Integrating Survivors of Stroke Into Exercise-Based Cardiac Rehabilitation Improves Endurance and Functional Strength.

Background Cardiac rehabilitation (CR) is a structured exercise program prevalent in the United States for people with cardiovascular disease that has been shown to increase cardiovascular endurance and improve quality of life. Despite similar cardiovascular risk factors, stroke is not among the covered diagnoses for CR. The purpose of this study was to examine the participant impact of integrating survivors of stroke into the exercise portion of an existing hospital-based CR program through measures of physical function and other health impacts and through qualitative evaluation of participant perception. Methods and Results Subacute and chronic survivors of stroke were integrated into a standard 12-week, 3 sessions per week, exercise-based CR program. A total of 29 began the program, 24 completed the program, and 18 were available for 6-month follow-up. Quantitative measures were compared preprogram with postprogram with t-test or equivalent, and preprogram with postprogram to 6-month follow-up with ANOVA or equivalent. Semistructured interviews were completed with 11 participants postprogram. Exercise-based CR had significant impacts on cardiovascular endurance preprogram to postprogram, with maintenance at 6-month follow-up. The participants improved on the 6-minute walk test on average by 61.92 m(95% CI, 33.99-89.84 m), and maximum metabolic equivalents improved by a median of 3.6 (interquartile range, 2.35). Five times sit to stand (functional strength) improved preprogram to postprogram by a median of 2.85 s (interquartile range, 4.03 s). Qualitative findings highlight additional health improvements. Most participants (83% [15/18]) reported continued exercise at follow-up. Conclusions Exercise-based CR has the potential to improve cardiovascular endurance, health status, and quality of life for survivors of stroke. Registration URL: https://www.clinicaltrials.gov; Unique identifier: NCT03706105.

6‐minute walk test

cardiac rehabilitation

physical activity

rating of perceived exertion

Stroke Impact Scale

Clinical Perspective

What Is New?

Including survivors of stroke into exercise‐based cardiac rehabilitation is an innovative way to reach an additional clinical population at risk for cardiovascular complications with structured exercise.

The current study demonstrates that the dosage and intensity of exercise‐based cardiac rehabilitation have the potential to improve cardiovascular endurance, health status, and quality of life, with some improvements persisting after 6 months.

What Are the Clinical Implications?

Medical and rehabilitation providers could have a new referral option to a standardized program shown to improve cardiovascular endurance, health status, and quality of life for survivors of stroke.

Currently, there are no standardized exercise programs for people poststroke, so survivors are left to manage their activity on their own.

Many survivors of stroke face barriers and are generally inactive. Exercise‐based cardiac rehabilitation would provide an option for survivors over self‐management.

Physical inactivity is a health concern for most of the 7 million survivors of stroke in the United States who face increased risk for additional stroke and cardiovascular disease. Exercise can mitigate these risks, but survivors of stroke are not exercising; 58% fail to meet stroke guidelines for physical activity (PA). , Although many survivors of stroke receive physical therapy immediately poststroke, time barriers and an emphasis on functional activities limit cardiovascular exercise intervention. , As a result, survivors remain deconditioned after traditional rehabilitation, when they transition from one‐on‐one care with a physical therapist to self‐directed individual activity. , The lack of appropriate community group exercise programs impedes continuation of supervised PA. Without guidance or knowledge on appropriate activity, most survivors of stroke do not continue to exercise or engage in PA postrehabilitation. , , In addition to deconditioning and risk factor reduction, exercise may positively impact poststroke depression, fatigue, and community participation. , ,

Structured exercise programs offer an opportunity to break the cycle of inactivity, reduce future cardiac risk, and improve perceived confidence and quality of life. , , In the United States, cardiac rehabilitation (CR) is a structured and prevalent exercise program for people with cardiovascular diagnoses, such as myocardial infarction. Participation in CR has been shown to increase functional exercise capacity and improve perceived health‐related quality of life for traditional participants. , , Previous studies of cardiovascular training in survivors of stroke have demonstrated that they can safely perform aerobic exercise and achieve health benefits. , , , , , , Variation in dosage, staffing, and mode of activity impacts the external validity of these studies , , , , , , ; therefore, more knowledge is required to determine if benefits translate into existing CR programs. Research in Canada suggests the potential for integration of survivors of stroke into existing CR programs; however, the dosage and insurance climate differ from US programs. , , Effectiveness in existing CR programs for survivors of stroke in the United States that follow Medicare guidelines has not been investigated. Evaluation of these programs is supported by the American Heart Association and the American Stroke Association.

