Stroke Recovery Program Incorporating Moderate-Intensity Aerobic Exercise Provides Sustained Improvement of Function: A 1-Year Follow-Up Study

HIGHLIGHTS

• •A comprehensive stroke recovery program sustains functional improvements at 1 year.
• •Improving long-term function of stroke survivors promotes functional independence.
• •Aerobic exercise benefits the cardiovascular and neurovascular systems after stroke.

Abstract

Objective

To evaluate the 1-year long-term functional improvements and sustainability of a comprehensive stroke recovery program (SRP) administered early in the subacute phase of recovery compared with stroke survivors who received usual care.

Design

A prospective matched cohort study.

Setting

Academic medical center with inpatient and outpatient rehabilitation facilities.

Participants

A total of 236 patients (N=236) with stroke admitted to an inpatient rehabilitation facility were included in this study. Of these, 152 participated in the SRP, and 84 patients received usual care therapies.

Interventions

Outpatient rehabilitation program including 36 sessions of moderate-intensity aerobic exercise and risk factor education administered in addition to usual care therapies.

Main Outcome Measures

Activity Measure for Post-Acute Care (AM-PAC) scores for Basic Mobility (BM), Daily Activity (DA), and Applied Cognitive (AC) were collected at admission and discharge from the hospital, at 30, 60, 90, 120 days after stroke, and 6-month and 12-month long-term follow-ups.

Results

The mixed effects linear regression model revealed a significant interaction effect between the intervention and time across all 3 functional domains (P<.001 each for BM, DA, AC). At the 1-year follow-up assessment, the mean score for AM-PAC BM was 61.5±10.7 for the SRP group compared with 55.1±15.5 in the control group (P<.001), 58.0±17.4 vs 52.7±17.5 for AM-PAC DA (P=.032), and 49.4±11.5 vs 44.5±12.4 for AM-PAC AC scores (P=.225), respectively.

Conclusions

An early outpatient comprehensive stroke rehabilitation program, including aerobic exercise and stroke risk factor education within the first 6 months of stroke, not only demonstrated functional improvements after the program but was able to achieve sustained functional improvements in mobility, self-care, and cognitive ability at 1 year. Improving the long-term function of stroke survivors promotes functional independence.

Stroke is a leading cause of serious long-term disability and a major public health concern.¹ The prevalence of stroke in the United States is on the rise, with an estimated 795,000 cases of stroke occurring annually, of which 1 in 4 individuals experience a recurrent stroke.^1,2 An estimated 7.6 million adults in the United States are currently affected by stroke, with approximately 50% of stroke survivors grappling with chronic disabilities.² Research indicates that up to 40% of individuals encounter limitations in performing activities essential for daily living.³ The long-term outlook for stroke survivors remains concerning, with studies suggesting that approximately 70% of stroke survivors are deceased or disabled within 5 years of their stroke.⁴ This underscores the urgent need to develop effective rehabilitation strategies that will have a long-term impact.

Although many studies have traditionally focused on data in the acute-subacute phase of recovery within the first 6 months after stroke, the evolving nature and the significant importance of long-term functional recovery prompt a shift in the conventional emphasis. Studies show that activities of daily living and functional independence do not stay at the same level after the initial recovery period between 3 months and 1 year after stroke.^5,6 Notably, the factors influencing long-term outcomes may differ from those affecting short-term recovery.7, 8, 9

As a result of the long-term disability associated with stroke, many stroke survivors require placement in long-term care facilities and nursing homes. For patients recovering from stroke, this greatly impacts their physical and mental well-being, as well as their ability to return to work and make future life plans.¹⁰ It has been proven to be a financial burden on our national health care system and significantly affects the quality of life of patients with stroke.³ Moreover, poststroke disability substantially increases the risk of adverse outcomes, including recurrent stroke and mortality.¹¹

