Duration of Supervised Exercise Necessary for Meaningful Improvement in Peripheral Artery Disease

Abstract

Background

Supervised treadmill exercise improves walking performance in people with lower extremity peripheral artery disease, but benefits are not immediate. This study identified the time course of attaining meaningful improvement in 6‐minute walk distance and patient‐reported outcome measures during a 6‐month supervised exercise intervention in people with peripheral artery disease.

Methods

Participants with peripheral artery disease were randomized to supervised treadmill exercise 3 time weekly or a nonexercise control group for 6 months. Six‐minute walk distance (large clinically important difference: 20 meters) and the Walking Impairment Questionnaire distance score (0–100 scale, 100 is best, clinically important difference: 5 points) were measured at the 6‐week, 12‐week, and 26‐week follow‐up using a mixed‐effects model for repeated measures.

Results

Of 210 randomized participants (mean age, 67.0±8.6 years, 82 [39%] women, 141 [66%] Black), 200 (95%) completed at least 1 follow‐up visit. Compared with controls, supervised exercise significantly improved 6‐minute walk distance by 13.0 m (P=0.049) at the 6‐week, 31.8 m (P<0.001) at the 12‐week, and 33.9 m (P<0.001) at the 26‐week follow‐up. Compared with controls, supervised exercise increased the Walking Impairment Questionnaire distance score by +2.63 (P=0.37) at the 6‐week, +6.59 (P=0.049) at the 12‐week, and +2.37 (P=0.49) at the 26‐week follow‐up.

Conclusions

In people with peripheral artery disease, >6 weeks of supervised treadmill exercise was necessary to attain a large meaningful gain in 6‐minute walk, and large meaningful gains were measurable by week 12 of supervised exercise. Meaningful improvement in participant reported walking ability was first observed at the 12‐week follow‐up, but this statistically significant benefit was gone by the 26‐week follow‐up.

Registration

URL: https://www.clinicaltrials.gov; Unique identifier: NCT01408901.

Nonstandard Abbreviations and Acronyms

GM‐CSF

granulocyte‐macrophage colony‐stimulating factor

PROPEL

Progenitor Cell Release Plus Exercise to Improve Functional Performance in Peripheral Artery Disease

WIQ

Walking Impairment Questionnaire

Clinical Perspective.

What Is New?

• In people with peripheral artery disease, >6 weeks of supervised treadmill exercise was necessary to attain a large meaningful improvement in 6‐minute walk distance.

• In people with peripheral artery disease, large meaningful gains in 6‐minute walk distance were measurable after 12 weeks of supervised treadmill exercise and did not further meaningfully increase after an additional 12 weeks of exercise at the 6‐month follow‐up; meaningful improvement in patient‐reported walking ability was first observed at the 12‐week follow‐up, but this statistically significant benefit was lost by the 6‐month follow‐up.

What Are the Clinical Implications?

• When counseling patients with peripheral artery disease, clinicians should advise that meaningful benefits in walking ability typically occur after >6 weeks of exercise. When counseling patients with peripheral artery disease, clinicians should advise that most of the benefits from supervised treadmill exercise are attained by the 12‐week follow‐up and can be maintained, but not meaningfully further increased, by an additional 12 weeks of supervised treadmill exercise.

Supervised exercise therapy is first‐line therapy for improving walking impairment in people with lower extremity peripheral artery disease (PAD). ¹ , ² The Centers for Medicare and Medicaid Services has covered supervised exercise training for patients with PAD since 2017. ² However, <1% of people with PAD who begin Medicare‐covered supervised exercise treatment complete all 36 supervised exercise sessions covered by Medicare. ³ Adhering to supervised exercise may be difficult for people with PAD, in part because supervised exercise does not immediately improve walking ability in PAD. Patients who begin an exercise program may become discouraged by lack of immediate improvement, potentially causing them to discontinue supervised exercise training before significant benefits from exercise are attained.