The primary aim of this study was to investigate the impact of an existing CR program for survivors of stroke through pilot measures for physical function (cardiovascular endurance, functional strength, and walking speed) and for other health impacts (quality of life, balance confidence, depression, and exercise habits). A secondary aim was to evaluate participant perception of program impact on physical function, health, and future exercise plans.

Methods

The study was conducted at multihospital health system's CR facility, located in the southeast United States. A mixed methods design combined a single group, preprogram‐postprogram design, with a pragmatic qualitative inquiry of participant perception to enhance interpretation of results. All data and materials have been made publicly available at Open Science Framework.

The project was approved by the health system Institutional Review Board and acknowledged by the University of South Carolina Institutional Review Board. The study was a registered clinical trial through the US National Library of Medicine (ClinicalTrials.gov identifier: NCT03706105). Participation was voluntary, and individuals were able to opt out at any time. The program was free for study participants, with program costs ($237 per participant) covered by study grant funding. Participants provided informed consent and an authorization for use and disclosure of protected health information for research purposes.

Recruiting and Study Criteria

Potential participants were recruited from health system rehabilitation clinicians, physicians, stroke team nurses, and CR providers. In addition, survivors of stroke were recruited directly from the community through stroke support groups and word‐of‐mouth referrals. Multiple referral targets were chosen to reflect likely general community program sources. Those who were interested were screened for eligibility by primary investigator (Figure 1).

Figure 1

Study flowchart.

COVID‐19 indicates coronavirus disease 2019; Phone, telephone; and Rehab, rehabilitation.

The following inclusion criteria were applied: (1) diagnosed with most recent stroke at least 3 months prior (stroke diagnosis code and date provided on physician referral; multiple stroke occurrences allowed); (2) completed physical and occupational therapy, if applicable; (3) cleared by treating physician or nurse practitioner to participate; (4) demonstrated ability to walk at least 40 m with or without an assistive device; (5) demonstrated ability to transfer from sit to stand without assistance; and (6) demonstrated ability to follow instructions and to communicate exertion, pain, and distress. Potential participants were excluded from the study for any of the following: (1) presence of a medical problem rendering exercise unsafe; (2) complaints of significant pain that interfered with movement; or (3) history of an additional, nonstroke, neurologic condition.

Once eligibility was determined, the primary investigator (a physical therapist) screened participants for safety, and participants completed a demographic intake form and a battery of outcome measures. Screening (one‐time mobility assessment at preprogram only) assessed range of motion, strength, balance limitations, and gait alterations. Initial determination of safety to participate was determined medically by the referring physician and from an independent mobility perspective by the study primary investigator.

Program Procedures and Progression

The mobility screening measures and the preprogram outcome measures were shared with the CR staff to establish initial exercise intensity goals and modifications to the standard CR exercise program. Participants were integrated into the standard CR exercise program. Aside from modifications provided by the participant evaluation, the exercise intervention did not differ from the standardized program. The program began with analysis to determine baseline levels of exercise intensity in metabolic equivalent tasks (METs) based on participant's 6‐minute walk test (6MWT) results. Target heart rate was estimated from resting heart rate and 6MWT completion heart rate. Target exercise rating of perceived exertion (RPE) levels were set from 11 to 14 (somewhat hard to hard) on a scale of 6 to 20. Activity plans were also individualized. Training sessions were 3 times a week for 12 weeks, with a target of 31 to 50 minutes of moderate aerobic activity each session. Additional optional activities included strengthening, stretching, and/or relaxation. Although components varied by session and individual, the general format was warm‐up, cardiovascular endurance activities (treadmill, recumbent step machine, recumbent bike, and over ground walking), cooldown, and optional activities.

Progression in the program was determined by participant‐reported RPE. If RPE was consistently rated ≤11, effort was increased to reach a rate of 14. Discontinuation of a session or the program was determined by standard health system protocols (Data S1).

Weekly formal educational sessions were not adapted for survivors of stroke. Sessions were available to survivors of stroke but were not used. Medical evaluation and monitoring for heart rate, blood pressure, heart rhythms, and blood glucose were completed on the basis of protocols and clinical expertise of supervising nurses and exercise physiologists. Providers educated participants informally on self‐monitoring, home exercise, diet, and nutrition during exercise sessions and recorded education in daily notes.

At the end of the 12‐week CR exercise program, all participants were reassessed using the study outcome measures and an additional inquiry of participant's postprogram exercise plans. Completion of the program included all participants with final outcome measures available at postprogram assessment.

