Treadmill Exercise and Resistance Training in Patients With Peripheral Arterial Disease With and Without Intermittent Claudication:  A Randomized Controlled Trial FREE

Lower extremity peripheral arterial disease (PAD) affects 1 in 16 US adults 40 years or older.¹ Men and women with PAD have greater functional impairment and more rapid rates of functional decline than those who do not have PAD.^2- 4

Supervised treadmill exercise improves treadmill walking performance in people with PAD who have symptoms of intermittent claudication.⁵ However, to what extent such training improves the physical capabilities of those with PAD without symptoms of intermittent claudication remains unanswered. Most patients with PAD do not have classic symptoms of intermittent claudication, either because they are asymptomatic or because they have exertional leg symptoms other than intermittent claudication.^2,6- 9 Individuals with PAD who do not have classic intermittent claudication symptoms have comparable or greater functional impairment and functional decline than those without PAD.^2- 3,6,9 However, no prior exercise interventions have been tested on PAD participants with and without classic symptoms of intermittent claudication.

Additionally, benefits of lower extremity resistance (strength) training in PAD are unclear. Adults with PAD have smaller calf muscle area and poorer leg strength than those without PAD, and these muscle characteristics are associated with greater functional impairment among people with PAD.^10- 11 However, clinical trials of lower extremity resistance training in persons with PAD have been small, have yielded mixed results, and have excluded PAD participants without classic symptoms of intermittent claudication.^12- 13

We performed a randomized controlled clinical trial involving men and women with PAD to address 2 clinical questions. First, we determined whether supervised treadmill exercise improves functional performance and other outcomes among participants with PAD with and without classic intermittent claudication symptoms. Second, we determined whether lower extremity-resistance training improves functional performance and other outcomes in participants with PAD with and without classic intermittent claudication symptoms. Each intervention group was compared with a control group.

The institutional review boards of the participating hospitals and medical centers approved the protocol. Participants provided written informed consent.

In this randomized controlled clinical trial, participants were assigned to 1 of 3 groups: a supervised treadmill exercise, supervised lower extremity resistance training, or control group. The resistance training intervention included only lower extremity exercises. Data collection and study interventions were performed at Northwestern University Feinberg School of Medicine between April 1, 2004, and August 19, 2008.

Eighty-five participants were recruited from newspaper and radio advertisements and 37 from among consecutive patients diagnosed with PAD in the noninvasive vascular laboratories and relevant clinics at Northwestern Memorial Hospital and other Chicago-area hospitals. Previous studies demonstrate that most patients diagnosed with PAD in noninvasive vascular laboratories do not have classic intermittent claudication symptoms.^3,6 Remaining participants were recruited from mailings to Chicago residents 60 years and older, posted flyers, recruitment mailings to people identified with PAD in the Lifeline of Screening program, and community outreach methods.¹⁴

The inclusion criterion was an ankle brachial index (ABI) of 0.95 or lower.^15- 17 Exclusion criteria were dementia, critical limb ischemia, foot ulcers, major amputation, nursing home residence, inability to walk on a treadmill, inability to exercise at the medical center 3 times weekly, failure to complete exercise run-in sessions, major surgery or a myocardial infarction within the past 3 months, major surgery planned in the next year, current participation in other clinical trials, baseline exercise comparable with that offered in either exercise group, abnormal baseline exercise stress test, walking limitation from a cause other than PAD, poorly controlled hypertension, and a baseline short physical performance battery score of 12 (ie, maximal possible score).