In addition to functional limitations, most of stroke survivors also exhibit significant cardiovascular disease.12, 13, 14 Stroke survivors exhibit reduced oxygen uptake, decreased exercise endurance, and lower cardiovascular capacity compared with age and sex-adjusted sedentary individuals.^15,16 In addition, stroke survivors face increased metabolic demands and higher energy costs for even basic activities such as walking.^15,17 Emphasizing the cardiovascular needs of stroke survivors, the American Heart Association recommends moderate-intensity aerobic exercise lasting 20-60 minutes from 3 to 5 times per week as exercise/physical activity recommendations for stroke survivors.^3,18 Programs addressing the cardiovascular health of stroke survivors alongside traditional neurorehabilitation have shown a significant impact on the recovery of function.19, 20, 21, 22, 23 In addition, secondary prevention strategies such as modifiable risk factor education, nutritional support, and psychological services improve the cardiovascular health and overall well-being of patients with stroke.24, 25, 26

Studies incorporating multidisciplinary comprehensive approaches are best suited to address the barriers and complex needs of stroke survivors and provide the best functional outcomes, demonstrating not only functional improvements but also improved cardiovascular capacity, cognitive function, depression, and blood pressure.^23,27, 28, 29 Our previous studies reveal that a comprehensive program incorporating aerobic exercise and risk factor reduction education led to a 22% reduction in 1-year all-cause hospital readmissions and an overall 76% decrease in 1-year all-cause mortality.^20,30

Limited research has been conducted on the long-term impact of rehabilitation programs administered in the early subacute phase of stroke recovery. This is of significant importance to patients with stroke as long-term functional deficits lead to increased dependence on caregiver assistance or placement in long-term care facilities.

We previously published the results of a large cohort study examining the effects of a comprehensive outpatient stroke recovery program (SRP), including multiple disciplines and resources to assist in recovery.²⁰ Our SRP used a multidisciplinary comprehensive approach focusing on cardiovascular conditioning and neurorehabilitation. The program administered in the subacute phase of the recovery (30-45d after stroke event) incorporated modified cardiac rehabilitation, traditional neurorehabilitation, and physician follow-up visits to provide risk factor management, rehabilitation psychology, smoking cessation, and nutrition education. Our results demonstrated a statistically and clinically significant improvement in overall function for mobility, self-care, and cognitive tasks at 4 months after stroke after the intervention.²⁰

The primary objective of this study was to evaluate the 1-year long-term functional improvements and the sustainability of this comprehensive SRP administered early in the subacute phase of recovery compared with stroke survivors who received usual care.

Methods

General design

The detailed description of the study’s design, recruitment, methods, and intervention is described in earlier publications.^19,20 The primary results up to the 120-day timepoint have also been published previously.²⁰ This study presents the analysis of follow-up data collected for up to 1 year. Briefly, this is a prospective matched cohort study of acute stroke survivors admitted to an inpatient rehabilitation facility between 2015 and 2020 and followed for 1 year after stroke. Participants were recruited during their inpatient rehabilitation hospitalization and received cardiac clearance before initiating the study intervention as an outpatient. Of the 2261 patients assessed for eligibility, 452 were excluded for not meeting the inclusion/exclusion criteria, and 113 declined to participate. A total of 1696 patients were enrolled in a longitudinal cohort study. Of these, 1323 received only the conventional neurorehabilitation therapies (physical therapy, occupational therapy, and speech therapy as prescribed) (control group), and 373 received conventional neurorehabilitation therapies in addition to the study intervention (stroke recovery program [SRP] group).²⁰ The study intervention consisted of a comprehensive hospital-based outpatient rehabilitation program comprising 36 sessions of a moderate-intensity aerobic exercise program, delivered 2-3 times a week in addition to conventional rehabilitation therapies as prescribed by their physicians. In addition, participants received secondary prevention risk factor education during physician visits.^19,20 All 1696 study participants were offered to enroll in the study intervention; the primary reason for participants’ decision not to engage in the SRP group was the distance to the facility. The groups were then matched for sex, race, age, type of stroke, medical complexity, and functional scores at the time of discharge from the inpatient rehabilitation hospital, to mitigate biases in nonrandomized selection of the sample, resulting in a final sample of 246 participants in the SRP group and 203 of nonparticipants in the control group used in the analysis²⁰ (supplemental fig S1, available online only at http://www.archives-pmr.org/). The Euclidean matching procedure was employed for the matched-pairs strategy.^31,32 A comprehensive overview of the matching algorithm, the variables used for matching, and their respective tolerance levels is detailed in previous studies.^19,20 The inclusion criteria for the study were as follows: individuals must be ≥18 years of age, demonstrate alertness, exhibit the ability to follow simple commands, and provide consent either directly or through a proxy. The exclusion criteria for the study included individuals who were unable to follow simple commands or who exhibited reduced alertness. All participants in the study, or their designated proxies, provided written informed consent before enrollment. The study was approved by the institutional review boards at Hackensack Meridian Health and Rutgers Robert Wood Johnson Medical School.