Six months after the end of the CR exercise program (follow‐up), program completers were invited to return for the last outcome measure assessment, which included an additional self‐report of current exercise habits.

Outcome Measures

Study primary investigator administered outcome measures preprogram, postprogram, and 6 months postprogram (Table 1 and Table S1). Maximum METs are a standard measure of exercise tolerance and functional capacity in CR programs. All remaining outcome measures have been validated in survivors of stroke. , , , , , , , , ,

Table 1

Outcome Measures

Outcome Measure	Assessment
6‐min Walk test	Cardiovascular endurance; walking capacity; initial fitness in CR programs 18 , 34 Measured as distance in meters
5 Times sit to stand test	Functional lower body strength 38 , 41 Measured as seconds to complete
10‐m Walk test	Self‐selected and fast walking speed 34 , 42 Measured as m/s
Maximum METs*	Standard measure of exercise tolerance and functional capacity in CR programs 33 Measured from 1 (very low) to 13 (very high) fitness
Activities‐Specific Balance Confidence Scale	Self‐perception of balance confidence 37 Total score, 0%–100% confidence A score <67% indicates an increased risk of falls 43
Stroke Impact Scale	Impact of stroke on 8 domains: mobility, participation, activities of daily living, hand function, strength, communication, emotion, and memory/thinking 35 Domain subscores from 0% (significant impact) to 100% (no impact)
Short Self‐Efficacy for Exercise and Short Outcome Expectations for Exercise Scales	Confidence to complete exercise behaviors, such as exercising alone or through fatigue 39 Exercise outcome expectations, such as belief that exercise improves mood or improves endurance 39 Scores are from 1 (low) to 5 (high)
Patient Health Questionnaire‐9	Depression 40 Total score of 0–27, with categories of 0 (no depression), 1–9 (minimum to mild depression), 10–14 (moderate depression), and 15–27 (moderately severe to severe depression)

CR indicates cardiac rehabilitation; and MET, metabolic equivalent task.

Maximum METs were calculated at the initial visit and final visits as part of the standard program and were not reassessed at 6‐month follow‐up.

Cardiovascular endurance; walking capacity; initial fitness in CR programs ,

Functional lower body strength ,

Self‐selected and fast walking speed ,

Standard measure of exercise tolerance and functional capacity in CR programs

Self‐perception of balance confidence

A score <67% indicates an increased risk of falls

Impact of stroke on 8 domains: mobility, participation, activities of daily living, hand function, strength, communication, emotion, and memory/thinking

Confidence to complete exercise behaviors, such as exercising alone or through fatigue

Exercise outcome expectations, such as belief that exercise improves mood or improves endurance

Depression

Statistical Analysis

Power analysis was conducted on the basis of findings from a previous study with a similar population and exercise intervention. Calculations suggested 22 participants would provide 80% power to detect preprogram‐postprogram changes moderate in magnitude (effect size d=0.56) in the 6MWT.

Participant demographic information and outcome measures were aggregated, with means, medians, and SDs calculated. Aggregate means and SDs for total number of sessions, session time, and minimum and maximum RPE were calculated for fidelity of the program to contextualize outcome measure results. The outcome measure data for the full sample of completers preprogram‐postprogram (n=24) were analyzed using a paired t‐test or Wilcoxon signed rank test (for those not normally distributed or ordinal variables). The α level was set at 0.01 because of multiple comparisons and the desire to minimize both type I and type II errors. Effect sizes (Cohen d values) were generated. Finally, for the subset of the sample in whom 6‐month follow‐up data were available (n=18), a repeated measures ANOVA or a Friedman test was completed for those measures found to be statistically different in the preprogram‐postprogram comparison. Bonferroni adjustments were made to the ANOVA and Friedman tests. Analysis was completed with IBM SPSS Statistics for Windows, version 26 (IBM Corp, Armonk, NY).

Qualitative Methods and Analysis

A pragmatic qualitative approach evaluated participant perspectives on program outcomes and future exercise plans. , Interview questions were developed on the basis of study aims and framed by the World Health Organization's International Classification of Function and Social Cognitive Theory. , The interview guide is provided in Table S2.

The sampling plan was nested within the original sample, and potential qualitative participants were identified sequentially as they completed the program. All participants who began the program and met qualitative eligibility requirements were invited to voluntarily participate. Participants who had previously participated in CR or had verbal communication limitations were ineligible for the qualitative portion. Participants in the qualitative portion of the study provided separate informed consent and received a $20 gift card as an incentive. Semistructured interviews were conducted in a private room at the time of postprogram outcome measure collection. Interviews were audio recorded and transcribed verbatim. Questions were piloted in the first 2 interviews and revised slightly. The number of participant interviews was determined by maximal voluntary participation of those eligible to achieve saturation of themes. Field notes and addition of quantitative data added rigor.