A handheld Doppler probe (Nicolet Vascular Pocket Dop II; Nicolet Biomedical Inc, Golden, Colorado) was used to obtain systolic pressures in the right and left brachial, dorsalis pedis, and posterior tibial arteries.^2- 3,18 Each pressure was measured twice. The ABI was calculated by dividing the mean of the dorsalis pedis and posterior tibial pressures in each leg by the mean of the 4 brachial pressures.¹⁸ Average pressures in the arm with the highest pressure were used when 1 brachial pressure was higher than the opposite brachial pressure in both measurement sets and the 2 brachial pressures differed by 10 mm Hg or higher in 1 measurement set.¹⁹

Medical history, race, and demographics were obtained using patient report.²⁰ Race data were obtained to determine whether black participants were equally distributed across the 3 study groups. Prior study indicates that among persons with PAD, blacks have poorer functional performance than whites.²¹

Leg symptoms were characterized using the San Diego claudication questionnaire.²² Intermittent claudication was defined as exertional calf pain that does not begin at rest, causes the participant to stop walking, and resolves within 10 minutes of rest.²¹ Participants without intermittent claudication were either asymptomatic (ie, no exertional leg symptoms) or had exertional leg symptoms not consistent with intermittent claudication.²²

Outcomes were measured before randomization and at the 6-month follow-up. Examiners were blinded to participant group assignment. The 2 prespecified primary outcomes were the 6-minute walk test and the short physical performance battery. The 6-minute walk was a primary outcome because persons with PAD are primarily limited in walking endurance.^2- 3,6 Treadmill walking performance was not a primary outcome because treadmill performance is susceptible to a learning effect²³ and participants randomized to treadmill exercise practiced treadmill walking during exercise sessions. Furthermore, corridor walking such as that measured in the 6-minute walk may better represent community walking performance than treadmill walking.^19,24- 26 The short physical performance battery is a functional performance measure that depends on leg strength and balance,^27- 28 which were targets of the lower extremity resistance training intervention. Lower short physical performance battery scores are associated with increased mobility loss among persons with PAD.⁴

Secondary outcomes were brachial artery flow mediated dilation and physical activity. Exploratory outcomes were treadmill walking performance, quality of life, and leg strength.

Six-Minute Walk. Following a standardized protocol,^2- 4,6,29 participants walk up and down a 100-foot hallway for 6 minutes after instructions to cover as much distance as possible. The distance completed after 6 minutes was recorded. The intraclass correlation coefficient for test retest reliability of the 6-minute walk was 0.90 (P < .001) among 155 PAD participants in our laboratory who completed 2 tests 1 to 2 weeks apart.

Short Physical Performance Battery. The short physical performance battery combines data from usual paced 4-m walking velocity, time to rise and stand from a seated position 5 times, and standing balance.^4,27- 28 Individuals receive a score of 0 for each task they are unable to complete. Scores of 1 to 4 are assigned for remaining tasks, according to established methods.^4,27- 28 Scores are summed to obtain the short physical performance battery, ranging from 0 to 12.^22- 23 Test retest reliability of the short physical performance battery was 0.77 (P < .001) among 151 PAD participants in our laboratory who completed 2 tests 1 to 2 weeks apart.

Repeated Chair Rises. Participants sit in a straight-backed chair with arms folded across their chest and stand 5 times consecutively as quickly as possible. Time to complete 5 chair rises is measured.^27- 28

Standing Balance. Participants are asked to hold 3 increasingly difficult standing positions for 10 seconds each: the side-by-side stand, semi-tandem stand (standing with feet parallel and the heel of 1 foot touching the base of the first toe of the opposite foot), and the full-tandem stand (standing with 1 foot directly in front of the other).^27- 28 Scores range from 0 (unable to hold the side-by-side stand for 10 seconds) to 4 (able to hold the full-tandem stand for 10 seconds).^27- 28

Four-Meter Walking Velocity. Walking velocity was measured with a 4-meter walk performed at usual pace, according to previously described methods.^{2- 4,22- 23} To account for possible learning effects, we measured the short physical performance battery and 6-minute walk at 2 separate baseline visits, approximately 1 to 2 weeks apart. Performance at the second visit was used as the baseline value. However, 5 participants had the short physical performance battery measured at only 1 baseline visit and their first (only) score was used as their baseline value.

