggests that increased skeletal muscle with minimal infiltration of fat is beneficial in people with T2DM.

The purpose of this study was to compare a diabetes exercise training program using combined aerobic and high-force eccentric resistance exercise with a program of aerobic exercise only. Specifically, we were interested in whether the addition of a high-force eccentric resistance component, previously reported to increase thigh lean tissue mass,15,22 would result in improvements in glucose control and physical performance similar to those of a program of aerobic exercise only.



Adults with T2DM were recruited from the accessible population of people receiving care for diabetes through local hospitals and physician offices. Interested subjects contacted the primary investigator (RLM), and a telephone interview followed to determine eligibility by screening.

Eligible subjects included those who were sedentary (not participating in regular aerobic or strengthening exercise over the 6 months prior to entering the study) and who were willing to commit to a 16-week, supervised exercise program. All subjects had a diagnosis of T2DM, confirmed by either an oral glucose tolerance test of >200 mg/dL or fasting blood glucose of >126 mg/dL on 2 separate occasions within the previous year. After informed consent was obtained, a medical history (including resting heart rate, blood pressure, date of T2DM diagnosis, medications, and comorbidities) was collected. Subjects were medically cleared for exercise by their primary care physician in the form a written prescription. If requested, a pre-exercise stress test was arranged to ensure clearance for participation. All subjects had normal cognition, as determined by a Folstein Mini-Mental Status Examination score of >23. Individuals with severe cardiac disease (New York Heart Association class III)23; uncontrolled hypertension (systolic blood pressure >165/95 mm Hg); orthopedic problems that limited their ability to use exercise equipment without pain; central or peripheral nervous system disorders; diabetic retinopathy; myopathy; inability to concentrate, follow directions, or work independently; neurologic insult that resulted in mobility impairment; rheumatological disease that affected mobility; impaired knee flexion of <90 degrees; or extreme claustrophobia were excluded from participation. In order to minimize threats to internal validity such as compensatory rivalry,24 subjects were sequentially enrolled in 1 of 2 groups, with the first 7 subjects in a combined aerobic and resistance exercise (AE/RE) group and the next 8 subjects in an aerobic exercise–only (AE) group.


Subjects reported to the exercise clinic 3 times per week for either a combined aerobic and resistance exercise progression or an aerobic exercise–only progression. Prior to each session, all participants’ blood glucose, resting blood pressure, and heart rate were recorded on an individual exercise log sheet. A pre-exercise blood pressure of <140/90 mmHg was required for the subjects to begin exercise.25 If the subjects’ pre-exercise blood pressure was ≥140/90 mm Hg, they were asked to sit quietly for 10 minutes and were reassessed. No exercise was permitted that day if a lower resting blood pressure was not achieved. If subjects arrived with a blood glucose level of <100 mg/dL, they were given the choice of having a 15-gm carbohydrate snack or exercising for 20 to 30 minutes and reassessing blood glucose levels to make sure that they was not dropping. If the pre-exercise blood glucose level was >300 mg/dL, subjects began exercise and were reassessed in 20 to 30 minutes to make sure the blood glucose level was not increasing. If the blood glucose level was increasing, exercise was stopped. Any time subjects had symptoms of hypoglycemia, they monitored their blood glucose level and had a carbohydrate snack if it was indicated (a blood glucose level of <70 mg/dL=15 gm of carbohydrate, a blood glucose level of <50 mg/dL=20 gm of carbohydrate, and a blood glucose level of <40 gm=30 gm of carbohydrate). After monitoring and controlling their blood glucose level, they were asked to sit for 15 minutes and then reassess their blood glucose level. If no increase in blood glucose level was noted, these steps were repeated until the blood glucose level rose above 70 mg/dL.26 Refer to the article by Gulve11 in this issue for additional information on exercise precautions for people with diabetes mellitus.

Initial aerobic exercise intensity was based on 60% of age-predicted heart rate or 60% of maximum heart rate achieved on a stress test. Each subject was progressed to 85% of this value over the 16-week training program. All subjects participated in an aerobic warm-up of at least 5 minutes, keeping their rating of perceived exertion (RPE) in the “light” range on the Borg Rating of Perceived Exertion Scale.27 Following the warm-up, subjects in the AE group were instructed to increase their intensity into their target heart rate range and an RPE commensurate with “somewhat hard.” Staff members met with each subject during their aerobic exercise to monitor heart rate and RPE to ensure that subjects were exercising at their prescribed intensity. Subjects were encouraged to increase their aerobic exercise time by 5 minutes each week until they reached 50 minutes, which they would maintain until the end of the 16 weeks. Subjects exercised in their respective groups under supervision. The equipment available consisted of treadmills, stationary bicycles, recumbent steppers, elliptical steppers, and rowing machines. Subjects were encouraged to use a variety of the available equipment and were instructed in their proper use, if necessary. All subjects used at least 2 different exercise machines each session, and each subject used at least 4 different exercise machines throughout the 16-week training period.