Study timeline and data collection

Demographic characteristics and medical history were collected from participants’ electronic medical records. Functional assessments were conducted for both groups upon admission to an inpatient rehabilitation hospital, at discharge from the hospital, and at 30, 60, 90, 120, 180, and 365 days after stroke. The primary results of the study evaluating the effects of the intervention at 120 days after stroke, after the conclusion of the program, have been previously published.²⁰ This study analyzed the follow-up data of functional assessments conducted at the 180-day and 365-day marks after stroke. The follow-up data were collected via a phone call from the study participants included in the previous study.²⁰ Each participant received 3 phone calls over a 4-week period. After 3 unsuccessful call attempts, the follow-up process for the participants was considered concluded and marked as a loss to follow-up.

Outcome measures

Functional performance was assessed using the Activity Measure for Post-Acute Care (AM-PAC).^33,34 AM-PAC is a functional outcomes instrument that is designed to track patients’ functional status as they move across the continuum of care. AM-PAC measures activity limitations in 3 distinct functional domains: Basic Mobility (BM), Daily Activities (DA), and Applied Cognitive (AC). The AM-PAC assessment provides a continuous functional score and correlates with real-world abilities categorized in 5 functional stages (see supplemental table S1, available online only at http://www.archives-pmr.org/). The AM-PAC has strong psychometric properties and has been used for longitudinal functional follow-up of stroke survivors for up to 1 year after their inpatient rehabilitation stay, demonstrating its sensitivity to changes over time.^33,35 Recent studies demonstrated the convergent validity of the AM-PAC by showing that its BM, DA, and AC domains align with the same latent traits measured by the Patient-Reported Outcomes Measurement Information System (PROMIS) Physical Function and Cognitive Function scales, allowing for the development of a valid crosswalk between the measures.^36,37 The instrument can be administered by a clinician with the patient or by a proxy.

Statistical analysis

General descriptive statistics were used to characterize the trends for all variables. Continuous variables were summarized with means, SDs, medians, ranges, and interquartile range, as appropriate. Categorical variables were summarized with frequencies and percentages. The AM-PAC subscales were tested for normality using the Shapiro-Wilk test. The final dataset had a 6% rate of missing data across all variables and assessment timepoints, with most of these missing data arising from the 6-month and 1-year follow-up timepoints. An analysis of baseline demographic and clinical variables revealed no significant differences between participants who were lost to follow-up and those who remained in the study at the 1-year follow-up. No imputation strategy for missing data was employed in this study. Analyses included 3 separate mixed effects models regressing each AM-PAC functional domain using linear regression on group (case or control) by follow-up timepoint (admission, discharge, and at 30, 60, 90, 120, 180, and 365d after stroke) with admission values set as a reference and with a random effect for participants to account for repeated measures. Post-hoc analyses were conducted to determine significant differences between groups at each follow-up timepoint. Model residuals were examined for symmetrical distribution around 0. Model results were reported with betas (slopes) and 95% CIs and visualized using boxplots. Effect sizes were calculated for each AM-PAC functional domain at the 1-year timepoint using Cohen’s d with pooled SDs. For the analysis of functional stages at the 1-year follow-up, Fisher Exact tests were conducted to assess the distribution of patients among AM-PAC stages across each functional domain and study group. Mann-Whitney U tests were performed to compare the outcomes of participants who were lost to follow-up with those who remained in the study at the 1-year follow-up. Results with a P value ≤.05 were considered statistically significant. All analyses were conducted using R version 4.3.3.^a