The researchers completed inductive thematic analysis using NVivo software version 12 (QSR International), with deidentified transcripts and observation notes. One researcher coded all interviews to phrases or sentences directly from the transcripts and structured observations. Results were reviewed with a second researcher. Both researchers then independently performed inductive categorization of the open coding. A final thematic codebook was agreed on. Each researcher updated independent coding to reflect the codebook. Data conflicting with primary themes were identified to present alternative viewpoints. Results were compared, and any discrepancies were resolved together. Final coding was reviewed with a third researcher, where naming conventions and minor alterations were made.

Results

Study referrals began in August 2018, with active program participation through November 2019 and 6‐month follow‐ups through February 2020. Of the 29 participants starting the program, there were 24 completers. Of the 5 noncompleters, 2 dropped out, 1 was unable to be independent because of agitation/cognitive issues, 1 had an additional mild stroke and no longer met the inclusion criteria, and 1 had recurrent bronchitis. Of the 24 completers, 18 returned for 6‐month follow‐up assessments (Figure 1). Six completers were lost to follow‐up: 1 declined, and 5 were impacted by coronavirus disease 2019 facility closure. Of the 13 eligible completers, 11 participated in the qualitative interviews.

Program participant demographics are presented in Table 2. The most common comorbid health conditions included high blood pressure (83.3%), diabetes mellitus (41.7%), cardiovascular disease (41.7%), and arthritis (29.2%). Completers time since stroke varied, with 38% (9) <1 year poststroke and 62% (15) ≥1 year poststroke. Of completers, 29% (7) reported a history of multiple strokes. In addition, 29% (7) of completers previously participated in CR for a traditional cardiac diagnosis. The average number of sessions per completer was 25.25 (95% CI, 22.91–27.92), with a range of 12 to 36 sessions. Completers averaged 38.93 (95% CI, 36.54–41.31) exercise minutes per session and met RPE targets of 11 (light) to 14 (somewhat hard), with minimum RPE median of 11 (interquartile range, 0.625) and maximum RPE median of 13 (interquartile range, 1.00) across all sessions. There were no safety events related to exercise intensity, and all participants met the prescribed RPE ranges.

Table 2

Demographics of Program Participants

Variable	Completers (n=24)	Noncompleters (n=5)
Sex, % (n)
Men	79 (19)	60 (3)
Women	21 (5)	40 (2)
Age, mean (SD), y	62.2 (12.4)	68.4 (15.0)
Race/ethnicity, % (n)
White	71 (17)	60 (3)
Black	25 (6)	40 (2)
Asian	4 (1)
Type of stroke, % (n)*
Ischemic	65 (15)	40 (2)
Hemorrhagic	12.5 (3)	40 (2)
Unknown	25 (6)	20 (1)
Time since stroke, mean (SD), mo	29.7 (29.9)	37.0 (41.8)
Initial 6MWT distance category, % (n) †
≥288 m	83.3 (20)	40 (2)
<288 m	16.7 (4)	40 (2)
No 6MWT		20 (1)
Initial SSWS, mean (SD), m/s	1.17 (0.21)	0.67 (0.28)
Preprogram exercise level, % (n)
None	12.5 (3)	40 (2)
<1× wk	12.5 (3)	0 (0)
1–3× wk	37.5 (9)	60 (3)
>3× wk	37.5 (9)	0 (0)

6MWT indicates 6‐minute walk test; and SSWS, self‐selected walking speed.

Stroke type was self‐reported by participant.

The 6MWT ≥288 m indicates community ambulator status.

Results of preprogram to postprogram comparisons are presented in Tables 3 and 4, , and results for preprogram, postprogram, and 6‐month postprogram comparisons are presented in Figure 2. Outcomes and qualitative themes are presented for the following: (1) cardiovascular endurance, (2) other physical outcomes and general health, (3) emotional health, (4) exercise self‐efficacy and outcomes expectations, and (5) postprogram exercise.