A clearer understanding of the number of weeks of supervised treadmill exercise necessary to meaningfully improve walking performance will help clinicians better educate patients about reasonable expectations on the time course of improvement in walking performance during supervised treadmill exercise participation. This study used data from the PROPEL (Progenitor Cell Release Plus Exercise to Improve Functional Performance in Peripheral Artery Disease) randomized clinical trial to evaluate the time course of improved walking performance in response to supervised treadmill exercise in patients with PAD. ⁴ PROPEL studied the effects of a 6‐month supervised treadmill exercise intervention on outcomes of 6‐minute walk distance, treadmill walking time, and patient‐reported outcome measures at the 6‐week, 12‐week, and 26‐week follow‐up. ⁴

METHODS

The institutional review board at Northwestern University and all recruitment sites approved the protocol. All participants provided written informed consent. This study was a post hoc analysis of data from the PROPEL randomized clinical trial. ⁴ The PROPEL clinical trial tested whether supervised treadmill exercise, with and without GM‐CSF (granulocyte macrophage colony stimulating factor), significantly improved 6‐minute walk distance, treadmill walking time, and patient‐reported outcome measures in people with PAD. Participants were randomized in Chicago, Illinois between January 6, 2012 and December 22, 2016, and final follow‐up testing occurred on August 15, 2017. ⁴ , ⁵ Participants with PAD were randomized to 1 of 4 groups for 6 months: supervised exercise + GM‐CSF (group 1), supervised exercise + placebo (group 2), health education control + GM‐CSF (group 3), and health education control + placebo (group 4). GM‐CSF had no significant effect on walking performance. ⁴ Therefore, these analyses combined the 2 groups randomized to supervised exercise (group 1 and group 2) and the 2 groups randomized to the health education control group (group 3 and group 4) for analyses. The study investigative team was diverse with regard to sex, specialty, age, and institution. ⁴ , ⁵ Data are available from the corresponding author upon reasonable request.

Participant Identification

Participants were identified through newspaper and radio advertisements or postcard mailings to people aged ≥55 years in Chicago, Illinois. People with PAD who previously participated in research with the principal investigator (M.M.M.) and expressed interest in future research studies were contacted. ⁴ , ⁵ Postcard mailings are highly effective at increasing diversity of study participants in people with PAD. ⁴ , ⁵

Inclusion Criteria

Inclusion criteria included an ankle‐brachial index (ABI) ≤0.90 at the baseline visit. Potential participants with an ABI >0.90 at the baseline visit were potentially eligible, if a hospital‐affiliated vascular laboratory provided further evidence of PAD. ⁴ , ⁵ , ⁶ , ⁷ Participants with an ABI between 0.90 and 1.00 at baseline and those with a normal ABI and prior lower extremity revascularization were eligible if their ABI dropped by 20% after a heel‐rise test.

Exclusion Criteria

Exclusion criteria were reported previously ⁴ , ⁵ and are summarized here. Exclusion criteria included previous below‐ or above‐knee amputation, major surgery or revascularization within the past 3 months or planned intervention within the next 6 months, foot ulcer or critical limb ischemia, wheelchair confinement, use of a walking aid other than a cane, walking impairment due to conditions other than PAD, inability or unwillingness to attend exercise sessions 3 times per week, or participation in current exercise level similar to the intervention. Potential participants who already planned to begin a new exercise program or who did not complete the study run‐in were excluded. For the run‐in, participants were asked to attend 1 weekly health education session and 1 treadmill exercise session over a 3‐week period.

Additional exclusion criteria consisted of current or recent participation in a clinical trial or cardiac rehabilitation, Parkinson Disease, oxygen requirement with activity, significant visual or hearing impairment, or Mini‐Mental Status Examination score <23. ⁸ Potential participants for whom exercise might have been unsafe, such as those with an increase in angina pectoris during the prior 6 months or an abnormal baseline exercise stress test were excluded. Participants treated for cancer in the past 2 years were excluded unless their cancer was early‐stage and their prognosis was excellent. Participants with greater than New York Heart Association class II heart failure or angina were excluded.

Randomization

Participants were randomized to 1 of 4 groups using an SAS computer program: GM‐CSF + supervised exercise, GM‐CSF + health education control (GM‐CSF only), placebo + supervised exercise (exercise only), or placebo + health education control. Randomization was stratified by diabetes. Block randomization was used, with block sizes randomly selected from 8 and 12.

Interventions

Supervised Treadmill Exercise

Supervised treadmill exercise was provided 3 times weekly with an exercise physiologist. Walking exercise duration was increased gradually, as tolerated by participants, until 50 minutes of exercise per session was achieved. ⁴ , ⁵ Walking exercise was individualized according to the ability of each participant. Participants were asked to walk to induce maximal symptoms of leg tightness, cramping, weakness, pain, or other ischemic symptoms. Treadmill speed and grade were adjusted periodically so that participants experienced ischemic symptoms within 8 to 10 minutes of each bout of exercise. When symptoms became severe, participants stopped to rest and resumed walking exercise after symptoms subsided.