Results

A total of 449 patients with stroke were included in the study. Full baseline demographic and medical characteristics were reported previously²⁰ and presented in supplemental table S2 (available online only at http://www.archives-pmr.org/). Briefly, the entire cohort included 251 (56%) men with a mean age of 68 years. Most participants presented with ischemic stroke, accounting for 94%, whereas 5% experienced hemorrhagic stroke, and 1% had subarachnoid hemorrhage. The most prevalent medical conditions observed in the cohort included hypertension (91%), dyslipidemia (56%), diabetes mellitus (45%), obesity (41%), and coronary artery disease (21%). Of 449 participants, 246 were included in the SRP group and 203 in the control group (usual care group). No significant difference was observed at the baseline regarding demographic or medical history characteristics between participants and controls (see supplemental table S2).

This study evaluated the long-term follow-up outcomes at 6 months and 1 year after stroke. At the 6-month follow-up, a total of 260 patients completed their functional assessments: 168 from the SRP group and 92 from the control group. During the 1-year follow-up, a total of 236 patients were assessed, 169 participants did not answer and return the phone calls (88 from SRP group and 81 from control group), 43 participants declined to complete the assessments (19 from SRP group and 24 from control), and 18 participants were deceased (4 from SRP group and 14 from control group), resulting in a total of with 152 from the SRP group and 84 from the control group, respectively (see table 1). Importantly, an analysis of clinical and demographic variables indicated no significant differences between participants lost to follow-up and those who were retained in the study (see supplemental table S3, available online only at http://www.archives-pmr.org/). Similarly, no significant differences were observed in the functional outcomes at 120-day after stroke between participants lost to follow-up and those who were retained in the study (see supplemental table S4, available online only at http://www.archives-pmr.org/). Furthermore, there was no notable difference in the number of completed sessions between participants who were lost to follow-up and those who remained in the study at the 1-year mark. On average, participants who were lost to follow-up completed 26.1 sessions, whereas those who remained in the study completed 27.3 sessions (P=.336; see supplemental table S5, available online only at http://www.archives-pmr.org/).

Table 1.

Descriptive statistics of the Activity Measure Post-Acute Care (AM-PAC) scores for all functional domains: Basic Mobility, Daily Activity, and Applied Cognitive across all data collection timepoints.