Table 3

Results Preprogram to Postprogram: Paired t‐Test Outcome Measures

Test	No.	Proprogram, Mean (SD)	Postprogram, Mean (SD)	Change, Mean (SD)	95% CI of Mean Change	t	df	Significance (P Value)	Effect Size (d)
6MWT, m*	24	397.80 (119.23)	459.71 (118.46)	↑ 61.92 (66.13) †	33.99–89.84	4.587	23	<0.001 ‡	0.94
FWS, m/s §	23	1.50 (0.42)	1.59 (0.50)	↑ 0.09 (0.18)	0.02–0.17	3.167	22	0.019

6MWT indicates 6‐minute walk test; and FWS, fast walking speed.

Higher distance indicates an improvement in score.

Greater than the minimal detectable change for stroke of 31 m.

Statistically significant changes.

Higher number indicates a faster walking speed.

Table 4

Results Preprogram to Postprogram: Wilcoxon Signed Rank Test Outcome Measures

Test	No.	Preprogam, Median (IQR)	Postprogram, Median (IQR)	Change, Median (IQR)*	Z	Significance (P Value)
Cardiovascular endurance measures
SIS‐Mobility (0%–100%)	24	72.22 (31.25)	77.78 (29.17)	↑ 6.94 (11.11) †	2.665	0.008 ‡
MET maximum (1 [low] to 13 [high])	24	2.95 (0.88)	6.00 (3.00)	↑ 3.6 (2.35)	4.199	<0.001 ‡
Physical function measures
FTSS test, s (lower score is better)	23	14.42 (11.14)	12.2 (6.47)	↓ 2.85 (4.03) §	−3.528	<0.001 ‡
SSWS, m/s (higher score is faster)	23	1.16 (0.34)	1.18 (0.38)	↑ 0.02 (0.16)	1.095	0.274
ABC Scale score, % confidence	24	73.44 (35.28)	86.38 (21.48)	↑ 1.78 (14.61)	1.686	0.092
SIS‐Physical (0%–100%)	24	62.50 (37.50)	75.00 (37.50)	− 0.00 (18.75)	1.350	0.177
Quality‐of‐life measures (SIS‐Other subscales)
SIS‐Mood (0%–100%)	24	77.78 (29.17)	86.11 (20.14)	↑ 4.17 (13.19)	1.869	0.062
SIS‐Memory (0%–100%)	24	78.57 (37.50)	82.14 (27.68)	− 0.00 (16.96)	2.076	0.038
SIS‐Communication (0%–100%)	24	87.50 (50.00)	83.93 (31.25)	− 0.00 (13.39)	1.623	0.105
SIS‐ADL (0%–100%) ‖	24	90.00 (31.25)	90.00 (16.90)	↑ 2.50 (9.38)	2.425	0.013
SIS‐Hand (0%–100%)	24	85.00 (43.75)	92.50 (40.00)	− 0.00 (10.00)	1.002	0.316
SIS‐Participation (0%–100%)	24	70.31 (53.91)	76.56 (39.06)	↑ 3.13 (21.09)	1.976	0.048
SIS‐Recovery (0%–100%)	24	80.00 (15.00)	82.50 (15.00)	↑ 5.00 (10.00)	1.715	0.086
Self‐efficacy measures (exercise and outcome expectations)
SSEE Scale (1 [low] to 5 [high])	22	4.20 (1.19)	4.50 (0.69)	↑ 0.25 (1.06)	2.023	0.043
SOEE Scale (1 [low] to 5 [high])	22	4.00 (0.60)	4.20 (1.60)	↑ 0.20 (0.65)	2.397	0.017

ABC indicates Activities‐Specific Balance Confidence; ADL, activities of daily living; FTSS, 5 times sit to stand; IQR, interquartile range; MET, metabolic equivalent task; SIS, Stroke Impact Scale; SOEE, Short Outcome Expectations for Exercise; SSEE, Short Self‐Efficacy for Exercise; and SSWS, self‐selected walking speed.

↑ Indicates improvement, ↓ indicates decline, and − indicates no change.

Change is greater than the clinically important difference rate of 4.5%.

Statistically significant changes <0.01.

Change is greater than the 1.14‐s minimal detectable change for survivors of stroke.

SIS‐ADL change score distribution was not symmetrical, so a sign test was completed instead of Wilcoxon signed rank test.

Figure 2

Changes over time.