Subjects in the AE/RE group participated in the above aerobic exercise program plus resistance exercise on a recumbent eccentric stepper.* Prior to training, the stepper seat setting was individually adjusted to each subject's leg length, and safety guidelines were reviewed. The recumbent eccentric stepper was powered by a 3-hp motor that drives the foot pedals in a “backward” direction (ie, toward the individual). Eccentric muscle contractions occurred when the subject attempted to resist this motion by pushing on the pedals (with verbal instruction to “try to slow down the pedals”) as the pedals moved toward the subject. Because the magnitude of the force produced by the stepper exceeded that of the subjects, the pedals continued to move toward the participant at a constant velocity, resulting in eccentric contractions of the knee and hip extensors, including the quadriceps femoris muscles (Fig. 1). The subjects began with a 5-minute session on the stepper and progressed to a maximum of 20 minutes over the next 3 to 4 weeks. The progression of the eccentric exercise work rate was determined as a function of the perceived exertion (RPE) using a “target” workload on a computer monitor and is summarized in Table 1. Once the subjects achieved an RPE of “somewhat hard,” they were instructed to maintain that RPE for the duration of the exercise program. A visual analog scale (VAS) was used to monitor muscle pain prior to each session, and heart rate and RPE values were collected at the halfway point of each session.

Figure 1.

Eccentric stepper. High muscle forces are generated on an eccentric stepper powered by a 3-hp motor that drives the pedals. As the pedals move toward the subject (blue arrow), the subject resists by applying force to the pedals (red arrow). Because the magnitude of force produced by the motor exceeds that produced by the subject, the leg extensors (green arrows) work eccentrically (lengthening), creating negative work.

Table 1.

Eccentric Resistance Exercise Progression for Combined Aerobic and Eccentric Resistance Exercise Group

Pretraining and Posttraining Measurements

All pretraining and posttraining measurements were taken within the week before beginning or following the cessation of the respective exercise programs. Glucose control and creatine kinase blood samples were collected simultaneously and at the same time of day both before and after training. Muscle structure and physical performance measurements, both before and after training, were collected on a day when the subjects did not participate in any exercise. Muscle pain scores were collected before each training session for the AE/RE group only. The mean pain scores from week 3 were compared with those from week 16 for the analysis. Week 3 pain scores were chosen for the comparison with week 16 pain scores because the subjects had had completed at least 2 moderate efforts of resistance exercise by the third week of training.

Glucose control.

Overall glucose control was measured as glycosylated hemoglobin (HbA1c or A1c), with a venous blood sample taken after a 12-hour overnight fast. A1c was measured with standard techniques,28 and all samples were measured by the same laboratory. A1c is a form of hemoglobin that is proportional to average blood glucose concentration over the previous 4 to 12 weeks. The normal reference range of A1c is approximately 4% to 5.9%, and a value of less that 7% is recommended for people with T2DM.28

Muscle structure.

Average mid-thigh cross-sectional areas (CSAs) of lean tissue and IMF were determined using magnetic resonance imaging (MRI) of both thighs. Participants were placed supine in the MRI magnet with legs relaxed. Imaging was performed on a 3.0-T Siemens Trio system using the spine coil incorporated in the scan table posteriorly combined with the torso array coil anteriorly. Two successive gradient-recalled echo sequences were performed, each of which acquired 2 echoes, with echo times of 3.15 and 5.25 milliseconds in the first acquisition and 4.20 and 6.30 milliseconds in the second acquisition. These echo times correspond to in- and out-of-phase configuration of fat and water at 3.0 T. Axial images (n=32) with slice thicknesses of 1 cm were obtained, giving a 32-cm field of view in the superior-inferior direction centered on the mid-thigh. An image matrix of 128 × 128 was used over a variable field of view chosen to adequately include the entire axial cross-section of the bilateral thighs, typically less than 36 cm. Sequence repetition time was 20 milliseconds, giving an imaging time of about 48 seconds per scan and resulting in a total imaging time of under 2 minutes. Readout bandwidth was 750 Hz per pixel, corresponding to a fat shift of just over one-half pixel.