Brachial Artery Flow-Mediated Dilation. Brachial artery flow-mediated dilation was measured after a 12-hour fast by trained registered diagnostic cardiac sonographers, using standard procedures.^30- 31 Participants were instructed to hold medications and not exercise or smoke prior to testing. Participants whose brachial systolic pressure differed by more than 15 mm Hg between the right and left arms, those with a radical mastectomy, and those with a history of Raynaud phenomenon were excluded from flow-mediated dilation testing. The proximal brachial artery was imaged (B-mode and Doppler) using a linear array vascular ultrasound transducer (Sequoia Model 256, Siemens Medical Solutions, Hoffman Estates, Illinois; frequency, 8 MHz; range, 5-8 MHz). A blood pressure cuff proximal to the visualized brachial artery segment was inflated for 4 minutes at 50 mm Hg above systolic pressure. Longitudinal images of the brachial artery and Doppler blood flow were obtained 60 seconds after cuff deflation.³⁰ Images were interpreted by a single reader at the University of Wisconsin Atherosclerosis Imaging Research Program Core Laboratory, using established standards.³¹ Measurement reproducibility in the core laboratory has a median flow-mediated dilation difference of 0.02% (interquartile range, −0.03 to 0.04) on blinded repeated readings.

Physical Activity. Habitual physical activity was measured objectively over 7 days using a vertical accelerometer (Caltrac, Muscle Dynamics Fitness Network, Inc, Torrance, California) according to established methods, yielding activity units.^32- 34

Treadmill Walking Performance. Maximal treadmill walking time and time to onset of ischemic leg symptoms were measured using the Gardner-Skinner protocol.^23,35 This protocol has a coefficient of variation of approximately 12% in persons with PAD.^23,35

Quality of Life. The Walking Impairment Questionnaire is a PAD-specific measure of self-reported walking limitations with 3 domains: walking distance, walking speed, and stair climbing.³⁶ Each domain is scored on a 0 to 100 scale, for which 0 represents the most extreme limitation and 100 represents no difficulty walking long distances, walking rapidly, or climbing 3 stair flights, respectively.³⁶ We used the Medical Outcomes Study Short-Form 36 (SF-36) to assess functional status in the physical functioning domain.³⁷

Strength Measures. Maximum isometric knee extension and plantar flexion strength were measured in newtons over 6 seconds with a computerized strength chair (Good Strength Chair, Metitur Oy, Jyvasklya, Finland).^11,38 Two trials were performed, and the maximum was used in analyses.¹¹ The Good Strength Chair has high test retest reliability (Pearson product moment correlations, 0.88-0.96).³⁸ Knee extension power was measured in watts using the Nottingham power rig.^11,39

Other Measures. Height and weight were measured at baseline. Body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared.²

After baseline testing, eligible participants were randomized by computer using a randomly permuted block method. Randomization was stratified by the presence vs absence of intermittent claudication.

The supervised treadmill exercise intervention consisted of treadmill exercise 3 times a week for 24 weeks, supervised by an exercise physiologist. Participants began with 15 minutes of exercise and increased to 40 minutes by week 8. Initial treadmill walking speed was 2.0 mph per hour or lower if the participant was unable to keep that pace. Between weeks 8 and 24, attempts to increase exercise intensity were made at least weekly either by increasing treadmill speed or by increasing the treadmill grade. Participants with leg symptoms were encouraged to exercise to near maximal leg symptoms. Asymptomatic participants were encouraged to exercise to a level of 12 to 14 (moderately hard) on the Borg rating of perceived exertion scale.⁴⁰

Supervised Lower Extremity Resistance Training. Participants in the lower extremity resistance training group exercised 3 times a week for 24 weeks with a certified trainer. Participants performed 3 sets of 8 repetitions of knee extension, leg press, and leg curl exercises using standard equipment. For each exercise, the 1 repetition maximum (1 rm) was measured at baseline and subsequently every 4 weeks. Participants began exercising at 50% of their 1 rm. Weights were increased over the first 5 weeks until participants were lifting 80% of their 1 rm. Weights were adjusted after each monthly 1 rm and as needed to achieve an exercise intensity of a rating of perceived exertion of 12 to 14. Participants also performed 3 sets of 8 repetitions of squat and toe rise exercises. The toe rise exercises were plantar extension exercises, in which participants assumed a tip-toe position and lowered themselves 8 times consecutively.