A, Improvements over time in mean 6‐minute walk test (6MWT) distance, ANOVA, n=18 (P=0.001), effect size 0.41. *Indicates 66.51‐m change (95% CI, 12.80–120.24 m): P=0.013. **Indicates 71.67‐m change (95% CI, 26.19–117.15 m): P=0.002. 6MWT distance mean (SD) preprogram, 403.18 m (108.34 m); postprogram, 474.85 m (101.49 m); and 6‐month follow‐up, 469.70 m (26.17 m). Error bars are SD. B, Improvements over time in the 5 times sit to stand (FTSS) test time in pairwise comparisons, Friedman test, n=17 (P<0.001). Faster time indicates a better score. Box plots show median, interquartile range, minimum, and maximum. FTSS median test time (interquartile range) preprogram, 14.23 s (7.15 s); postprogram, 11.88 s (5.44 s); 6‐month follow‐up, 10.75 s (6.92 s). One participant used one upper extremity for support to rise to standing during testing. **P<0.001.

Cardiovascular Endurance

The 6MWT, the Stroke Impact Scale (SIS)‐Mobility subscale, and maximum METs measured cardiovascular endurance. The 6MWT, the primary outcome measure for aerobic and walking capacity, improved by 61.92 m (95% CI, 33.99–89.84 m) preprogram‐postprogram, with a large effect size (0.94), which is greater than the minimal detectable change of 34 m for survivors of stroke (Table 3). , Improvements in 6MWT distances were maintained at 6‐month follow‐up results (Figure 2A).

The SIS‐Mobility subscale had a statistically significant median improvement postprogram of 6.94%, which is greater than the clinically important difference of 4.5% (Table 4). However, comparisons including the 6‐month follow‐up did not find a statistically different change over time (P=0.057).

Maximum METs progressed with a median difference of 3.6 (interquartile range, 2.35) from the beginning of the program (first session) to the end of the program (final session at week 12) (Table 4).

Qualitative themes related to endurance included improved stamina, improved stair climbing, and needing less rest breaks during activity. For some, improved endurance impacted their PA tolerance, and they were able to do more of what they enjoy.

“I think that, because of improving my stamina and my endurance, that has um, helped me in other things. So, um it, it's allowed me to do a little bit more dancing, and a little bit and, and not have to constantly be resting as much … ”

“Um, yes it's … think … just walking and uh, uh just general um physical activities and I think … I don't wanna over say it, it but ah I have to think that, ah, it's improved my every day, ah, activity tolerance.”

Other Physical and General Health Outcomes

Other physical outcome measures included strength, walking speed, stability and balance, and general health impacts. The 5 times sit to stand test measured lower extremity strength, which improved by a median of 2.85 s (interquartile range, 4.03 s) (Table 4). The 5 times sit to stand test gains remained at 6‐month follow‐up (Figure 2B). Participants maintained but did not improve their walking speed, Activities‐Specific Balance Confidence Scale score, or SIS‐Physical subscale score preprogram to postprogram. The proportion of participants in the highest fall risk category (Activities‐Specific Balance Confidence Scale score, <67%) was 33.3% (n=8) preprogram and 20.8% (n=5) postprogram. A few participants noticed balance improvements, with qualitative themes noting improved reaction times and better balance confidence. Several participants noted improvements in their walking, often related to improved stability, balance, and strength.

“My reflexes are getting quicker. I can, I can look both ways quicker on the crosswalk, and I can run across the street and I can read the car coming at me quicker.”

General health outcome measures included the remaining Stroke Impact Scale subscales (SIS‐Activities of Daily Living, SIS‐Hand, SIS‐Communication, SIS‐Memory, SIS‐Participation, and SIS‐Recovery [overall self‐rated stroke recovery]), all without statistically significant changes. A few participants noted health changes not covered above; themes included weight loss/improved physical appearance, positive medication changes, and improved awareness of importance of health.

“… I have been thinking about my health and how to live the best life that I can and I think this program has encouraged this thinking on my part.”

“Okay, and what are you thinking you need to do to live the best life? Like are you thinking about changes you need to make?”

“Well I got a referral for speech therapy and I am doing that now, and I am not sure that would have occurred to me before. And um I think the eating has been better.”

Emotional Health

Several study outcomes measured emotional health: the Patient Health Questionnaire‐9, the SIS‐Mood subscale, and an analysis of qualitative interviews. The SIS‐Mood subscale did not have statistically significant changes preprogram‐postprogram. Twenty‐three participants had initial Patient Health Questionnaire‐9 depression screen scores at preprogram: 11.5% (n=2) in the moderately severe to severe depression category, 17.4% (n=4) in moderate depression category, 69.6% (n=16) in the minimum to mild depression category, and 4.3% (n=1) in the no depression category. These depression category proportions remained mostly unchanged at postprogram, where 24 participant scores were available, with 11.5% (n=2) in the moderately severe to severe depression category, 11.5% (n=2) in moderate depression category, 70.8% (n=17) in the minimum to mild depression category, and 12.5% (n=3) in the no depression category. Although a few participants noted no changes to mood or outlook as a result of the program, many participants noted improvements in emotional health, including reduced depression, contributions to a positive attitude, and improved self‐perception. Participants noted a new or renewed sense of enthusiasm for exercise or for engaging in activities and feeling more confident about their abilities.

