- Email: [email protected]
Combined Aerobic and Resistance Training for Cardiorespiratory Fitness, Muscle Strength, and Walking Capacity after Stroke: A Systematic Review and Meta-Analysis
ARTICLE IN PRESS
Combined Aerobic and Resistance Training for Cardiorespiratory Fitness, Muscle Strength, and Walking Capacity after Stroke: A Systematic Review and Meta-Analysis Junghoon Lee, MS, and Audrey J. Stone, PhD
Background: Cardiorespiratory fitness, measured as peak oxygen consumption, is a potent predictor of stroke risk. Muscle weakness is the most prominent impairment after stroke and is directly associated with reduced walking capacity. There is a lack of recommendations for optimal combined aerobic training and resistance training for those patients. The purpose of this study was to systematically review and quantify the effects of exercise training on cardiorespiratory fitness, muscle strength, and walking capacity after stroke. Methods: Five electronic databases were searched (until May 2019) for studies that met the following criteria: (1) adult humans with a history of stroke who ambulate independently; (2) structured exercise intervention based on combined aerobic training and resistance training; and (3) measured cardiorespiratory fitness, muscle strength, and/or walking capacity. Results: Eighteen studies (602 participants, average age 62 years) met the inclusion criteria. Exercise training significantly improved all 3 outcomes. In subgroup analyses for cardiorespiratory fitness, longer training duration was significantly associated with larger effect size. Likewise, for muscle strength, moderate weekly frequency and lower training volume were significantly associated with larger effect size. Furthermore, in walking capacity, moderate weekly frequency and longer training duration were significantly associated with larger effect size. Conclusions: These results suggest that an exercise program consisting of moderateintensity, 3 days per week, for 20 weeks should be considered for greater effect on cardiorespiratory fitness, muscle strength, and walking capacity in stroke patients. Key Words: Cardiac rehabilitation—exercise—peak oxygen consumption— skeletal muscle—gait © 2019 Elsevier Inc. All rights reserved.
Introduction Cardiorespiratory fitness (CRF) is one the strongest predictors of stroke risk. Measured as peak oxygen consumption (VO2peak), CRF was inversely associated with a 2.30-fold risk for any type of stroke.1 Compared to agematched sedentary groups, individuals with stroke have about 50% decreased CFR2 and are at the risk of From the Department of Kinesiology and Health Education, The University of Texas at Austin, Austin, Texas. Received July 10, 2019; revision received October 11, 2019; accepted October 23, 2019. Financial Disclosure: There is no received funding for this study. Address correspondence to Audrey J. Stone, PhD, Department of Kinesiology and Health Education, The University of Texas at Austin, 2109 San Jacinto Blvd, Stop D3700, Austin, TX 78712. E-mail: [email protected]. 1052-3057/$ - see front matter © 2019 Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.jstrokecerebrovasdis.2019.104498
cardiovascular events by over 25%.3 Muscle weakness is the most prominent impairment after stroke,4 directly associated with reduced walking speed and endurance.5 Hemiparesis, affecting 65% of stroke victims 3, can double physiological energy for normal walking, compared to that of healthy persons.6 In this regard, exercise training (ET) or physical activity is required for the rehabilitation of stroke populations. Physical activity is defined as any bodily movement produced by skeletal muscle that substantially increases energy expenditure.7 Exercise is a subset of physical activity that is planned, structured, and repetitive and is performed deliberately for the purpose of improving or maintain physical fitness.8 The participation of physical activity should be fundamentally encouraged because low levels of physical inactivity is predominant in stroke survivors, which increases cardiovascular risk.9 In particular, well-designed exercise programs are necessary to
Journal of Stroke and Cerebrovascular Diseases, Vol. &&, No. && (&&), 2019: 104498
1
ARTICLE IN PRESS J. LEE AND A.J. STONE
2
facilitate their participation in physical activity and more effectively improve CFR, muscle strength, and walking capacity as mentioned above. A recent meta-analysis study found that high-intensity aerobic training (AT) increases VO2peak further than AT with low-intensity, but walking capacity, measured as walking speed or distance, is not improved in ambulatory persons with stroke.10 Another meta-analysis showed that resistance training (RT) increases muscle strength but not walking distance, measured using the 6 minutes walking test (6MWT).11 Meanwhile, regarding non-ambulatory stroke survivors, one meta-analysis found that ET increases both VO2peak and walking distance.12 However, these previous meta-analyses only included a small number of studies and did not provide appropriate exercise guidelines. Although both AT and RT are essential for stroke patients, no meta-analysis has investigated the effects of combined AT and RT on the key elements of physical fitness: CFR, muscle strength, and walking capacity. Billinger et al.7 provided comprehensive physical activity and exercise recommendation for stroke survivors based on systematic literature reviews. However, structured exercise intervention with detailed demographic and training factors are needed to be statistically reviewed by a meta-analysis combining results from comparable studies which enhance the validity and reliability of conclusions. For the improvement in all 3 factors, CFR, muscle strength, walking capacity, combined AT and RT is ideal compared to any singular exercise modality. Therefore, the primary purpose of this study was to evaluate the effects of combined AT and RT on CFR, muscle strength, and walking capacity in ambulatory persons with stroke by conducting a systematic review and metaanalysis. This review carefully examined the specific variables of the ET regimens (i.e., intensity, duration, frequency, and volume), identified different types of stroke populations (i.e., age, gender, body mass index (BMI), stroke type, and post-stroke period), and reported adverse events to provide sufficient evidence for establishing optimal exercise guidelines. Moreover, this study also attempted to reveal current issues in this field and inspire future research to resolve health issues associated with stroke.
Materials and Methods This current systematic review followed the strategy of the PRISMA statement.13 Ethics committee approval was not sought for the present study because this meta-analysis study was based on the results of previously published studies.
Data Sources Five electronic databases (CINAHL, EMBASE, PubMed, SportDiscus, and Web of Science) were searched for eligible studies published in English from the earliest date available to May 2019. The following keywords were used for searches: ‘exercise training or physical therapy’, ‘cardiorespiratory fitness or aerobic capacity or muscle strength or walk’ and
‘stroke’. Manual searches of reference lists were conducted to ensure all relevant studies were captured. One reviewer (JH) searched all articles and applied the inclusion and exclusion criteria to the titles and abstracts searched. When the information was not clear, the full text papers of the studies were obtained for review. Corresponding authors of potentially eligible studies were contacted if studies reported data for which it was impossible to discriminate.