Control. Those in the control group participated in 11 nutritional information sessions over 6 months. The 1-hour sessions were intended to provide regular contact with participants in the control group. Session topics included nutritional supplements, healthful restaurant eating, and increasing fruit and vegetable consumption. Sessions were led by registered dieticians but were not designed to change behavior.

Power calculations assumed that 50 participants in each group would complete 6-month follow-up and that 2 separate 2-sample t tests using a 2-sided α of .05 would be conducted. The study was designed to have 80% power to detect a difference of 30 m change in the 6-minute walk distance and a difference of 0.97 change in the short physical performance battery between baseline and 6-month follow-up between each exercise and control group. These differences represent clinically meaningful change in study outcomes.⁴¹ Power analyses assumed pooled standard deviations of 54 m and 1.74 units, respectively. Because the number of participants who actually completed 6-month follow-up testing was slightly lower, power was reduced to 0.76 to 0.79 for comparisons of each exercise group with the control group.

χ² Tests and 1-way analyses of variance were used to compare characteristics of participants across the 3 groups at baseline. Two sample, 2-sided t tests were used to compare changes in outcomes between baseline and 6-month follow-up between each exercise group and the control group, respectively, without adjustments for multiple comparisons. A priori, the P value considered statistically significant was P < .05. Because of skewed distributions for quality-of-life measures and brachial artery flow-mediated dilation, differences in median values for these outcomes were compared using Kruskal-Wallis analysis of variance and Wilcoxon rank sum tests, with Hodges-Lehman 95% confidence intervals (CIs) computed for between-group differences.⁴²

Intention-to-treat analyses were performed.⁴³ Analyses were repeated using multiple imputation for persons who died or dropped out before completing 6-month follow-up testing.^44- 45 The multiple imputation for missing 6-month data was performed using SAS Proc MI (SAS Institute Inc, Cary, North Carolina) to obtain 5 imputed data sets. Results were combined using Proc Combine. Variables used to impute data sets included age, ABI, BMI, sex, race, smoking status, baseline outcome values, leg symptoms, and comorbidities. Imputations were performed with and without 4 participants: 2 who died, 1 diagnosed with end-stage renal disease, and 1 diagnosed with lung cancer. Analyses were performed using SAS computer software version 9.2.

Of 1009 potential participants who scheduled a baseline visit, 263 failed to show for their appointment and 261 had an ABI higher than 0.95. Of the remaining 485, there were 329 who met an exclusion criterion, leaving 156 eligible participants (Figure). Among those who scheduled a baseline visit, the average age of randomized vs nonrandomized participants was 73.73 vs 70.62 (P = .002). Randomized vs nonrandomized participants included 52% vs 46% women (P = .17) and 39.7% vs 36.0% blacks (P = .37). There were no differences in baseline characteristics across the 3 groups (Table 1).

Among participants in the treadmill exercise, resistance training, and control groups, median session attendance rates were 85%, 92%, and 100%, and 6-month follow-up data were obtained in 98%, 89%, and 91%, respectively. Two participants (1 in the resistance training group and 1 in the control group) died from cancer before the 6-month follow-up.

For the 6-minute walk, those in the supervised treadmill exercise group increased their distance from baseline by a mean of 20.9 m vs those in the control group whose distance decreased from baseline by a mean of −15.0 m, for a mean increase of 35.9 m between groups (95% CI, 15.32-56.54 m; P < .001). Participants in the lower extremity resistance training group did not experience change in their 6-minute walk performance compared with the control group (12.4 m, 95% CI, −8.42 to 33.25 m; P = .24)(Table 2).