“Overall experience was, it was, it was kind of life changing. Kind of life saving. Um, definitely haven't been nearly as depressed as I was before I came in here. Not at all. Um, and that doesn't just have to do with [life change]. It was, it was night and day difference. After about 2 weeks of being in here, it was night and day difference. From being really dark and, and in a really bad way. Um, really depressed, and, and trying to almost, uh, not really sure what to do with it, and I kind of starting, getting faith again, hope, feeling good, wanting to take care of myself, and, and just being happy.”

Exercise Self‐Efficacy, Exercise Outcome Expectations, and Postprogram Exercise

Participants had high initial scores for both the Short Self‐Efficacy for Exercise Scale (median, 4.20 of 5) and the Short Outcome Expectations for Exercise Scale (median, 4 of 5), indicating their confidence to exercise was high and that they anticipated benefits from exercise. Changes postprogram were not statistically significant (P>0.01).

Postprogram Exercise Plans

All completers had plans to continue exercise postprogram. Plans included continuing at CR through the self‐pay maintenance program, participating in group‐based exercise classes, joining a gym for aerobic and strength activities, doing exercise at home, and working with a personal trainer.

At 6‐month follow‐up, 83.3% (15/18) of participants reported engaging in exercise at least once a week, 44.4% (8/18) with a frequency of 1 to 3 times a week, and 38.9% (7/18) with a frequency of >3 times a week. Reported activities included walking (50%), gym strengthening (22.2%), gym aerobic (50%), home aerobic (33.3%), home strengthening (11.1%), group exercise (22.2%), and other (22.2%), which included swimming, yardwork, horseback riding, and running.

Discussion

After participation in exercise‐based CR, survivors of stroke made improvements in cardiovascular endurance, functional strength, and perceived mobility. Improvements in cardiovascular endurance and functional strength were maintained at the 6‐month follow‐up, suggesting the possibility of lasting changes. Qualitative results confirmed endurance and mobility improvements and highlighted additional improvements in emotional health. Improvements occurred regardless of self‐reported prior activity levels. Previous exercise experience combined with high levels of exercise self‐efficacy may have been a driver for initial participation.

Survivors Improve Cardiovascular Endurance

Survivors of stroke integrated into exercise‐based CR demonstrated improvements in cardiovascular endurance. The 6MWT improvements suggest better community walking status and real‐world walking capability. , This increase in capacity is especially important to survivors of stroke who have mobility impairments that result in a higher energy cost for walking. The 6MWT improvements were maintained at 6‐month follow‐up, supporting maintenance of gains after exercise‐based CR. The 61.92‐mchange preprogram‐postprogram was greater than a pooled mean change of 53.3 m from a recent meta‐analysis of aerobic programs for stroke survivors with similar dosage to CR. Maximum METs had a median increase of 3.6 METs preprogram‐postprogram. These changes are important measures of overall health. A meta‐analysis by Kodama et al found that in healthy individuals, for each one MET increase in exercise capacity, all‐cause mortality was reduced by 13% and incidence of coronary heart disease and cardiovascular disease was reduced by 15%. Similar results have been found for traditional CR participants. The SIS‐Mobility scale measures participant perception of home and community mobility capabilities, and improvement preprogram to postprogram corroborates the link between capacity and participation. These results were not maintained at 6‐month follow‐up, however. The addition of social support from other participants and from staff during the exercise‐based CR program may have impacted the SIS‐Mobility results, which did not continue in the follow‐up period. Qualitative themes of improved stamina impacting participant's daily activities support the quantitative findings of improved cardiovascular endurance and overall mobility.

Survivors Improve Functional Strength

Lower extremity strength improved preprogram to postprogram and was maintained at 6‐month follow‐up. In addition to measuring strength, the 5 times sit to stand test has speed and control components and functional correlates. , , For survivors of stroke, taking longer to complete the 5 times sit to stand test correlates with lower bilateral knee flexor strength and increased risk of falls. , For geriatric populations, which often include survivors of stroke, a slower time is predictive of less independence in activities of daily living within 3 years. Collectively, improvements in cardiovascular endurance and strength support the positive health and fitness benefits of integration of survivors of stroke into US‐based CR exercise programs.