Study Selection The inclusion criteria for eligible studies were as follows: (1) adult humans with a history of stroke who ambulate independently with or without an assistive device; (2) structured exercise intervention based on combined AT and RT; and (3) measured CFR, muscle strength, and/or walking capacity. Studies including additional interventions such as dietary supplements were excluded to focus on the effects of ET alone. Duplicate studies or sub-studies of included trials were also excluded from the analysis.
Quality Assessment One reviewer assessed the quality of the included studies using the PRISMA recommendations.13 The quality assessment consisted of 6 items: (1) appropriate generation of random allocation sequence; (2) concealment of the allocation sequence; (3) blinding of the assessment and collection outcomes; (4) proportion of participants lost to follow-up; (5) complete outcome data; (6) the intention-to-treat principle.13
Data Extraction Data were extracted from all selected studies to record the detailed information in terms of participant characteristics, study methods, interventions, outcomes, and adverse events. We used means and standard deviation (SD); where standard errors or 95% confidence interval (CI) were provided, they were converted to SD. Population characteristics, such as age, gender, BMI, stroke type, post-stroke period, and the number of participants were recorded to compare the similarity of participants between trials. The primary outcomes were CFR (VO2peak), muscle strength (one-repetition maximum (1-RM) or dynamometry), and walking capacity (6WMT or gait speed). If muscle strength was reported as upper and lower body, the lower body or affected leg was used. Interventions including total duration, frequency (days per week), intensity, session duration, repetition, set, rest periods, type and order of exercise, names of exercise machine or tool, supervisors, places of intervention, and whether or not RT was performed to muscular fatigue were recorded to compare the similarity of training methods between trials. Measurement technique and region were also extracted. The median values were used for calculation if the studies reported a range of data (e.g., 16 for 15-17 repetitions). Detailed interventions about control groups (CON) and any additional interventions were recorded.
ARTICLE IN PRESS EXERCISE TRAINING FOR STROKE SURVIVORS
Data Analysis Heterogeneity between studies were assessed using the Cochran Q statistic14 and the I2 test.15 I2 ranges from 0%100%: a value less than 25% indicates a low risk of heterogeneity, 25%-75 % indicates a moderate risk of heterogeneity, and greater than 75% indicates a high risk of heterogeneity. In each study, the effect size (ES) for the intervention was calculated by the difference between the means of the postmeasurement and premeasurement at the end of the intervention using Cohen's definition. Separate meta-analyses of trials with CFR, muscle strength, and walking capacity were performed to generate the mean ES and 95% CI. ESs were classified according to Cohen’s definition (1988), where .2 is considered small, .5 moderate, and .8 large.16 This study used a fixed-effects model when homogeneity was verified or a random-effects model when heterogeneity was shown by the Q statistic.15 Publication bias was assessed using the Egger’s regression test.17 To evaluate whether an individual cohort had undue influence on the overall meta-analysis result, sensitivity analyses were performed in all 3 outcomes by omitting one of the trials at a time and determining whether statistical conclusion remained the same. All calculations were conducted with SPSS version 20, Microsoft Excel 2016, and STATA version 14.2. Subgroups analyses were performed where sufficient numbers of trials existed in subgroups in order to identify
3
potential factors influencing the effect of exercise on outcomes variables. These subgroup analyses also accounted for heterogeneity between studies: (1) age less than 65 versus age greater than or equal to 65; (2) less than 2 post-stroke years versus greater than or equal to 2 and less than 4 poststroke years versus greater than or equal to 4 post-stroke years; (3) less than or equal to 12 weeks versus greater than 12 weeks; (4) 2 days per week versus 3 days per week versus greater than or equal to 4 days per week; (5) moderate-intensity (AT: 40%-60% of heart rate reserve (HRR), RT: 50%-70% of 1RM) versus high-intensity (AT: 60%-85% of HRR, RT: 70%-90% of 1RM); (6) less than 50 sets per week versus greater than or equal to 50 sets per week; (7) less than 500 repetitions per week versus greater than or equal to 500 repetitions per week; and (8) supervised versus unsupervised training. Random effects meta-analysis regression was conducted to compare the effect estimates (ES) in different subgroups by considering the meta-analysis results from each subgroup separately. To interpret the results of subgroup analyses, P value (P < .05) between study variation was considered for the statistical difference between subgroups.
Results Study Selection and Characteristics The search resulted in 2817 potential studies (Fig 1). From the titles and abstracts, 2742 studies were excluded
Figure 1. Study search and selection process.
ARTICLE IN PRESS J. LEE AND A.J. STONE
4
based on the inclusion criteria, and then 75 full text studies were reviewed. Of these, 18 articles met the criteria. In selected studies, 3 AT groups18,19 and one Chinese martial art group20 were excluded from this analysis. One CON conducting combined AT and RT21 was included in exercise group (EX). Consequently, nineteen ET cohorts in eighteen studies were selected.
Participants Table 1 shows the characteristics of all of the studies included. Articles were published from April 199922 to March 2018.18 A total of 602 participants completed their interventions (EX: 526, CON: 157, female%: 37.8%) ranging from 923 to 12024 participants. The average age of the participants was 62.1 § 10.2 years (EX: 61.2 § 10.23, CON: 64.8 § 10.0 years). Average post-stroke period was 2.4 § 3.3 years (EX: 2.5 § 3.5, CON: 2.2 § 2.7 years).