Health Education Control

Participants randomized to the health education control group attended weekly 1‐hour educational sessions on topics of interest to the typical patient with PAD, including cancer screening, immunizations, and hypertension.

GM‐CSF and Placebo

GM‐CSF (250 μg/m² per day) or placebo was administered subcutaneously 3 times weekly for 2 weeks in a double‐blinded fashion.

Outcomes

Outcome data were collected by individuals blinded to group assignment. All outcomes were measured before randomization, and at the 6‐week, 12‐week, and 26‐week follow‐up.

Six‐Minute Walk Test

The 6‐minute walk test is a well‐validated measure of walking endurance for people with PAD. ⁹ , ¹⁰ Following a standardized protocol, participants walked up and down a 100‐foot hallway for 6 minutes with instructions to cover the longest possible distance. The distance completed after 6 minutes was recorded. A large meaningful change has been defined as 20 meters. ¹¹ , ¹²

Treadmill Walking Performance

Maximal treadmill walking distance was measured using the Gardner‐Skinner protocol. Maximum treadmill walking time and the time to onset of ischemic leg symptoms were recorded. ¹³ A moderate clinically important difference for improvement in maximal treadmill walking time is 2.35 minutes, and a moderate clinically important difference for improvement in pain‐free treadmill walking time is 0.53 minutes. ¹¹

Walking Impairment Questionnaire

The Walking Impairment Questionnaire (WIQ) is a measure of self‐reported walking impairment specific for patients with PAD (score range, 0–100; 100 indicates the best score). ¹⁴ The WIQ distance score measures limitations in walking distances ranging from walking across a small room to up to 1500 feet. The WIQ speed score measures patient‐reported difficulty walking varying speeds over 1 block, ranging from slow to jogging. The WIQ stair‐climbing score measures difficulty climbing 1, 2, and 3 flights of stairs. A small clinically important difference for the WIQ distance score is 5.0. ¹¹

Additional Measures

Ankle‐Brachial Index

A handheld Doppler probe (Nicolet Vascular Pocket Dop II, Golden, CO) was used to measure systolic blood pressures after the participant rested supine for 5 minutes. ⁴ , ⁵ , ¹⁰ Pressure was measured twice in the following arteries: right brachial, dorsalis pedis, and posterior tibial arteries, and left dorsalis pedis, posterior tibial, and brachial arteries. ⁴ , ⁵ , ¹⁰ The ABI was calculated by dividing average pressures in each leg by the average of the 4 brachial pressures. ⁴ , ⁵ , ¹⁰ , ¹⁵ Among participants with baseline resting ABI of 0.90 to 1.00, heel‐rise testing, consisting of 50 heel rises at a rate of 1 per second followed by immediately repeating the ABI, was performed. ⁷ Potential participants who had an ABI drop of 20% or more after heel‐rise testing were included. ⁷

Medical History, Demographics, Body Mass Index

Medical history and demographics were obtained using the patient report. Height and weight were measured at baseline. Body mass index was calculated as weight (kilograms)/height (meters squared).

Leg Symptoms

Leg symptoms were characterized using the San Diego Claudication Questionnaire. ¹⁶ , ¹⁷ Intermittent claudication was defined as exertional calf pain that did not begin at rest, caused the participant to stop walking, and resolved within 10 minutes of rest. Participants without intermittent claudication were classified as either asymptomatic (ie, reported no exertional leg symptoms) or with exertional leg symptoms that did not meet criteria for intermittent claudication.

Power Considerations

Power considerations for the primary and secondary outcomes were reported previously. ⁴ Power calculations were not performed for these post hoc exploratory analyses.