		Basic Mobility		Daily Activity		Applied Cognitive
Time	Measure	Control	SRP Group	Control	SRP Group	Control	SRP Group
Adm	Mean (SD)	35.9 (6.29)	37.5 (4.94)	35.8 (7.40)	36.5 (7.38)	33.9 (14.4)	36.6 (14.6)
	Median (Q1-Q3)	38.4 (34.4-39.1)	39.1 (37.0-39.1)	34.2 (31.0-41.0)	35.1 (32.6-41.0)	33.7 (27.0-42.6)	36.2 (30.0-44.8)
	n	203	246	203	246	203	246
Dis	Mean (SD)	44.6 (11.0)	46.5 (10.2)	43.4 (10.6)	44.7 (10.3)	38.9 (11.9)	40.9 (12.4)
	Median (Q1-Q3)	39.1 (39.1-48.5)	41.5 (39.1-51.7)	41.1 (35.1-49.2)	43.1 (36.0-50.5)	37.0 (31.6-44.8)	40.2 (33.7-51.3)
	n	203	246	203	246	203	246
30 d	Mean (SD)	44.2 (11.0)	54.1 (8.99)	43.1 (10.8)	50.5 (12.9)	38.5 (12.7)	45.5 (10.2)
	Median (Q1-Q3)	39.1 (39.1-48.0)	53.6 (50.1-59.7)	39.9 (35.1-49.0)	46.6 (43.4-55.6)	36.1 (31.4-44.8)	44.7 (37.6-53.4)
	n	203	246	203	246	203	246
60 d	Mean (SD)	49.6 (13.3)	58.2 (8.95)	47.1 (14.8)	54.9 (14.3)	41.9 (12.1)	48.1 (11.0)
	Median (Q1-Q3)	50.9 (39.1-58.3)	57.5 (53.1-64.0)	44.3 (36.0-54.0)	51.3 (45.8-62.9)	40.2 (34.0-50.4)	47.4 (39.9-57.5)
	n	203	246	203	246	203	246
90 d	Mean (SD)	51.1 (14.0)	60.6 (9.12)	48.1 (15.3)	57.6 (16.1)	42.8 (11.3)	48.7 (11.0)
	Median (Q1-Q3)	52.6 (39.1-61.5)	59.7 (54.0-66.8)	45.0 (38.2-54.9)	52.3 (46.1-66.7)	40.7 (34.6-50.9)	48.7 (39.5-57.5)
	n	203	246	203	246	203	246
120 d	Mean (SD)	52.0 (13.5)	61.7 (8.96)	48.3 (14.8)	58.9 (15.1)	42.7 (11.5)	50.0 (10.9)
	Median (Q1-Q3)	52.7 (41.3-61.5)	61.3 (54.8-68.1)	45.8 (38.2-54.1)	56.6 (46.9-69.2)	41.1 (35.1-50.8)	50.9 (41.1-57.5)
	n	203	246	203	246	203	246
180 d	Mean (SD)	54.8 (13.4)	61.9 (10.3)	50.6 (17.3)	58.2 (16.5)	44.9 (12.7)	48.5 (10.8)
	Median (Q1-Q3)	55.7 (51.1-61.7)	61.9 (54.8-68.2)	46.4 (41.4-58.8)	51.8 (46.0-68.9)	43.9 (37.1-52.9)	47.1 (39.8-57.5)
	n	92	168	92	168	92	168
365 d	Mean (SD)	55.1 (15.5)	61.5 (10.7)	52.7 (17.5)	58.0 (17.4)	44.5 (12.4)	49.4 (11.5)
	Median (Q1-Q3)	55.3 (49.7-62.3)	61.9 (55.4-67.0)	50.5 (42.4-63.6)	51.9 (46.2-65.8)	44.0 (36.5-53.5)	49.8 (40.2-57.5)
	n	84	152	84	152	84	152

Abbreviations: Adm, admission; Dis, discharge.

As reported previously,²⁰ participants in the SRP group exhibited significant improvements in their functional abilities compared with the control group at the end of the intervention, that is, 120-day after stroke. In the follow-up assessments conducted at 6 months and 1 year, participants in the SRP group sustained their improvements and exhibited significantly higher scores. The mixed effects linear regression model, which accounted for both time and treatment groups, demonstrated an overall significant interaction between treatment and time in relation to AM-PAC BM and AM-PAC DA scores. Importantly, at both 6 months and 1 year, participants in the SRP group demonstrated significantly higher scores on BM (β=6.14, 95% CI [3.73-8.55], P<.001 and β=4.96, 95% CI [2.47-7.45], P<.001, respectively) and DA (β=8.19, 95% CI [5.10-11.28], P<.001 and β=5.11, 95% CI [1.91-8.30], P=.002, respectively) compared with the control group (see table 2). For AM-PAC AC scores, the treatment group trended toward higher scores at the 6-month follow-up timepoint (β=2.21, 95% CI [–0.40 to 4.83], P=.097), but there was no significant difference at the 1-year follow-up timepoint (β=1.40, 95% CI [–1.31 to 4.10], P=.311; table 2). The wide range of CIs reflects the broad inclusion/exclusion criteria of the study, thus representing patients with various levels of functional deficits. This is particularly evident in the AM-PAC DA measures at the 365-day follow-up. The differences between the mean and median scores, along with a large SD (see table 1), indicate a skewed distribution of functional outcomes and deficits in upper extremity function. Despite the non-normal distribution of these variables, the final statistical model for each AM-PAC subscale was examined, and the residuals displayed a symmetrical distribution around 0 without any discernible patterns (see supplemental fig S2, available online only at http://www.archives-pmr.org/).