Interventions All interventions, except for 3 studies implementing unsupervised intervention,18,24,25 were supervised by qualified trainers or researchers in research/community centers or participants’ homes. The mean training period was 15 weeks (minimum-maximum: 421-2418,24-27 weeks). The mean training frequency was 3 days per week (minimum-maximum: 226,28-518,24 days). The intensities of AT were expressed as a percent of maximum heart rate (HRmax), HRR or VO2peak or the Borg rating of perceived exertion (RPE) or revolution per minute (rpm). The intensities ranged from low (40 rpm29) to high (80% of HRmax30). The intensities of RT were expressed as a percentage of 1-RM or 10-RM (a workload which enables the participant to complete 10 repetitions) or RPE. The intensities ranged from low (50%-60% of 1RM24,25) to high (80% of 1RM22,31). All trials, with the exception of 5 studies,19-21,26,27 progressively increased their session duration or intensity over the duration of the intervention on the basis of the progression of aerobic capacity and muscle strength tested at baseline and/or mid-study.
Measurements Of all 18 studies, 9 assessed CFR by measuring VO2peak by the graded exercise test18,19,23-25,27,29,32,33 and 1 used estimated VO 2peak by the submaximal stationary ergometer test.30 For the assessment of muscle strength, 6 studies measured maximum weight moved.18,20,25,26,33,34 and 5 measured maximal voluntary isometric contraction. 22,24,29,31,32 For walking capacity, 11 studies measured 6MWT,18,21,23-30,34 1 measured 12MWT,35 and 3 used gait speed.20,22,31
Effect of Exercise Training Cardiorespiratory Fitness ET significantly increased CFR in ten trials (mean ES = .41, 95% CI = .25 to .56, P < .0001) (Fig 2). The absolute increase of CFR was 12%. Univariate meta-regression did not show heterogeneity between studies (Q = 4.29, df = 9, P = .891, I2 = .0%). In subgroup analyses, younger subgroups aged less than 65 years resulted in a significantly greater improvement of CFR (P < .0001) than elderly subgroups aged greater than or equal to 65 years (P = .06). Subgroups with less than 2 post-stroke years resulted in a significantly greater CFR (P < .0001) than subgroups with greater than or equal to 2 and less than 4 post-stroke years (P = .09). Subgroups training for greater than 12 weeks resulted in a significantly greater CFR (P < .0001) than training for less than or equal to 12 weeks (P = .10). There was no significant difference in effect between subgroups training 3 days per week and greater than or equal to 4 days per week (P = .75), with moderate-intensity (40%-60% of HRR) and high-intensity (60%- 85% of HRR) (P = .34), and with supervision and nonsupervision (P = .29).
Muscle Strength ET significantly increased muscle strength in 11 trials (mean ES = .59, 95% CI = .32 to .86, P < .0001) (Fig 3). The absolute increase of muscle strength was 33%. Univariate meta-regression showed moderate heterogeneity between studies (Q = 29.83, df = 10, P = .001, I2 = 66.5%). In subgroup analyses, there was no significant difference in effect between subgroups aged less than 65 years and aged greater than or equal to 65 years (P = .84). Subgroups with less than 2 post-stroke years resulted in a significantly greater improvement of muscle strength (P < .0001) than subgroups with greater than or equal to 4 post-stroke years (P = .14). Subgroups training for less than or equal to 12 weeks resulted in a significantly greater muscle strength (P < .0001) than training for greater than 12 weeks (P < .01). Subgroups training 3 days per week resulted in a significantly greater muscle strength (P < .01) than training greater than or equal to 4 days per week (P = .10). There was no significant difference in effect between subgroups training with moderate-intensity (50%-70% of 1RM) and high-intensity (70%-90% of 1RM) (P = .71). Subgroups training with lower volume (<50 sets and <500 repetitions per week) resulted in a significantly greater muscle strength (P < .05) than training with higher volume (50 sets and 500 repetitions per week) (P = .27). Subgroups training with supervision resulted in significantly greater muscle strength (P < .001) than training with non-supervision (P < .05).
Walking Capacity ET significantly increased walking capacity in fifteen trials (mean ES = .45, 95% CI = .25 to 0.65, P < .0001) (Fig 4).
Study
Number (%female)
ET
CON
ET
CON ET
CON
Carr and Jones (2003)
30-82
N/A
20
N/A
N/A
N/A
VO2peak, knee isometric extension
16
3
N/A
8ⅹ2ⅹ10
Duncan et al. (2003)
68.5§ 9.0
70.2§ 44 11.40
48
.2
.2
13
3
30
7ⅹ2ⅹ10
Eng et al. (2003)
63.2§ 8.5
N/A
25
N/A
4.2§ N/A 2.9
VO2peak, knee isometric extension, 6 MWT 12 MWT
8
3
60
9ⅹN/AⅹN/A
Jørgensen et al. (2010)
60.4§ 5.7
N/A
14
N/A
2.1§ N/A 1.9
estimated VO2peak, 6MWT
12
3
N/A
4ⅹ3-5ⅹ6-12
fairly light to somewhat hard on RPE 80% of HRmax
Kim et al. (2014) Kluding et al. (2011) Lee et al. (2015)
53.95.82 54.1§ 20 11.7 63.7§ N/A 9 9.1 64.0§ 63.0§ 14 7.4 5.5
N/A
1.1§ 1.2§ 6MWT 0.2 0.3 3.2§ N/A VO2peak, 0.8 6MWT 6.0§ 5.8§ 6MWT, grip 3.3 2.5 strength
4
3
50
N/A
12
3
30
4ⅹ1ⅹ10
16
3
20
Marzolini et al. (2012)
63.6§ 13.5
N/A
41
N/A
1.4§ N/A 2.6
24
3
N/A
14ⅹ2-3ⅹ10-15 1-8wks: 50-60% light to hard Supervised on RPE 9-16wks: 60%-70% of HRR N/Aⅹ 70% of HRR 50%-60% Unsupervised N/Aⅹ10-15 of 1RM
Marzolini et al. (2014)
63.8§ 12.7
N/A
120 N/A
2.0§ N/A 3.3
24
5
20-60
10ⅹ N/Aⅹ10-15
N/A
33
24
5
20-60
11ⅹ1-2ⅹ10-15
N/A 12
N/A