The images corresponding to echo times of 3.15, 4.2, and 5.25 milliseconds were used to create separate fat and water images with the 3-point Dixon method.29 T1 weighting effects in the fat and water images then were corrected based on the gradient-recalled echo signal equation, assuming a T1 of 340 milliseconds for the signal component at the fat resonant frequency and a T1 of 640 milliseconds for the signal component at the water resonant frequency. These T1 values had been previously experimentally determined in the thigh of a volunteer using an inversion-recovery technique with exponential curve fitting.

Fat-water separation imaging gives images corresponding to tissue components at the resonant frequency of fat and at the resonant frequency of water. Only fat contributes to the fat resonant frequency images. The water resonant frequency image includes signal from water in nonfat soft tissues, including muscle, nerves, blood, and vessels. No significant signal is received from cortical bone. Water-only images in the thigh represent primarily muscle, with minimal contribution from nerves and vessels.

After fat and water image signal intensities were corrected for T1 effects, the percent volume fraction of fat and nonfat tissue was calculated for each image pixel using a published algorithm.30 Images were created showing the percent volume fraction of fat and nonfat tissue in each pixel. Total fat and lean tissue then were calculated by manually drawing a region of interest defined by the fascia latae and summing the value of percent fat fraction and percent nonfat tissue fraction over all pixels within this region of interest with Scion Image for Windows, version This sum was multiplied by the area of each pixel to give total fat and lean tissue CSAs within the region of interest (Fig. 2).

Figure 2.

Representative pretraining (left) and posttraining (right) magnetic resonance images of an eccentrically trained individual. Region of interest is drawn around the fascia latae of the thigh. White material within the region of interest is intramuscular fat.

This method accurately measures fat and lean tissue in pixels that contain both. Such pixels are not assigned wholly to fat or lean tissue based on a threshold value, but instead are allowed to make fractional contributions to the fat and lean tissue CSA calculations. This allows microscopic fat within muscle tissue as well as thin planes of fat adjacent to fascial planes to be taken accurately into account, even when image resolution is inadequate to delineate them visually.

The same investigator, blinded to time point of the scan and slice location, performed measurements of individual subjects before and after training. This method has been used previously in our laboratory, resulting in an average interclass correlation coefficient .99 (range=.89–.99) across multiple images.22

Physical performance.

The Six-Minute Walk Test (6MWT), a measure of the distance a person walks in 6 minutes, was used to assess overall physical performance. Subjects were asked to cover as much distance as possible within 6 minutes without running. The 6MWT has been shown to be reliable and valid in detecting differences in mobility performance and has high test-retest reliability.31

Muscle damage.

Acute exposure to eccentric muscle contraction has previously been associated with insulin resistance, possibly due to a muscle damage response.17 Therefore, we evaluated serum creatine kinase (CK) concentrations and muscle pain VAS ratings as indirect measures of muscle damage. Serum CK samples were collected via venous blood draws before and after the 16-week training program, and the serum CK concentrations (in units per liter) were calculated using commercial enzymatic analysis. Muscle pain was determined (in centimeters) by the use of a 10-cm VAS anchored at 0 cm (“no pain”) and at 10 cm (“worst possible pain”). Subjects marked their leg muscle pain before each training session.

Data Analysis

Data were analyzed with SPSS version 16.0.§ Descriptive statistics were calculated for demographic variables and dependent measures. The assumptions of parametric statistical tests were tested via tests of normality and homogeneity of variance. In all cases, the assumptions were met and, therefore, parametric tests were performed. Baseline characteristics of the groups were compared using t tests for independent samples. To compare the effects of training type on muscle composition, glucose control, and physical performance, separate 2 × 2 repeated-measures analyses of variance were done. Time was the within-subjects factor, with 2 levels (pretraining and posttraining). Group was the between-subjects factor, with 2 levels (AE/RE group and AE group). When an interaction effect was found, the mean between-group difference and 95% CI were reported. To gain a clearer picture of the differential response of the groups, the magnitude of effect from pretraining test to posttraining test was estimated using calculations of the mean within-group difference and 95% CIs for all dependent variables. The level of significance was set at P<.05. Finally, to assess muscle damage, we compared serum CK concentrations and muscle pain VAS scores for each group before and after training with separate 2 × 2 repeated-measures analyses of variance.