Statistical Analysis

Because GM‐CSF did not significantly affect study outcomes, compared with placebo, ⁴ the 2 supervised exercise groups (exercise + placebo and GM‐CSF + exercise) and the 2 exercise control groups (placebo + health education control and GM‐CSF + health education control) were combined in the analyses to compare supervised treadmill exercise and the control group at each time point. Participants' data were included in analyses regardless of their adherence to study interventions, according to the intent‐to‐treat principle. Baseline characteristics of participants were compared between the exercise and health education control groups. Mixed‐effects models for repeated measures regression were used to compare temporal changes in each outcome between the exercise and control groups. Changes in outcome measures (6‐minute walk distance, maximal treadmill walking time, pain‐free treadmill waking time, and the WIQ distance score) from baseline to the 6‐week, 12‐week, and 26‐week follow‐up were treated as longitudinal outcomes, with an unstructured correlation matrix in the mixed‐effects models for repeated measures regression. The treatment indicator and their interaction with visit time (as categorical variables) were independent variables of the primary interest. The mixed‐effects models for repeated measures regression also adjusted for the corresponding baseline measure for each outcome and a GM‐CSF assignment indicator. The regression coefficients representing the between‐group difference at the 6‐week, 12‐week, and 26‐week follow‐up were estimated and the corresponding 95% CIs constructed. Inferences based on mixed‐effects models for repeated measures regression are valid under the missing at random assumption. Simple complete case analyses were performed as sensitivity analysis. Subset analyses and statistical tests for interaction were performed for sex, race, diabetes, presence versus absence of college education, and presence versus absence of Medicaid insurance and the effects of supervised treadmill exercise on 6‐minute walk distance. P values <0.05 were considered statistically significant. The statistical analyses were conducted using SAS version 9.4.

RESULTS

There were 210 people who met the inclusion criteria and were randomized (Figure 1). Of these, 200 (95.2%) completed at least 1 follow‐up visit and were included in the analyses. Of the 200 participants included in the analyses, the mean age was 67.1±8.5 years, 40% were women, 66% were Black, and 38% had diabetes (Table 1).

Figure 1. Number of participants and follow‐up rates for the PROPEL clinical trial of supervised exercise for PAD.

GM‐CSF indicates granulocyte macrophage colony stimulating factor; PAD, peripheral artery disease; and PROPEL, Progenitor Cell Release Plus Exercise to Improve Functional Performance in Peripheral Artery Disease.

Table 1.

Baseline Characteristics of Participants With PAD (N=200)

Baseline characteristics	Overall	Exercise group	Control group
	N=200	n=100	n=100
Age, y, mean±SD	67.1±8.5	67.3±9.1	66.9±7.8
Ankle‐brachial index, mean±SD	0.70±0.19	0.69±0.20	0.70±0.18
Body mass index, mean±SD	30.7±6.6	31.0±6.4	30.4±6.8
Women, n (%)	79 (39.5)	43 (43.0)	36 (36.0)
Black, n (%)	132 (66.0)	63 (63.0)	69 (69.0)
Current smoker, n (%)	66 (33.0)	37 (37.0)	29 (29.0)
Intermittent claudication, n (%)	64 (32.0)	37 (37.0)	27 (27.0)
Leg pain, not intermittent claudication, n (%)	127 (63.5)	58 (58.0)	69 (69.0)
No exertional leg symptoms, n (%)	9 (4.5)	5 (5.0)	4 (4.0)
Diabetes, n (%)	76 (38.0)	38 (38.0)	38 (38.0)
Myocardial infarction, n (%)	45 (22.5)	20 (20.0)	25 (25.0)
Hypertension, n (%)	168 (84.0)	85 (85.0)	83 (83.0)
Cancer, n (%)	36 (18.0)	17 (17.0)	19 (19.0)
Stroke, n (%)	30 (15.0)	14 (14.0)	16 (16.0)
Heart failure, n (%)	25 (12.5)	10 (10.0)	15 (15.0)
Angina, n (%)	40 (20.0)	22 (22.0)	18 (18.0)
Six‐minute walk distance, m, mean±SD	340.0±98.4	337.9±101.3	342.1±95.8
Total treadmill time, min, mean±SD	7.58±4.61	7.13±4.97	8.04±4.19
Treadmill time at onset of leg symptom, min, mean±SD	3.49±3.05	3.96±3.87	3.91±3.36
WIQ distance score, 1–100 (100 is best), mean±SD	33.4±27.0	34.0±27.7	32.7±26.3
WIQ speed score, 1–100 (100 is best), mean±SD	36.0±24.8	36.9±24.9	35.0±24.8
WIQ stair‐climbing score, 1–100 (100 is best), mean±SD	49.4±28.9	52.1±28.9	46.7±28.8

PAD indicates peripheral artery disease; and WIQ, Walking Impairment Questionnaire.