Table 2.

Mixed effects linear regression models of AM-PAC scores across all 3 functional domains during the 1-year follow-up time period

	Basic Mobility			Daily Activity			Applied Cognitive
Characteristic	β	95% CI*	P Value	β	95% CI*	P Value	β	95% CI*	P Value
Group	1.7	–0.9 to 3.6	.095	0.7	–1.8 to 3.2	.588	2.7	0.5-4.9	.018
Timepoint
Admission	—	—		—	—		—	—
Discharge	8.8	7.3-10.2	<.001	7.6	5.8-9.5	<.001	5.0	3.4-6.6	<.001
30 d	8.3	6.9-9.8	<.001	7.3	5.5-9.2	<.001	4.6	3.1-6.2	<.001
60 d	13.7	12.3-15.1	<.001	11.3	9.5-13.2	<.001	8.0	6.5-9.6	<.001
90 d	15.3	13.8-16.7	<.001	12.3	10.4-14.2	<.001	8.9	7.3-10.5	<.001
120 d	16.1	14.7-17.6	<.001	12.6	10.7-14.4	<.001	8.8	7.3-10.4	<.001
180 d	18.2	16.3-20.1	<.001	13.8	11.4-16.3	<.001	10.2	8.1-12.2	<.001
365 d	18.7	16.7-20.6	<.001	16.5	14.0-19.0	<.001	11.0	8.8-13.1	<.001
Group by timepoint
Group: discharge	0.2	–1.8 to 2.2	.848	0.6	–1.9 to 3.1	.645	–0.7	–2.8 to 1.5	.544
Group: 30 d	8.2	6.2-10.2	<.001	6.7	4.1-9.2	<.001	4.3	2.1-6.4	<.001
Group: 60 d	7.0	5.0-9.0	<.001	7.1	4.6-9.7	<.001	3.5	1.3-5.6	.001
Group: 90 d	7.9	5.9-9.8	<.001	8.9	6.3-11.4	<.001	3.2	1.1-5.4	.003
Group: 120 d	8.1	6.1-10.0	<.001	9.9	7.3-12.4	<.001	4.5	2.4-6.7	<.001
Group: 180 d	6.1	3.7-8.6	<.001	8.2	5.1-11.3	<.001	2.2	–0.4 to 4.8	.097
Group: 365 d	5.0	2.5-7.5	<.001	5.1	1.9-8.3	.002	1.4	–1.3 to 4.1	.311

NOTE. Admission scores were used as a reference point.

Beta represents the slope of the model, indicating the direction and magnitude of differences in the AM-PAC scores.

Figure 1 illustrates the trajectory of changes in functional scores across all 3 functional domains from baseline to 120 days, 6 months, and 1 year after stroke. The Shapiro-Wilk test indicated that all AM-PAC subscales were non-normally distributed; therefore, the data are presented as boxplots. The median semi-interquartile range (±SIQR) score at 1 year for AM-PAC-BM was 61.88±6.30 for the SRP group compared with 55.31±6.30 in the control group (P<.001), 51.88±9.81 vs 50.47±10.59 for the AM-PAC DA score (P=.032), and 49.77±8.67 vs 43.96±8.51 for the AM-PAC AC score (P=.225), respectively (see fig 1, table 1). In addition, effect sizes calculated for AM-PAC BM, AM-PAC DA, and AM-PAC AC at 1-year follow-up timepoint were 0.51, 95% CI [0.24-0.78], 0.30, 95% CI [0.03-0.57], and 0.41, 95% CI [0.14-0.68], respectively.