The sample size was determined using previous studies of exercise effects on glucose control in individuals with T2DM.4,8,9,32 These studies have yielded effect sizes ranging from 0.7 to 1.2. Based on a conservative effect size of 0.80, an alpha level set at .05, and a desired power level of .80, it was estimated that 14 participants (7 per group) would be necessary to detect a significant statistical and clinical change in glucose control.


Fifteen subjects with T2DM completed the 16-week supervised exercise-training program. Six subjects were required by their referring physicians to undergo a pre-exercise stress test, and the remaining subjects were cleared by their physicians to exercise without a stress test. No subjects were screened out because of unacceptable performance on the pre-exercise stress test. Both the AE/RE and AE groups were similar in body mass index (BMI), HbA1c, and 6MWT performance before training. The AE/RE group was significantly younger than the AE group (Tab. 2).

Table 2.

Subject Characteristicsa

The AE/RE group completed a mean of 2.6 (SD=0.28) exercise sessions per week. Estimated eccentric work increased from 15.5 kJ (52 W) per session to 194.8 kJ (162 W) per session, and perceived exertion was incrementally increased from “very light” (8.5) to “somewhat hard” (13.0) over the first 3 weeks of training. Subjects maintained their perceived exertion levels at “somewhat hard” throughout the rest of the 16-week program. Mean maximum heart rates during the eccentric exercise were 113 bpm during week 3 (ie, after the first 2 weeks of ramping-up exertion and work) and 122 bpm during week 16. Aerobic exercise time in the AE/RE group increased to a mean of 45.2 (SD=4.6) minutes per session by the end of the 16-week program. Perceived exertion was maintained at “somewhat hard” (13.0) over the 16 weeks of training. Mean maximum heart rates during the aerobic exercise were 130 bpm during week 1 and 130 bpm during week 16.

The AE group participated in 2.5 (SD=0.31) exercise sessions per week, and their aerobic exercise time increased to a mean of 52.3 (SD=3.8) minutes per exercise session by the end of the 16-week program. Perceived exertion was maintained at “somewhat hard” (13.0) over the 16 weeks of training. Mean maximum heart rates were 142 bpm during week 1 and 142 bpm during week 16.

There was no significant interaction between groups with respect to change in A1c measurements. There was a significant main effect for time (P=.02). The mean within-group A1c change was −0.59% (95% CI=−1.5 to 0.28) for the AE/RE group and −0.31% (95% CI=−0.60 to −0.03) for the AE group. There was a significant interaction (P<.01) between group and time with respect to change in thigh lean CSA, with greater increases experienced by the AE/RE group (mean between-group difference=20.4 cm2 [95% CI=13.2 to 27.7]). There was no significant interaction between groups with respect to change in mean thigh IMF CSA. There was a significant main effect for time (P<.01). The mean within-group thigh IMF CSA change was −1.2 cm2 (95% CI=−2.6 to 0.26) for the AE/RE group and −2.2 cm2 (95% CI=−3.5 to −0.85) for the AE group. There was no significant interaction between groups with respect to change in mean 6MWT distance. There was a significant main effect for time (P<.01). The mean within-group thigh 6MWT distance change was 45.5 m (95% CI=7.5 to 83.6) for the AE/RE group and 29.9 m (95% CI=−7.7 to 67.5) for the AE group. There was a significant interaction (P<.01) between group and time with respect to change in BMI, with greater reduction experienced by the AE/RE group (mean between group difference=−2.1 kg/m2 [95% CI=−3.4 to −0.9]) (Tab. 3).

Table 3.

Outcomes for Glucose Control, Muscle Structure, Physical Performance, Muscle Damage, and Body Mass Indexa

The mean CK levels did not differ significantly and the CK values did not exceed the threshold of muscle damage at any time point (P=.28) (study overall mean=128 U/L muscle; damage threshold ≥200 U/L). Muscle pain VAS scores were minimal (study overall mean=0.8). Muscle pain VAS scores at week 3 (following at least 2 eccentric exercise sessions at an RPE of “somewhat hard”) averaged 2.0 (SD=2.3). The most common complications were blisters, which occurred in 3 of the AE/RE group participants. None of the blisters led to ulcerations. Some subjects in the AE/RE group experienced mild patellofemoral soreness that was relieved by temporarily decreasing exercise intensity. In more than 600 total exercise sessions, there were less than 5% hypoglycemic episodes. All hypoglycemic episodes were resolved with the steps identified in the “Method” section. No subject was stopped from participation because of a hypoglycemic episode.