Post-stroke years
N/A
Outcomes
Intervention Duration Frequency AT Volume RT Volume (A ⅹ B ⅹ C) (weeks) (week-1) (min)
VO2peak, leg extension, 6MWT VO2peak, knee isometric extension, 6MWT
AT Intensity
RT Intensity
Supervised or Unsupervised
1-5wks: 40-50% 6-10wks: 50-60% 11-16wks: 60%-70% of original test wattage 40 rpm
N/A
Supervised
N/A
Supervised
N/A
Supervised
Supervised
N/A
the highest possible training intensity N/A
Supervised
50% of VO2peak
N/A
Supervised
40%-70% of VO2peak
50-60% of 1RM
Unsupervised
Unsupervised 5
(Continued)
ARTICLE IN PRESS
Age (years)
EXERCISE TRAINING FOR STROKE SURVIVORS
Table 1. Summary of included studies
6
Table 1 (Continued) Study
Age (years)
Number (%female)
ET
ET
61.7§ 10.0
Rand et al. (2010) Rimmer et al. (2000) Rimmer et al. (2009) Tang et al. (2010) Taylor-Piliae et al. (2014) Teixeira-Salmela et al. (1999)
67.0§ 10.8 53.2§ 8.3 63.7§ 9.1 64.5§ 12.2 69.6§ 9.40 67.7§ 9.2
CON ET
CON
1.2§ 1.3
N/A
11
VO2peak,knee isometric extension, 6MWT 6 MWT, Knee strength VO2peak, leg press VO2peak
53.2§ 35 8.3 N/A 18
35 N/A
N/A
N/A
N/A
2.5§ N/A VO2peak, 2.3 6 MWT 2.8§ 3.2§ gait speed, 4.9 4.0 leg strength 7.6§ N/A gait speed, 9.4 knee isometric extension
24
2
20
N/A
12
3
30
8ⅹ1ⅹ15-20
14
3
N/A
24
2-3
30
12
3
N/A
10
3
20
7ⅹ3ⅹ10
7ⅹ3ⅹ10
68.2§ 44 10.3 N/A 13
N/A
Teixeira-Salmela 67.7§ et al. (2001) 9.2
N/A
N/A
7.7§ N/A 9.4
knee isometric extension, gait speed
10
3
20
ToledanoZarhi et al. (2011)
65.0§ 14 12.0
14
N/A
6MWT
6
2
35-55
65.0§ 10.0
13
48
N/A
Intervention Duration Frequency AT Volume RT Volume (weeks) (week-1) (min) (A ⅹ B ⅹ C)
4.4§ N/A 2.0 N/A N/A
38
N/A
Outcomes
N/A
N/A 9ⅹ2ⅹ15 N/A
N/A
AT Intensity
RT Intensity
60%-80% of VO2peak
70% of 1RM
somewhat hard on RPE 50%-70% of HRR N/A
N/A
Supervised
70% of 10RM N/A
Supervised
60%-80% of HRR N/A
60% of 1RM N/A
80% 1-5wks: of 1RM 50%-70% 6-10wks: 70% of aerobic working capacity 1-5wks: 50-70% 80% of 1RM 6-10wks: 70% of aerobic working capacity 50%-70% N/A of HRmax
Supervised or Unsupervised
ARTICLE IN PRESS
Marzolini et al. (2018)
CON
Post-stroke years
Supervised Supervised Supervised Supervised
Supervised
Supervised
J. LEE AND A.J. STONE
Abbreviations: A ⅹ B x C, number of exercise x sets x repetitions; AT, aerobic exercise training; CON, control group; ET, exercise training group; N/A, not available; RM, repetition maximum; RT, resistance exercise training; VO2peak, peak oxygen consumption; 6MWT, six minutes walking test; 12MWT, twelve minutes walking test. Values are means § SD.
ARTICLE IN PRESS EXERCISE TRAINING FOR STROKE SURVIVORS
7
Figure 2. Forest plot of effect sizes and 95% confidence intervals for 10 cohorts representing cardiorespiratory fitness, based on the fixed effects meta-analysis results.
The absolute increase of walking capacity was 14%. Univariate meta-regression showed moderate heterogeneity between studies (Q = 23.82, df = 14, P = .048, I2 = 41.2%). In subgroup analyses, there was no significant difference in effect between subgroups aged less than 65 years and aged greater than or equal to 65 years (P = .52). Subgroups with greater than or equal to 4 post-stroke years resulted in a significantly greater improvement in walking capacity (P < .01) compared to subgroups with greater than or equal to 2 and less than 4 post-stroke years (P = .09). Subgroups training for greater than 12 weeks resulted in a significantly greater walking capacity (P < .001) than training for less than or equal to 12 weeks (P < .05). Subgroups training 3 days per week resulted in a significantly greater walking capacity (P < .0001) than training 2 days (P = .28) and greater than or equal to 4 days per week (P = .17). Significant difference in effect was not found between subgroups training with moderate-intensity and high-intensity (P = .60). Subgroups training with supervision resulted in a significantly greater walking capacity (P < .001) than training with nonsupervision (P = .06).
Quality Assessment and Potential Bias In the quality assessment, 89% reported appropriate generation of a random allocation sequence (16 of 18),
33% presented concealment of the allocation sequence (6 of 18), 56% described blinding of the assessment and collection outcomes (10 of 18), 94% explained proportion of participants lost to follow-up (17 of 18), 100% exhibited complete outcome data (18 of 18), and 39% reported that the intention-to-treat principle was used for statistical analyses (7 of 18). Egger’s test showed no significant publication bias for all CFR, muscle strength, and walking capacity (P = .94, P = .98, and P = .19 respectively) (Fig 5).
Sensitivity Analysis Sensitivity analysis reported that by excluding any of all cohorts from the meta-analysis, the estimated effects would still be within the 95% CI of the mean ES in all three outcomes, suggesting the results of the meta-analysis would not significantly change after the removal of any one cohort.
Adverse Events The presence or absence of adverse events was recorded in 7 of the 18 studies. Five studies reported there were no adverse events. One study reported 2 participants experienced complications during the exercise testing.33 Another study reported 3 participants had diagnoses of second strokes, but there were no deaths or heart attacks.29
ARTICLE IN PRESS J. LEE AND A.J. STONE
8
Figure 3. Forest plot of effect sizes and 95% confidence intervals for 11 cohorts representing muscle strength, based on the random effects meta-analysis results.