There were no differences in change in short physical performance battery score between the treadmill exercise and control groups or between the resistance training and the control groups at the 6-month follow-up (Table 2). Results for primary outcomes were not substantially changed when analyses were repeated using multiple imputation methods with and without the 4 participants with extreme outcomes (death or serious illness). Similarly, results were not substantially changed when analyses were repeated substituting baseline values for the 6-month follow-up time point for participants who did not return for follow-up, with and without exclusion of the 4 participants with extreme outcomes. Findings would be unchanged with multiple comparison adjustment (P < .025).

One hundred-one participants met inclusion criteria for brachial artery flow-mediated dilation testing and adhered to requirements for testing. At the 6-month follow-up, participants in the treadmill exercise group had more favorable changes in brachial arterial flow-mediated dilation than the control group (1.53%; 95% CI, 0.35%-2.70%; P = .02 and 0.06 mm; 95% CI, 0.00-0.11 mm; P = .04). Changes in brachial artery flow-mediated dilation among participants in the resistance training group were not different from the control group (Table 2). At the 6-month follow-up, there were no differences in changes in accelerometer-measured physical activity in either exercise group compared with the control group (Table 2).

Participants in the treadmill exercise and in the resistance training groups each had significantly greater increases in mean maximum treadmill walking time at the 6-month follow-up, than the control group (3.44 minutes, 95% CI, 2.05-4.84 minutes; P < .001 for the treadmill group and 1.90 minutes, 95% CI, 0.49-3.31 minutes, P = .009 for the resistance training group; Table 3). Participants in the treadmill exercise group had greater mean increases in treadmill time to onset of ischemic leg symptoms than the control group (1.65 minutes; 95% CI, 0.34-2.95; P = .01; Table 3). Similarly, the treadmill exercise group had significantly greater mean improvement in their SF-36 physical functioning score (7.50; 95% CI, 0.00-15.0; P = .02) and in their walking impairment distance score (10.7; 95% CI, 1.56-19.9; P = .02) than the control group (Table 3). The resistance training group had greater mean improvement in their SF-36 physical functioning score (7.50; 95% CI, 0.00-15.0; P = .04) and in their walking impairment distance (6.92; 95% CI, 1.07-12.8; P = .02) and stair climbing scores (10.4; 95% CI, 0.00-20.8; P = .03) than the control group (Table 3).

Participants in the treadmill group did not have greater improvement in any leg strength measure at 6 months than the control group (Table 3). The resistance training group improved isometric knee extension strength compared with the control group (mean, 80.2 N; 95% CI, 36.8-124 N, P < .001; Table 3).

Associations of treadmill exercise with greater absolute change in brachial artery flow-mediated dilation and associations of lower extremity resistance training with improved SF-36 physical functioning score compared with the control group would not be statistically significant after multiple comparison adjustment at P < .025.

The study was not designed to have statistical power for comparing differences in outcomes between the 2 exercise groups. However, the treadmill exercise group had greater increases in the 6-minute walk distance (23.5 m; 95% CI, 2.8-44.2 m; P = .03) and in the maximum treadmill walking time (1.54 minutes; 95% CI, 0.17-2.92 minutes; P = .03) compared with the resistance group. Participants in the resistance group experienced greater increases in knee extension isometric strength (83.5 N; 95% CI, 39.4-127.6 N; P <.001) and in knee extension power (14.4 W; 95% CI, 1.43-27.4 W; P = .03) compared with the treadmill exercise group.

The study also was not designed to have statistical power for comparing differences in outcomes between different leg symptom groups. The magnitude of change for our primary outcomes was reasonably similar between participants with vs without intermittent claudication and between asymptomatic participants vs symptomatic participants (data not shown).

Several adverse events occurred related to study participation. One participant had a cardiac arrest during treadmill exercise, 4 months after randomization. The participant was successfully resuscitated and hospitalized. The patient did not have a myocardial infarction but was found to have nonobstructive coronary disease and did not have an inducible arrhythmia on electrophysiological testing. He was discharged home on medical therapy but experienced persistent memory loss. Another participant randomized to the treadmill exercise group developed chest pain during an exercise session. A subsequent coronary catheterization revealed coronary atherosclerosis without a flow-limiting lesion. The participant was treated medically and returned to exercise. A third participant fell during the follow-up 6-minute walk test and fractured her arm. A fourth participant with a history of prior ankle injury developed severe ankle pain after a run-in resistance training session. The participant could not exercise for several weeks, until the ankle pain resolved. He was ultimately randomized to treadmill exercise and completed the trial uneventfully.