In this study, we sought to compare a combined aerobic and high-force eccentric resistance exercise program with a program of aerobic exercise only and to evaluate the efficacy of high-force eccentric resistance exercise on glucose control, muscle structure, and physical performance. Significant improvements in long-term glycemic control (A1c levels), fat composition in the muscle structure (IMF), and physical performance (6MWT distance) occurred in both the AE/RE and AE groups after participating in a 16-week training program. The AE/RE group demonstrated additional increases in thigh lean tissue CSA and decreases in BMI.

Improved Glucose Control

The glucose control (A1c) improvements seen in our study of individuals with T2DM (−0.6% in the AE/RE group, −0.3% in the AE group) are clinically relevant and consistent with those reported in the literature for aerobic, resistance, and combined aerobic and resistance exercise programs of similar duration. In studies of middle-aged and older men and women participating in aerobic, resistance, or combined exercise programs, average A1c reductions of 0.6% (range=0.2%−1.3%) have been reported.3336 A recent meta-analysis of the effects of different modes of exercise training on glucose control in adult patients with T2DM identified mean effects of all 3 modes of exercise in studies lasting ≥12 weeks (and up to 2 years) as having reductions of 0.8% (SD=0.3%).1 This meta-analysis suggested that people with more-severe disease experienced the greatest benefit from exercise. Because our subjects had good glucose control before training, (overall pretraining mean A1c of 6.7%), we expected a more modest reduction in A1c. Although the effect of a specific reduction of A1c on health outcomes is not fully understood, the reduction in A1c from exercise in this study and in other studies3739 is similar to that of long-term drug or insulin therapy and diet (0.6%−0.8%).

Several potential mechanisms associated with improved glucose control following chronic exposure to exercise have been proposed. They include biochemical and structural adaptations of skeletal muscle and systemic influences on physical activity. Biochemical adaptations include an upregulation of mitochondrial proteins involved in respiration (citrate synthase),40 increased glycogen synthase activity,41 and increases in GLUT4 protein content.2,41 Structural adaptations from resistance training include increases in contractile protein content (hypertrophy), resulting in a higher basal metabolic rate13 and, therefore, potentially greater absolute glucose uptake.42 Endurance exercise results in increased mitochondrial proteins and improvements in the capillary to muscle fiber ratio, thereby increasing the distribution of substrates.43 Finally, regional adiposity, specifically visceral and intramuscular fat stores, is directly related to insulin insensitivity via fat-specific cytokine-mediated pathways, as well as a direct influence of intramyocellular fat storage on insulin receptor function within muscle tissue.44 For a full review on fat, see the article by Stehno-Bittel45 in this issue. Therefore, reduction in fat mass via exercise reduces the adverse influence of these factors. Whether any of these proposed mechanisms were responsible for the improved glucose control in our subjects is beyond the scope of the current project. We can speculate, however, that the improvements stemmed from more than the addition of lean tissue, as the AE group, which did not gain thigh lean tissue, also demonstrated significant reductions in A1c. The observed decrease in regional intramuscular fat stores, however, is an alluring and testable hypothesis for future study. See the article by Hilton et al46 in this issue for additional data on the relationship among intermuscular fat, muscle tissue, and functional limitations in people with diabetes and peripheral neuropathy.

Benefit of Combined Aerobic and Resistance Exercise in People With T2DM

Our results appear to indicate that aerobic exercise should be pursued by people with T2DM, but not in isolation. Aerobic exercise, long considered the exercise of choice for individuals with T2DM, when used in isolation, actually may promote lean tissue loss,47 an outcome demonstrated with our AE group. Individuals with T2DM have been reported to have less strength and muscle quality than age-matched control subjects who were healthy.48 This loss of protein reserve may predispose people with T2DM to muscle function impairments. Low muscle mass has been reported to be associated with poor lower-extremity function and mobility limitations in older individuals.49 Increasing lean tissue, particularly in the lower extremities, has been shown to improve functional mobility.15 Recent evidence also suggests a relationship between T2DM and fall risk, particularly in older women with diabetes.50 Although the mechanisms responsible for this association are not clear, the pathway from T2DM to a fall might be characterized as a multisystem reduction in an individual's functional reserve, which includes a loss of muscle mass and strength that is associated with, and perhaps accelerated in, people with T2DM.