Discussion The primary results of this meta-analysis study were that ET significantly increased CRF, muscle strength, and walking capacity by 12%, 32%, and 14%, respectively, in ambulatory persons with stroke of an average age of 61.2 years. Moderate heterogeneity was found between studies on muscle strength and walking capacity except for CFR. In subgroup analyses for CFR, longer training duration was significantly associated with larger ES. For muscle strength, moderate weekly frequency and lower training volume were significantly associated with larger ES. For walking capacity, moderate weekly frequency and longer training duration were significantly associated with larger ES. Interestingly, subgroups training for middle/long-term (>12 weeks) resulted in significantly greater effects of ET both on CRF and walking capacity, whereas muscle strength in short-term (12 weeks) training of subgroups increased significantly more than middle/long-term training. This may be a result of physiological adaptations in muscular responses to exercise, which can diminish in middle/longterm training.36 These results suggest that middle/longterm is an essential training element for improving both CRF and walking capacity with ET, but it is necessary to change variables of RT during mid/long-term to consistently obtain ET-induced strength gains.
In terms of CRF, there was no significant difference in effect between subgroups training 3 days per week and greater than or equal to 4 days per week, and training with moderate and high intensity. Subgroup analyses were not conducted based on training volume, as most studies conducted AT with 20-30 minutes of session duration. Thus, 3 days per week and moderate-intensity of AT for relatively short session duration combined with RT can be appropriate to improve CRF in stroke patients. These results seem to contradict the results of a meta-analysis37 suggesting that high weekly volume of treadmill ET were significantly associated with greater improvement in VO2peak. In this study, supplementing RT to AT might contribute to increasing CRF by enhanced muscle strength. Marzolini et al.18 found that combined AT and RT resulted in similar VO2peak, but a 2-fold improvement in VO2 at the ventilatory threshold as well as greater muscle strength and mass gains compared with AT alone. Accordingly, moderate weekly training frequency and volume of AT combined with RT could be sufficient for significant improvement of CRF in this study. Contrary to expectations, training with lower volume (<50 sets and <500 repetitions per week) significantly increased muscle strength, whereas subgroups training with higher volume (50 sets and 500 repetitions per week) had no significant effects on strength. Similarly,
ARTICLE IN PRESS EXERCISE TRAINING FOR STROKE SURVIVORS
9
Figure 4. Forest plot of effect sizes and 95% confidence intervals for 15 cohorts representing walking capacity, based on the random effects meta-analysis results.
subgroups training 3 days per week resulted in significantly greater muscle strength than training greater than or equal to 4 days per week. No significant differences between subgroups training with moderate-intensity (50%-70% of 1RM) and high-intensity (70%-90% of 1RM) were identified. As already mentioned, short-term training (12 weeks) was associated with greater strength gains. Consequently, it is possible to suggest that RT 3 days per week with moderate-intensity and lower training volume for short-term combined with AT can induce
significant strength gains in stroke patients. On the other hand, a meta-analysis study investigating the effects of RT alone on strength after stroke reported that moderatehigh intensity and long duration is required for strength gains.11 A possible interpretation for the discrepancy is adding AT on RT as well as the different characteristics of participants (such as age, BMI, post-stroke period, and levels of baseline strength), might influence the change of strength with ET. Future research is warranted to ascertain the influence of these variables on muscle strength.
Figure 5. Funnel plots of publication bias in all 3 outcomes. Abbreviations: SE, standard error; SMD, standardized mean difference.
ARTICLE IN PRESS J. LEE AND A.J. STONE
10
Walking capacity significantly increased in subgroups training 3 days per week, whereas there was no significant effect in subgroups training 2 or greater than or equal to 4 days per week. There was no significant difference in effect between subgroups training with moderate and high intensity. Walking capacity was continuously improved in proportion to training duration. Consequently, moderate frequency and longer training duration should be prioritized over high-intensity for greater improvement in walking capacity. In contrast with this study, a meta-analysis by Veerbeek et al.38 found intense, highly repetitive, and taskoriented ET induces the largest positive effect on walking capacity, suggesting that different methods of ET can be applied depending on the purpose of rehabilitation. In terms of supervision, subgroups training under supervision significantly increased muscle strength and walking capacity compared to subgroups training without supervision; however, there was no significant difference in effect of CFR between the two subgroups. Although the participants in unsupervised training cohorts were asked to record their interventions, there is a possibility they did not completely perform their assigned training without supervision, which would diminish the effects of exercise. In addition, some adverse events were reported during or after interventions in included studies. Supervision under exercise specialists is integral to safely maintaining greater ET-induced positive effects. Compared to subgroups aged greater than or equal to 65 years, subgroups aged less than 65 years had greater effects of ET both on CRF and walking capacity. These results suggest that aging might attenuate physiological response or adaptation to exercise,39 consequently indicating that ET is necessary for the elderly population, distinct from the younger stroke population. Other results suggest that shorter poststroke duration (<2 years) was significantly associated with greater improvement in CRF and muscle strength. This suggests that aging without rehabilitation might diminish the effects of ET in stroke patients; thus, ET early after stroke is important for greater ETinduced those benefits.39,40 On the contrary, the largest ES of walking capacity was shown in subgroups with a longest post-stroke duration (4 years). As most of these subgroups had lowered walking capacity at baseline, these subgroups might increase walking capacity greater than subgroups with shorter post-stroke duration and higher baseline walking capacity after ET. Moreover, there was no significant difference in the effect of walking capacity between the younger and elderly subgroups. Therefore, it can be suggested that walking capacity can be significantly improved by ET, and is less affected by a long post-stroke duration and aging. There are some limitations in this study. First, studies only published in English were retrieved, which could increase a risk of bias; however, significant publication bias was not found in all 3 outcomes. Second, all studies were conducted on stroke patients who ambulate independently, thus
limiting the generalizability of this review to non-ambulatory stroke patients. Third, most included studies did not involve CON. However, differences in ESs were insignificant between randomized and nonrandomized trials.41 Lastly, this meta-analysis did not involve the facilitators and barriers to exercise. There are various factors influencing the participation in exercise, such as stroke severity, comorbid conditions, fatigue, and learning ability.7 Further studies are warranted to investigate those crucial elements to increase participation with exercise. Despite these limitations, this meta-analysis study has significant strengths. This was the first study to investigate the effects of ET on CRF, muscle strength, and walking capacity in stroke patients. To enhance the validity of the research, this review excluded studies conducting solely AT or RT and/or targeting nonambulatory stroke patients. Furthermore, this review determined that all gains in CRF, muscle strength, and walking capacity in stroke patients can be significantly induced by ET, and that longer duration for AT and moderate-intensity and lower volume for RT are the priority for greater effects. Thus, it is believed that this study expands our knowledge to provide a more optimized exercise strategy for stroke patients.