Six‐Minute Walk Distance

Compared with controls, supervised treadmill exercise significantly improved the 6‐minute walk distance by 13.0 m (95% CI, +0.04–+25.9; P=0.049) at the 6‐week follow‐up, 31.8 m (95% CI, +15.7–+47.9; P < 0.001) at the 12‐week follow‐up, and 33.9 m (95% CI, +16.4–+51.3; P < 0.001) at the 26‐week follow‐up (Table 2, Figure 2). The 13.0‐m improvement at the 6‐week follow‐up represented 38.3% of the 33.9‐m improvement observed at the 26‐week follow‐up, whereas the 31.8‐m improvement at the 12‐week follow‐up represented 93.8% of the improvement at the 26‐week follow‐up. The effects of supervised exercise on changes in the 6‐minute walk distance did not significantly differ by sex, race, presence versus absence of diabetes, presence or absence of college education, or presence or absence of Medicaid insurance (Table S1).

Table 2.

Temporal Changes in Walking Outcomes in Response to Supervised Exercise in Peripheral Artery Disease

Group	No.	Baseline, mean±SD	Within‐group difference at 6‐wk follow‐up from baseline, least‐squares mean±SD	Between‐group difference at 6‐wk change, least‐squares mean (95% CI)	Within‐group difference at 12‐wk follow‐up from baseline, least‐squares mean (SE)	Between‐group difference at 12‐wk change, least‐squares mean (95% CI)	Within‐group difference at 6‐mo follow‐up, least‐squares mean (SE)	Between‐group difference at 6‐mo change, least‐squares mean (95% CI)
Six‐minute walk distance (m)
Exercise	100	337.9±101.3	15.3±4.6	13.0 (0.04 to 25.9) P=0.0494	23.7 (5.7)	31.8 (15.7 to 47.9) P=0.0001	27.5 (6.2)	33.9 (16.4 to 51.3) P=0.0002
Control	100	342.1±95.8	2.4±4.7		−8.1 (5.8)		−6.4 (6.3)
Maximal treadmill walking time (min)
Exercise	100	7.12±4.97	2.84±0.29	2.47 (1.64 to 3.29) P < 0.0001	3.83 (0.34)	3.64 (2.68 to 4.60) P < 0.0001	4.37 (0.37)	3.73 (2.69 to 4.77) P < 0.0001
Control	99	7.99±4.19	0.37±0.30		0.19 (0.34)		0.64 (0.38)
Pain‐free treadmill walking time (min)
Exercise	100	3.96±3.87	1.40±0.25	0.95 (0.24 to 1.65) P=0.0087	2.34 (0.30)	1.54 (0.70 to 2.39) P=0.0004	2.71 (0.32)	1.70 (0.81 to 2.59) P=0.0002
Control	99	3.86±3.34	0.45±0.25		0.80 (0.31)		1.01 (0.32)
WIQ distance score, 0–100 scale (100 is best)
Exercise	100	34.0±27.7	1.44±2.07	2.63 (−3.16 to 8.41) P=0.3721	7.05 (2.33)	6.59 (0.04 to 13.15) P=0.0488	6.55 (2.38)	2.37 (−4.33 to 9.08) P=0.4859
Control	99	32.7±26.3	−1.19±2.08		0.45 (2.37)		4.18 (2.42)
WIQ speed score, 0–100 scale (100 is best)
Exercise	100	36.91±24.89	2.78±1.92	3.26 (−2.12 to 8.64) P=0.2335	1.43 (2.00)	−2.13 (−7.74 to 3.49) P=0.4564	5.66 (2.11)	1.68 (−4.24 to 7.59) P=0.5772
Control	100	34.98±24.77	−0.48±1.94		3.56 (2.03)		3.98 (2.13)
WIQ stair‐climbing score, 0–100 scale (100 is best)
Exercise	99	52.10±28.91	−2.55±2.27	−1.74 (−8.09 to 4.61) P=0.5897	0.65 (2.17)	0.19 (−5.89 to 6.27) P=0.9512	4.82 (2.35)	4.13 (−2.43 to 10.70) P=0.2160
Control	100	46.67±28.75	−0.81±2.28		0.46 (2.18)		0.69 (2.35)

WIQ indicates Walking Impairment Questionnaire.

Figure 2. Changes over time in 6‐minute walk distance among participants randomized to 26 weeks of supervised treadmill exercise or a nonexercise control group.