Fig 1.

Comparison of the changes in Activity Measure Post-Acute Care (AM-PAC) scores during the 1-year follow-up period for both the SRP and Control groups.

The AM-PAC measure serves as a valuable tool for assessing functional performance, either as a continuous score to monitor changes over time or by categorizing scores into 5 functional stages in accordance with predetermined criteria^33,34 (see supplemental table S1). These stages are organized hierarchically to reflect different levels of performance within each functional domain. Figure 2 demonstrates the distribution of participants across each functional stage at the 1-year follow-up. The analysis revealed statistically significant differences between participants in the SRP group and those in the control group across stages in BM, DA, and AC at the 1-year follow-up (P=.009, P=.026, P=.023, respectively, Fisher Exact test, table 3). For example, the percentage of SRP participants classified in stage 4 for the BM domain, indicative of the ability to “move around outdoors,” was 25%; in contrast, only 14% of control group participants were in this stage. On the other hand, the percentage of control participants in stage 2, which signifies “limited moving indoors,” was 25%, compared with just 12% of SRP participants (see table 3). Similar patterns were observed in “daily activities” and “applied cognitive” functioning, with a higher proportion of SRP participants reaching advanced functional stages such as independence with self-care and achieving multistep complex tasks (see table 3).

Fig 2.

Comparison of the proportion of patients across each of the 5 Activity Measure Post-Acute Care (AM-PAC) functional stages at the 1-year follow-up for both the SRP and Control groups.

Table 3.

Analysis of the proportions of patients at the 1-year follow-up across the 5 Activity Measure Post-Acute Care (AM-PAC) functional stages for both the SRP and Control groups

AM-PAC	Control n (%)	SRP Group n (%)	P Value*
Basic mobility			.009
Stage 1	5 (6.0)	2 (1.3)
Stage 2	21 (25.0)	18 (11.8)
Stage 3	44 (52.4)	91 (59.9)
Stage 4	12 (14.3)	38 (25.0)
Stage 5	2 (2.4)	3 (2.0)
Daily activity			.026
Stage 1	17 (20.2)	11 (7.2)
Stage 2	30 (35.7)	69 (45.4)
Stage 3	15 (17.9)	19 (12.5)
Stage 4	18 (21.4)	41 (27.0)
Stage 5	4 (4.8)	12 (7.9)
Applied cognitive			.022
Stage 1	4 (4.8)	1 (0.7)
Stage 2	12 (14.3)	10 (6.6)
Stage 3	21 (25.0)	39 (25.7)
Stage 4	32 (38.1)	55 (36.2)
Stage 5	15 (17.9)	47 (30.9)

^⁎Refers to results of Fisher Exact test.

Discussion

The results of this study show that an early outpatient comprehensive SRP including aerobic exercise and stroke risk factor education within the first 6 months of stroke was associated with functional benefits that were maintained for at least 1 year after stroke. Significant improvements in mobility and self-care were observed lasting well beyond the completion of the intervention period, measured at 1 year after stroke. Notably, all 3 functional domains demonstrated a medium effect size. The program was administered early in the subacute phase, during the critical time window after a stroke, which is known to have the most benefit to impact neuroplasticity.^38,39 Although previous results have shown an improvement in function after intervention,²⁰ there has been no in-depth investigation about the long-term sustained effects of a SRP including aerobic exercise and stroke risk factor education delivered in the subacute phase after stroke.