Summary and Conclusion This systemic review and meta-analysis found that ET significantly improved all CRF, muscle strength, and walking capacity in ambulatory persons with stroke. Through subgroup analyses, this study suggests that moderate-intensity (AT: 40%-60% of HRR, RT: 50%-70% of 1RM) and 3 days per week for 20 weeks should be considered as a priority in ET program for greater effect on all 3 outcomes. For detailed training variables, a volume of 30 minutes (AT), 2 sets of 10-12 repetitions, and 8 exercises (RT) can be additionally recommended, and ET should start early after a stroke has occurred. To continuously improve muscle strength, it is necessary to change training elements during mid/long-term training (>12 weeks). Different ET is necessary for the elderly stroke patients aged 65 years and should be under supervision. In future investigations, researchers may wish to examine various training components in detail to establish further optimized exercise guidelines.
Authors’ Contributions Junghoon Lee: Study design, data collection, data analysis, data interpretation, manuscript writing; Audrey J Stone: Manuscript writing, manuscript review.
Ethics Approval and Consent to Participate Not applicable
Consent for Publication Not applicable
ARTICLE IN PRESS EXERCISE TRAINING FOR STROKE SURVIVORS
Availability of Data and Material The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Acknowledgments We would like to thank the authors from the included studies who provided additional information.
Conflict of Interest The authors declare that they have no conflict of interest.
References 1. Kurl S, Laukkanen JA, Rauramaa R, et al. Cardiorespiratory fitness and the risk for stroke in men. Arch Intern Med 2003;163:1682-1688. 2. Macko RF, Katzel L, Yataco A, et al. Low-velocity graded treadmill stress testing in hemiparetic stroke patients. Stroke 1997;28:988-992. 3. Mohan KM, Wolfe CD, Rudd AG, et al. Risk and cumulative risk of stroke recurrence: a systematic review and meta-analysis. Stroke 2011;42:1489-1494. 4. Bohannon RW. Muscle strength and muscle training after stroke. J Rehabil Med 2007;39:14-20. 5. Yang Y-R, Wang R-Y, Lin K-H, et al. Task-oriented progressive resistance strength training improves muscle strength and functional performance in individuals with stroke. Clin Rehabil 2006;20:860-870. 6. Hill TR, Gjellesvik TI, Moen PMR, et al. Maximal strength training enhances strength and functional performance in chronic stroke survivors. Am J Phys Med Rehabil 2012;91:393-400. 7. Billinger SA, Arena R, Bernhardt J, et al. Physical activity and exercise recommendations for stroke survivors: a statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2014;45:2532-2553. 8. Committee PAGA. Physical activity guidelines advisory committee report. Washington (DC): US Department of Health and Human Services; 2008. 9. Gebruers N, Vanroy C, Truijen S, et al. Monitoring of physical activity after stroke: a systematic review of accelerometry-based measures. Arch Phys Med Rehabil 2010;91:288-297. 10. Boyne P, Welge J, Kissela B, et al. Factors influencing the efficacy of aerobic exercise for improving fitness and walking capacity after stroke: a meta-analysis with metaregression. Arch Phys Med Rehabil 2017;98:581-595. 11. Wist S, Clivaz J, Sattelmayer M. Muscle strengthening for hemiparesis after stroke: a meta-analysis. Ann Phys Rehabil Med 2016;59:114-124. 12. Lloyd M, Skelton DA, Mead GE, et al. Physical fitness interventions for nonambulatory stroke survivors: a mixed methods systematic review and meta analysis. Brain Behav 2018;8:e01000. 13. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Med 2009;6:e1000100.
11 14. Cochran WG. Some methods for strengthening the common x 2 tests. Biometrics 1954;10:417-451. 15. Higgin J, Thompson S, Deeks J, et al. Measuring inconsistency in meta-analysis. Br Med J 2003;327:557-560. 16. Cohen J. Statistical power analysis for the behavioral sciences. Lawrence Earlbaum Associates; 1988. p. 20-26. 17. Egger M, Smith GD, Schneider M, et al. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315:629-634. 18. Marzolini S, Brooks D, Oh P, et al. Aerobic with resistance training or aerobic training alone poststroke: a secondary analysis from a randomized clinical trial. Neurorehabil Neural Repair 2018;32:209-222. 19. Rimmer JH, Rauworth AE, Wang EC, et al. A preliminary study to examine the effects of aerobic and therapeutic (nonaerobic) exercise on cardiorespiratory fitness and coronary risk reduction in stroke survivors. Arch Phys Med Rehabil 2009;90:407-412. 20. Taylor-Piliae RE, Hoke TM, Hepworth JT, et al. Effect of Tai Chi on physical function, fall rates and quality of life among older stroke survivors. Arch Phys Med Rehabil 2014;95:816-824. 21. Kim J, Park JH, Yim J. Effects of respiratory muscle and endurance training using an individualized training device on pulmonary function and exercise capacity in stroke patients. Med Sci Monit 2014;20:2543. 22. Teixeira-Salmela LF, Olney SJ, Nadeau S, et al. Muscle strengthening and physical conditioning to reduce impairment and disability in chronic stroke survivors. Arch Phys Med Rehabil 1999;80:1211-1218. 23. Kluding PM, Tseng BY, Billinger SA. Exercise and executive function in individuals with chronic stroke: a pilot study. J Neurol Phys Ther 2011;35:11. 