Maximal Treadmill Walking Time

Compared with controls, supervised treadmill exercise significantly improved maximal treadmill walking time by 2.47 minutes (95% CI, +1.64–+3.29; P < 0.001) at the 6‐week follow‐up, 3.64 minutes (95% CI, +2.68–+4.60; P < 0.001) at the 12‐week follow‐up, and 3.73 minutes (95% CI, +2.69–+4.77; P < 0.001) at the 26‐week follow‐up (Table 2 and Figure 3A). The 2.47‐minute improvement at the 6‐week follow‐up represented 66.2% of the 3.73‐minute improvement observed at the 26‐week follow‐up, whereas the 3.64‐minute improvement at the 12‐week follow‐up represented 97.6% of the improvement at the 26‐week follow‐up.

Figure 3. Changes over time in maximum and pain‐free treadmill walking time among participants randomized to 26 weeks of supervised treadmill exercise or a nonexercise control group.

A, Temporal change in maximal treadmill walking time. B, Temporal change in pain‐free treadmill walking time.

Pain‐Free Treadmill Walking Time

Compared with controls, supervised treadmill exercise significantly improved pain‐free treadmill walking time by 0.95 minutes (95% CI, +0.24–+1.65; P=0.009) at the 6‐week follow‐up, 1.54 minutes (95% CI, +0.70–+2.39; P=0.0004) at the 12‐week follow‐up, and 1.70 minutes (95% CI, +0.81–+2.59; P=0.0002) at the 26‐week follow‐up. The 0.95‐minute improvement at the 6‐week follow‐up represented 55.9% of the 1.70‐minute improvement in pain‐free walking distance observed at the 26‐week follow‐up, and the 1.54‐minute improvement at the 12‐week follow‐up represented 90.6% of the improvement at the 26‐week follow‐up.

Patient‐Reported Outcome Measures

Compared with control, supervised treadmill exercise had no meaningful or significant effect on the WIQ distance score (score range, 0–100 [100 is best]) at the 6‐week follow‐up (+2.6 [95% CI, −3.2 to +8.4]; P=0.37), significantly improved the WIQ distance score by 6.6 (95% CI, +0.04–+13.15; P=0.049) at the 12‐week follow‐up, and had no meaningful or significant effect on the WIQ distance score at the 26‐week follow‐up (+2.4 [95% CI, −4.3 to +9.1]; P=0.49) (Table 2 and Figure 4). Compared with control, there were no significant effects of supervised treadmill exercise on WIQ speed or WIQ stair‐climbing scores at any follow‐up time point (Table 2).

Figure 4. Changes over time in the WIQ distance score among participants randomized to 26 weeks of supervised treadmill exercise or a nonexercise control group.

WIQ indicates Walking Impairment Questionnaire.

DISCUSSION

In this post hoc analysis of the PROPEL trial, compared with a control group, supervised treadmill exercise improved 6‐minute walk distance by 13.0 m after 6 weeks of supervised exercise, consistent with a modest meaningful improvement, and representing just 38.3% of the maximal improvement in 6‐minute walk distance observed after 26 weeks of exercise. These results demonstrate that supervised treadmill exercise has a gradual effect on improving walking endurance, and that improvement in 6‐minute walk distance after 6 weeks of exercise is meaningful but modest, representing approximately one‐third of the meaningful, large, and statistically significant benefit attained after 26 weeks of exercise.

In contrast to findings for the 6‐minute walk distance, compared with controls, supervised treadmill exercise improved maximal treadmill walking distance by 2.47 minutes at the 6‐week follow‐up, consistent with a clinically important improvement, and representing about two‐thirds of the maximal treadmill walking time attained at the 26‐week follow‐up. These results demonstrate a much more rapid effect of supervised treadmill exercise on improvement in maximal treadmill walking time, compared with the more gradual improvement in 6‐minute walk distance. The more rapid improvement in treadmill walking distance, after initiation of a supervised treadmill exercise program, could be due to a learning effect, which has been documented previously for supervised treadmill exercise and its effect on treadmill walking performance. ¹⁸