Emotional Health and Self‐Efficacy for Exercise Had Ceiling Effects and Mixed Results

The initial scores measuring this construct indicated low initial depression in the sample (73.9% with no to mild depression on the Patient Health Questionnaire‐9) and higher initial mood (median of 77.78% on the SIS‐Mood subscale), leaving little room for change. Qualitative results suggest that participation in exercise‐based CR may impact emotional health for individuals, and this is supported by existing research. , , , The qualitative themes related to emotional health were participants finding renewed self‐confidence and sense of self. Higher self‐esteem is known to positively impact self‐perception of identity after stroke. , , A qualitative study by Erikson et al found that finding a positive new self‐identity after stroke was tied to engaging with others through meaningful activities, which a program like CR can provide.

Self‐efficacy for exercise and outcome expectation for exercise scores were high at preprogram, suggesting good to excellent confidence in exercise abilities (Short Self‐Efficacy for Exercise Scale: median, 4.2) and belief in benefits of exercise (Short Outcome Expectations for Exercise Scale: median, 4.0). With a maximum score of 5 on both the Short Self‐Efficacy for Exercise Scale and the Short Outcome Expectations for Exercise Scale, achieving significant changes was difficult because of a ceiling effect. The high initial scores in this sample may be related to the importance of having self‐efficacy and positive outcome expectations to commit to structured exercise programs. , All of the completers had concrete plans for continued exercise at the completion of the program. At 6 months, most were still active, suggesting that 12 weeks may be long enough to build habits for maintained activity. However, improved exercise habit results require more investigation, as this sample had a high proportion of participants with high self‐efficacy for exercise and some exercise experience before the program, both key drivers of ongoing PA in survivors of stroke.

Study Limitations

There are several study limitations related to the use of a single‐group pilot design with a convenience sample at a single exercise‐based CR program. Deliberately broad inclusion criteria allowed for capturing survivors in both the subacute and chronic phases of recovery. Chronic stroke survivors face persistent significant deficits in endurance, strength, and gait speed and the cardiovascular disease risk factors accompanying these deficits. The amount of benefit for those in the chronic phase is promising and suggests continued study for this population. Future studies with larger sample sizes may determine more clear pathways to the program and address potentially divergent needs between the 2 subgroups. Although a diverse sample of mobility impairments, sex, age, and racial/ethnic diversity was desired, most participants were White men with few mobility limitations and relatively high initial fitness levels. Future studies for survivors with less functional mobility may require further screening and expansion of CR protocols. In addition, future studies can expand to multiple health system sites and use a randomized control trial design with recruiting plans targeting participants with specific characteristics. Low participation rate for referred potential study participants mirrors participation rates for traditional CR participants, which can be as low as 20% to 30%. Institutional barriers, healthcare provider barriers, and individual barriers require exploration before successful implementation of future phase trials and community implementation. Uptake rates may be addressed through institutional buy‐in, education for healthcare providers and survivors of stroke, and elimination of barriers, where possible, before implementation. Study participants' failure to use the available education sessions was a missed opportunity for additional potential behavior change. Confusion by staff on study participant eligibility and participant belief that it was not necessary or appropriate to them were barriers that could be addressed with better education of staff in future studies.

Conclusions

Exercise‐based CR for survivors of stroke had a positive impact on cardiovascular endurance and functional strength. Exercise‐based CR also influenced participants' perception of their home and community mobility, their walking capability, and their emotional health. Improvements in METs correspond to reduced risk for mortality and cardiovascular disease. Despite similar cardiovascular risk factors to traditional CR participants and potential health benefits from participation, stroke is not among the covered diagnoses for CR services in the United States. Findings support the use of CR exercise programs for survivors of stroke after rehabilitation to improve endurance, health status, and quality of life. Further investigations can confirm findings and explore integrating survivors of stroke as a standard of care.

Sources of Funding

This work was supported by the University of South Carolina (USC) Behavioral‐Biomedical Interface Program (National Institute of General Medical Sciences/National Institutes of Health T32 2T326M081740‐11A1), 2019 American Heart Association (AHA) Pre‐Doctoral Fellowship, 2018 American Physical Therapy Association Health Policy and Administration Research Grant, 2019 USC Support to Promote Advancement of Research and Creativity Grant, AHA Grant 15SDG24970011, Promotion of Doctoral Studies–Level I Scholarship from the Foundation for Physical Therapy Research, and the Arnold Fellowship from the Arnold School of Public Health, USC.

Disclosures

None.

Data S1

Tables S1–S2

Click here for additional data file.