24. Marzolini S, Tang A, McIlroy W, et al. Outcomes in people after stroke attending an adapted cardiac rehabilitation exercise program: does time from stroke make a difference? J Stroke Cerebrovasc Dis 2014;23:1648-1656. 25. Marzolini S, Oh P, McIlroy W, et al. The effects of an aerobic and resistance exercise training program on cognition following stroke. Neurorehabil Neural Repair 2013;27:392-402. 26. Rand D, Eng JJ, Liu-Ambrose T, et al. Feasibility of a 6month exercise and recreation program to improve executive functioning and memory in individuals with chronic stroke. Neurorehabil Neural Repair 2010;24:722-729. 27. Tang A, Marzolini S, Oh P, et al. Feasibility and effects of adapted cardiac rehabilitation after stroke: a prospective trial. BMC Neurol 2010;10:40. 28. Toledano-Zarhi A, Tanne D, Carmeli E, et al. Feasibility, safety and efficacy of an early aerobic rehabilitation program for patients after minor ischemic stroke: a pilot randomized controlled trial. NeuroRehabilitation 2011;28:85-90. 29. Duncan P, Studenski S, Richards L, et al. Randomized clinical trial of therapeutic exercise in subacute stroke. Stroke 2003;34:2173-2180. 30. Jørgensen JR, Bech-Pedersen DT, Zeeman P, et al. Effect of intensive outpatient physical training on gait performance and cardiovascular health in people with hemiparesis after stroke. Phys Ther 2010;90:527-537. 31. Teixeira-Salmela LF, Nadeau S, Mcbride I, et al. Effects of muscle strengthening and physical conditioning training on temporal, kinematic and kinetic variables during gait in chronic stroke survivors. J Rehabil Med 2001;33:53-60. 32. Carr M, Jones J. Physiological effects of exercise on stroke survivors. Top Stroke Rehabil 2003;9:57-64. 33. Rimmer JH, Riley B, Creviston T, et al. Exercise training in a predominantly African-American group of stroke survivors. Med Sci Sports Exerc 2000;32:1990-1996.
ARTICLE IN PRESS 12 34. Lee YH, Park SH, Yoon ES, et al. Effects of combined aerobic and resistance exercise on central arterial stiffness and gait velocity in patients with chronic poststroke hemiparesis. Am journal Phys Med Rehabil 2015;94:687-695. 35. Eng JJ, Chu KS, Kim CM, et al. A community-based group exercise program for persons with chronic stroke. Med Sci Sports Exerc 2003;35:1271. 36. Herrick AB, Stone WJ. The effects of periodization versus progressive resistance exercise on upper and lower body strength in women. J Strength Cond Res 1996;10:72-76. 37. Abbasian S, MM MR. Is the intensity or duration of treadmill training important for stroke patients? A meta-analysis. J Stroke Cerebrovas Dis 2018;27:32-43.
J. LEE AND A.J. STONE 38. Veerbeek JM, van Wegen E, van Peppen R, et al. What is the evidence for physical therapy poststroke? A systematic review and meta-analysis. PLoS One 2014;9:e87987. 39. Mehta S, Pereira S, Viana R, et al. Resistance training for gait speed and total distance walked during the chronic stage of stroke: a meta-analysis. Top Stroke Rehabil 2012;19:471-478. 40. Stoller O, de Bruin ED, Knols RH, et al. Effects of cardiovascular exercise early after stroke: systematic review and meta-analysis. BMC Neurol 2012;12:45. 41. Benson K, Hartz AJ. A comparison of observational studies and randomized, controlled trials. N Engl J Med 2000;342:1878-1886.
Combined Aerobic and Resistance Training for Cardiorespiratory Fitness, Muscle Strength, and Walking Capacity after Stroke: A Systematic Review and Meta-Analysis Junghoon Lee, MS, and Audrey J. Stone, PhD
Background: Cardiorespiratory fitness, measured as peak oxygen consumption, is a potent predictor of stroke risk. Muscle weakness is the most prominent impairment after stroke and is directly associated with reduced walking capacity. There is a lack of recommendations for optimal combined aerobic training and resistance training for those patients. The purpose of this study was to systematically review and quantify the effects of exercise training on cardiorespiratory fitness, muscle strength, and walking capacity after stroke. Methods: Five electronic databases were searched (until May 2019) for studies that met the following criteria: (1) adult humans with a history of stroke who ambulate independently; (2) structured exercise intervention based on combined aerobic training and resistance training; and (3) measured cardiorespiratory fitness, muscle strength, and/or walking capacity. Results: Eighteen studies (602 participants, average age 62 years) met the inclusion criteria. Exercise training significantly improved all 3 outcomes. In subgroup analyses for cardiorespiratory fitness, longer training duration was significantly associated with larger effect size. Likewise, for muscle strength, moderate weekly frequency and lower training volume were significantly associated with larger effect size. Furthermore, in walking capacity, moderate weekly frequency and longer training duration were significantly associated with larger effect size. Conclusions: These results suggest that an exercise program consisting of moderateintensity, 3 days per week, for 20 weeks should be considered for greater effect on cardiorespiratory fitness, muscle strength, and walking capacity in stroke patients. Key Words: Cardiac rehabilitation—exercise—peak oxygen consumption— skeletal muscle—gait © 2019 Elsevier Inc. All rights reserved.