This randomized controlled clinical trial demonstrates that a 6-month supervised treadmill exercise intervention increases walking endurance, measured by the 6-minute walk and treadmill walking performance, in patients with PAD both with and without classic intermittent claudication symptoms. Supervised treadmill exercise also increased brachial arterial flow-mediated dilation and improved quality of life. A 6-month lower extremity resistance training intervention did not improve 6-minute walk distance in PAD participants. However, resistance training improved maximal treadmill walking time and quality-of-life measures, particularly stair climbing ability. Supervised treadmill exercise was associated with greater increases in 6-minute walk performance than the resistance-training group.

To our knowledge, this is the first randomized controlled clinical trial of exercise in PAD to include participants with and without classic symptoms of intermittent claudication. Most patients with PAD do not have classic intermittent claudication symptoms.^6- 9 Results reported herein indicate that clinicians should recommend supervised treadmill exercise to PAD patients, with or without classic symptoms of intermittent claudication. To our knowledge, this is also the first randomized controlled clinical trial to demonstrate that supervised treadmill exercise improves brachial artery flow-mediated dilation in persons with PAD. Patients with PAD typically have severe systemic atherosclerosis and poorer endothelial function than patients without PAD.⁴⁶ Among persons with PAD, poorer brachial arterial flow-mediated dilation is associated with higher cardiovascular event rates.^47- 48 Our findings suggest that supervised treadmill exercise confers a favorable systemic vascular effect that may reduce cardiovascular events in persons with PAD.

Reasons for lack of improvement in the short physical performance battery in either exercise group are unclear. Balance and leg strength are important determinants of the short physical performance battery score. Our results suggest that treadmill exercise does not improve balance and leg strength significantly. One potential explanation for the lack of improvement in the short physical performance battery in the resistance-training group is that baseline leg strength was relatively high in our cohort.⁴⁹ Resistance training may more effectively improve components of the short physical performance battery in persons with poorer baseline strength. The poorer test retest reliability of the short physical performance battery compared with the 6-minute walk may also make meaningful change in the short physical performance battery more difficult to achieve.

Despite their associations with improved walking performance, the exercise interventions did not increase accelerometer-measured physical activity during daily life. Interventions specifically targeted to physical activity behavior may be necessary to increase daily physical activity in persons with PAD.

This study has several limitations. First, our exclusion criterion of the maximum short physical performance battery score of 12 and the relatively large proportion of potential participants who were unwilling to attend on-site exercise sessions 3 times a week reduces the generalizability of our findings. Second, missing data at follow-up were more common in more frail participants. However, sensitivity analyses demonstrated that these missing data are not likely to have significantly altered our findings. Third, median session attendance rates varied from 85% for the treadmill exercise group to 100% for the control sessions. This difference may have influenced results. Fourth, because of additional exclusion criteria for flow-mediated dilation and patient refusal to adhere to test requirements, the sample size for the flow-mediated dilation measurement was lower than that for the entire cohort. Sample sizes for secondary and exploratory outcomes were smaller than for the primary outcomes. Finally, the sample size was not large enough to allow meaningful comparisons in outcomes between categories of leg symptoms.

Based on findings reported in this trial, physicians should recommend supervised treadmill exercise programs for PAD patients, regardless of whether they have classic symptoms of intermittent claudication. Our findings regarding brachial artery flow-mediated dilation suggest that supervised treadmill exercise improves global vascular health in patients with PAD. Lower extremity resistance training improves treadmill walking performance and quality of life, particularly stair climbing, in patients with PAD with and without intermittent claudication.