Results reported here for the 6‐minute walk distance suggest that people with PAD may not appreciate large meaningful improvements in walking during their daily lives until >6 weeks after beginning a supervised exercise program. The gradual improvement in 6‐minute walk distance in response to treadmill exercise is likely to be particularly pertinent to experiencing walking improvement in daily life, because the 6‐minute walk better reflects walking during daily life than treadmill walking. ⁹ Consistent with this hypothesis, compared with controls, supervised treadmill exercise did not significantly improve the WIQ distance score, a patient‐reported outcome measure, at the 6‐week follow‐up, and only attained a significant and clinically meaningful improvement in the WIQ distance score at the 12‐week follow‐up. In contrast to findings for the 6‐minute walk distance and treadmill walking time, beneficial effects of supervised walking exercise on the WIQ distance score at the 12‐week follow‐up were lost by the 26‐week follow‐up. It is possible that participants in the supervised exercise group adapted gradually to their improved walking ability and no longer appreciated the magnitude of their improved walking performance at the 26‐week follow‐up, compared with baseline. ¹⁹ The lack of effect of supervised treadmill exercise on improvements in the WIQ walking speed and stair‐climbing scores may be due to the fact that supervised treadmill exercise specifically improves walking endurance, but is not well‐suited to improving walking speed over short distances or stair‐climbing ability.

Gardner et al randomized 142 participants with PAD to supervised treadmill exercise or a usual care control group for 6 months and measured outcomes at 2 months, 4 months, and 6 months after randomization. One hundred seven participants (75%) completed 6‐month follow‐up testing. Results showed that most of the effects of supervised treadmill exercise on maximal treadmill walking, pain‐free treadmill walking time, and the WIQ distance score were observed at the 2‐month follow‐up. ²⁰ The magnitude of the effect of exercise on the WIQ distance score was greatest at the 2‐month follow‐up (+13 points compared with controls) and diminished over time (+9 points compared with controls at the 6‐month follow‐up). Based on these results, Gardner et al concluded that a 2‐month supervised exercise program may be sufficient for people with PAD. In contrast, results reported here suggest that 6 weeks of supervised treadmill exercise is not sufficient to attain a large and meaningful improvement in either the objective 6‐minute walk distance measure or a patient‐reported outcome measure for walking distance.

Based on findings reported here, clinicians prescribing supervised treadmill exercise for people with PAD should advise them that large improvements are not likely to be perceived in daily life until >6 weeks after starting the supervised exercise program. This duration of time represents more than half of the 12‐week supervised exercise program covered by the Centers for Medicare and Medicaid Services. Although results reported here demonstrated that 12 weeks of supervised exercise was sufficient to attain near maximal gains in 6‐minute walk distance and maximal treadmill walking distance, few people with PAD who begin Centers for Medicare and Medicaid Services‐covered supervised exercise therapy complete all of the sessions. ³ Future study is needed to determine whether advising patients about the gradual nature of improvement following initiation of a supervised treadmill exercise program can improve adherence to supervised exercise. ¹⁹

Limitations

This study has several limitations. First, outcomes were measured at the 6‐week, 12‐week, and 26‐week follow‐up. Effects of supervised treadmill exercise on outcomes measured at other time points, such as the 8‐week follow‐up or 52‐week follow‐up, were not measured. Second, this clinical trial provided transportation for the exercise sessions, and the supervised exercise was administered by study investigators without a copay. The findings may not be fully generalizable to people engaged in supervised treadmill exercise covered by the Centers for Medicare and Medicaid Services. Third, additional potentially useful outcomes, such as physical activity, were not measured. Fourth, in addition to comparing supervised exercise and placebo, participants in this trial were randomized to GM‐CSF or placebo, using a 2 × 2 factorial design. However, because GM‐CSF did not significantly improve any study outcomes, ⁴ participants randomized to GM‐CSF were included in the current analyses, and analyses adjusted for GM‐CSF randomization assignment.

CONCLUSIONS

In people with PAD, >6 weeks of supervised treadmill exercise were necessary to attain a large meaningful gain in 6‐minute walk, and large meaningful gains were measurable by week 12 of supervised exercise. Meaningful improvement in participant‐reported walking ability was first observed at the 12‐week follow‐up, but this statistically significant benefit was gone by the 26‐week follow‐up.

Sources of Funding

This work was funded by the National Heart Lung and Blood Institute (R01‐HL107510) and supported in part by the intramural branch of the National Institute on Aging.

Disclosures

Dr McDermott has received research support from Mars and ACI Medical in the past 12 months. She has received consulting fees from Eli Lilly and has ownership in a Vanguard health care Exchange Traded Funds. The remaining authors have no disclosures to report.

Supporting information

Table S1