Other programs incorporating aerobic exercise with risk factor education have shown improvements after intervention in multiple aspects of cardiovascular health,^23,40 for example, improvement in cardiovascular capacity,³⁶ stroke vascular risk factors,^22,41,42 multiple aspects of the cardiometabolic profile, such as waist circumference, and BMI.42, 43, 44 Additional benefits beyond the cardiovascular system for stroke survivors include improvements in health-related quality of life, cognitive function, and depression.45, 46, 47

Importantly, several studies have reported that exercise induces improvements in motor function as a result of the facilitation of neuroplasticity after stroke.48, 49, 50, 51 Studies have demonstrated that engaging in long-term aerobic exercise induces significant changes in the brain repair and recovery processes such as neurogenesis, synaptogenesis, and angiogenesis as well as increasing the levels of brain-derived neurotrophic factor, insulin-like growth factor, and vascular endothelial growth factors.⁵² These long-term outcomes sustained for 1 year are likely attributable to the substantial benefits to both the cardiovascular and neurovascular systems.^20,41,53,54

Although our results showed statistically significant improvements with the AM-PAC BM and DA scores at 1-year follow-up, analysis of the AM-PAC AC scores did not reach statistical significance at the 1-year timepoint (table 1). These results emphasize the existing literature suggesting that patients with stroke may require long-term interventions for sustained speech and cognitive improvements.55, 56, 57 Nevertheless, the AM-PAC functional stage designation for the AC domain gives a better understanding of how the group as a whole was functioning. Our findings indicate that participants in the SRP group (36.2%+30.9%=67.1%) were at higher functioning stages (stage 4 and 5) compared with the control group (38.1%+17.9%=56%). Future studies examining cognitive function with different outcome measures that are sensitive to cognitive changes may help elucidate the impact of aerobic exercise on long-term cognitive function.

It is important to note that the intervention was accepted and well tolerated by patients with stroke in their subacute phase of recovery. The completion rates demonstrated that most (over 63%) of the patients were able to complete at least 27 sessions of the intervention.²⁰ Importantly, there were no adverse events or injuries reported in relation to the intervention.^19,20 This evidence highlights that patients with stroke are able to participate in a structured aerobic exercise program during the subacute phase of recovery from stroke. Notably, the completion rates reported in our previous studies for patients with stroke are significantly higher than completion rates reported for traditional cardiac rehabilitation programs (30%-52%) for patients with cardiac disease.^58,59

In addition to aerobic exercise, the intervention included education for risk factor modification and fostering the behavioral lifestyle changes for long-lasting risk factor management after stroke. Having more confidence participating in aerobic exercise for participants in the SRP group may potentially lead to a higher level of physical activity, which could account for the sustained functional improvement observed at 1 year. Our results showed that a higher number of SRP participants were able to ambulate outdoors, perform more advanced self-care tasks, and showed higher cognitive/communication skills compared with usual care (control group) alone (table 3). Future studies should examine the physical activity levels of chronic stroke survivors and their impact on long-term functional status.

Study limitations

There are several limitations to this study. As in any follow-up study, data acquisition was challenged by the loss of follow-up of the participants. In addition, activity levels after discharge from the program for individual study participants were not quantified and included in the analysis. The level of physical activity and functional ability of each participant may play a role in their 1-year functional status and overall recovery. Another limitation of this study is that it was not conducted as a randomized clinical trial. The original study was designed as a longitudinal cohort study, as it was the first of its kind for patients with stroke in the subacute phase, to allow inclusion of a large number of patients with broad inclusion and exclusion criteria to examine the feasibility, safety, and impact of this program. Future randomized controlled trials are needed to confirm the findings of this study. Nevertheless, the results of this current cohort study present evidence that the comprehensive SRP, including aerobic exercise and risk factor management, has sustained effects on functional recovery after stroke. Future large randomized controlled trials are warranted to solidify these findings, mitigate potential bias introduced by the study design, and account for participant activity levels outside the study intervention.

Conclusions

An early outpatient comprehensive SRP including aerobic exercise, and stroke risk factor education within the first 6 months of stroke not only demonstrated functional improvements after the program, but was able to achieve sustained functional improvements in mobility, self-care, and cognitive ability at 1 year. Improving long-term function in stroke survivors may help increase their level of independence, thereby giving them a better quality of life, decreasing caregiver burden, and decreasing the cost burden on an already challenged health care system.

Supplier

a. R, version 4.3; R Core Team.

Disclosure

The investigators have no financial or nonfinancial disclosures to make in relation to this project.