This study is the first to utilize high-force eccentric training as the resistance training stimulus in an exercise and diabetes training study. The high-force–producing characteristic of eccentric exercise has previously been reported to result in an amplified muscle hypertrophic response in other patient populations.15,22 The hypertrophic and body composition improvements noted here, although not causally linked to improved glycemic control and metabolic outcomes in the present study, may be important because they have been shown previously to be related to improved glucose metabolism.51,52 Likewise, because aging individuals with T2DM can become progressively plagued by a diminishing exercise tolerance,53 due, in part, to the clinical complications of diabetes54 and fueled by further decreases in physical activity, eccentric exercise may be ideally suited as an exercise paradigm, particularly as an adjunct to aerobic exercise. In this study, the low energetic cost associated with eccentric exercise55,56 may have contributed favorably to the high adherence of the subjects to our diabetes exercise program. This may be especially important because people with T2DM often do not willingly participate in exercise.57

Study Limitations

There are several limitations to our study. First, the subjects were not randomly assigned to the different exercise groups. In order to minimize communication between the 2 groups in our diabetes exercise program, we chose to sequentially enroll subjects in each group. Second, the AE/RE group exercised for a longer period of time (up to 20 additional minutes) per exercise session compared with the AE group. Therefore, we cannot rule out that any additional benefits in the AE/RE group were simply due to the additional exercise time, regardless of the mode of exercise. Given that the AE group demonstrated lean tissue loss, it is unlikely that an additional 20 minutes of aerobic exercise would have resulted in any increase in lean thigh tissue CSA. Finally, the sample sizes in both groups were small, and any of the nonsignificant findings may have been due to a lack of statistical power.


There is a high prevalence of T2DM in physical therapist practice settings (unpublished data). One of the most frequently used interventions in physical therapy is therapeutic exercise. Exercise is clearly beneficial for individuals with T2DM, although very often this group does not willingly participate in exercise. People with T2DM often demonstrate low exercise tolerance and decreased physical activity.53 This often results in increased BMI and an increased total and regional storage of fat, along with decreases in lean tissue. Exercise can mitigate these detrimental body composition changes. Aerobic exercise should be pursued by people with T2DM, but not in isolation. Resistance exercise, by definition, is less aerobically challenging than aerobic exercise, and it can increase lean tissue CSA and benefit the individual with T2DM in ways that aerobic exercise alone cannot. Utilizing eccentric resistance exercise may be ideally suited to maximize lean tissue outcomes, at a fraction of the cardiovascular cost of concentric and isometric resistance exercise. Based on these results, we recommend that therapeutic exercise for people with T2DM include both aerobic and resistance components and that eccentric resistance exercises should be included.


  • Dr Marcus, Mr Smith, Dr Morrell, and Dr LaStayo provided concept/idea/research design. Dr Marcus, Mr Smith, Dr Morrell, Dr Dibble, Ms Wahoff-Stice, and Dr LaStayo provided writing. Dr Marcus, Mr Smith, Dr Morrell, and Dr Addison provided data collection. Dr Marcus, Mr Smith, and Dr Dibble provided data analysis. Dr Marcus and Mr Smith provided project management. Dr Marcus, Mr Smith, and Dr LaStayo provided fund procurement. Mr Smith, Ms Wahoff-Stice, and Dr LaStayo provided participants. Dr Morrell provided facilities/equipment and institutional liaisons. Mr Smith and Dr LaStayo provided clerical support. Mr Smith, Dr Morrell, Dr Dibble, and Dr LaStayo provided consultation (including review of manuscript before submission).

  • Dr LaStayo has served as an ad hoc, nonpaid consultant for the company (Eccentron LLC) that developed the commercial eccentric stepper device used as a resistance exercise device in this study. Dr LaStayo, in conjunction with the Arizona Board of Regents, holds a patent (United States: patent #7083547) related to the methods and an apparatus for a torque-controlled eccentric exercise training device. Dr LaStayo does not have a financial interest in the company, nor has he received any financial incentives from the company or from any results stemming from this or any other eccentric-related research.

  • This study was approved by the University of Utah Institutional Review Board.

  • This research was supported by the Utah Building Interdisciplinary Research Careers in Women's Health Program (National Institutes of Health grant 5K12HD043449-04) to Dr Marcus and a University of Utah Center for Rehabilitation Research grant to Dr Marcus and Ms Smith.

  • * Eccentron LLC, 570 Detroit St, Denver, CO 80206.

  • Siemens Corp, 153 E 53rd St, 56th Floor, New York, NY 10022.

  • Scion Corp, 82 Worman's Mill Ct, Suite H, Frederick, MD 21701.

  • § SPSS Inc, 233 S Wacker Dr, Chicago, IL 60606.

  • Received April 24, 2008.
  • Accepted July 17, 2008.


View Abstract