Introduction Cardiorespiratory fitness (CRF) is one the strongest predictors of stroke risk. Measured as peak oxygen consumption (VO2peak), CRF was inversely associated with a 2.30-fold risk for any type of stroke.1 Compared to agematched sedentary groups, individuals with stroke have about 50% decreased CFR2 and are at the risk of From the Department of Kinesiology and Health Education, The University of Texas at Austin, Austin, Texas. Received July 10, 2019; revision received October 11, 2019; accepted October 23, 2019. Financial Disclosure: There is no received funding for this study. Address correspondence to Audrey J. Stone, PhD, Department of Kinesiology and Health Education, The University of Texas at Austin, 2109 San Jacinto Blvd, Stop D3700, Austin, TX 78712. E-mail: [email protected]. 1052-3057/$ - see front matter © 2019 Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.jstrokecerebrovasdis.2019.104498
cardiovascular events by over 25%.3 Muscle weakness is the most prominent impairment after stroke,4 directly associated with reduced walking speed and endurance.5 Hemiparesis, affecting 65% of stroke victims 3, can double physiological energy for normal walking, compared to that of healthy persons.6 In this regard, exercise training (ET) or physical activity is required for the rehabilitation of stroke populations. Physical activity is defined as any bodily movement produced by skeletal muscle that substantially increases energy expenditure.7 Exercise is a subset of physical activity that is planned, structured, and repetitive and is performed deliberately for the purpose of improving or maintain physical fitness.8 The participation of physical activity should be fundamentally encouraged because low levels of physical inactivity is predominant in stroke survivors, which increases cardiovascular risk.9 In particular, well-designed exercise programs are necessary to
Journal of Stroke and Cerebrovascular Diseases, Vol. &&, No. && (&&), 2019: 104498
1
ARTICLE IN PRESS J. LEE AND A.J. STONE
2
facilitate their participation in physical activity and more effectively improve CFR, muscle strength, and walking capacity as mentioned above. A recent meta-analysis study found that high-intensity aerobic training (AT) increases VO2peak further than AT with low-intensity, but walking capacity, measured as walking speed or distance, is not improved in ambulatory persons with stroke.10 Another meta-analysis showed that resistance training (RT) increases muscle strength but not walking distance, measured using the 6 minutes walking test (6MWT).11 Meanwhile, regarding non-ambulatory stroke survivors, one meta-analysis found that ET increases both VO2peak and walking distance.12 However, these previous meta-analyses only included a small number of studies and did not provide appropriate exercise guidelines. Although both AT and RT are essential for stroke patients, no meta-analysis has investigated the effects of combined AT and RT on the key elements of physical fitness: CFR, muscle strength, and walking capacity. Billinger et al.7 provided comprehensive physical activity and exercise recommendation for stroke survivors based on systematic literature reviews. However, structured exercise intervention with detailed demographic and training factors are needed to be statistically reviewed by a meta-analysis combining results from comparable studies which enhance the validity and reliability of conclusions. For the improvement in all 3 factors, CFR, muscle strength, walking capacity, combined AT and RT is ideal compared to any singular exercise modality. Therefore, the primary purpose of this study was to evaluate the effects of combined AT and RT on CFR, muscle strength, and walking capacity in ambulatory persons with stroke by conducting a systematic review and metaanalysis. This review carefully examined the specific variables of the ET regimens (i.e., intensity, duration, frequency, and volume), identified different types of stroke populations (i.e., age, gender, body mass index (BMI), stroke type, and post-stroke period), and reported adverse events to provide sufficient evidence for establishing optimal exercise guidelines. Moreover, this study also attempted to reveal current issues in this field and inspire future research to resolve health issues associated with stroke.
Materials and Methods This current systematic review followed the strategy of the PRISMA statement.13 Ethics committee approval was not sought for the present study because this meta-analysis study was based on the results of previously published studies.
Data Sources Five electronic databases (CINAHL, EMBASE, PubMed, SportDiscus, and Web of Science) were searched for eligible studies published in English from the earliest date available to May 2019. The following keywords were used for searches: ‘exercise training or physical therapy’, ‘cardiorespiratory fitness or aerobic capacity or muscle strength or walk’ and
‘stroke’. Manual searches of reference lists were conducted to ensure all relevant studies were captured. One reviewer (JH) searched all articles and applied the inclusion and exclusion criteria to the titles and abstracts searched. When the information was not clear, the full text papers of the studies were obtained for review. Corresponding authors of potentially eligible studies were contacted if studies reported data for which it was impossible to discriminate.
Study Selection The inclusion criteria for eligible studies were as follows: (1) adult humans with a history of stroke who ambulate independently with or without an assistive device; (2) structured exercise intervention based on combined AT and RT; and (3) measured CFR, muscle strength, and/or walking capacity. Studies including additional interventions such as dietary supplements were excluded to focus on the effects of ET alone. Duplicate studies or sub-studies of included trials were also excluded from the analysis.
Quality Assessment One reviewer assessed the quality of the included studies using the PRISMA recommendations.13 The quality assessment consisted of 6 items: (1) appropriate generation of random allocation sequence; (2) concealment of the allocation sequence; (3) blinding of the assessment and collection outcomes; (4) proportion of participants lost to follow-up; (5) complete outcome data; (6) the intention-to-treat principle.13
Data Extraction Data were extracted from all selected studies to record the detailed information in terms of participant characteristics, study methods, interventions, outcomes, and adverse events. We used means and standard deviation (SD); where standard errors or 95% confidence interval (CI) were provided, they were converted to SD. Population characteristics, such as age, gender, BMI, stroke type, post-stroke period, and the number of participants were recorded to compare the similarity of participants between trials. The primary outcomes were CFR (VO2peak), muscle strength (one-repetition maximum (1-RM) or dynamometry), and walking capacity (6WMT or gait speed). If muscle strength was reported as upper and lower body, the lower body or affected leg was used. Interventions including total duration, frequency (days per week), intensity, session duration, repetition, set, rest periods, type and order of exercise, names of exercise machine or tool, supervisors, places of intervention, and whether or not RT was performed to muscular fatigue were recorded to compare the similarity of training methods between trials. Measurement technique and region were also extracted. The median values were used for calculation if the studies reported a range of data (e.g., 16 for 15-17 repetitions). Detailed interventions about control groups (CON) and any additional interventions were recorded.
ARTICLE IN PRESS EXERCISE TRAINING FOR STROKE SURVIVORS
Data Analysis Heterogeneity between studies were assessed using the Cochran Q statistic14 and the I2 test.15 I2 ranges from 0%100%: a value less than 25% indicates a low risk of heterogeneity, 25%-75 % indicates a moderate risk of heterogeneity, and greater than 75% indicates a high risk of heterogeneity. In each study, the effect size (ES) for the intervention was calculated by the difference between the means of the postmeasurement and premeasurement at the end of the intervention using Cohen's definition. Separate meta-analyses of trials with CFR, muscle strength, and walking capacity were performed to generate the mean ES and 95% CI. ESs were classified according to Cohen’s definition (1988), where .2 is considered small, .5 moderate, and .8 large.16 This study used a fixed-effects model when homogeneity was verified or a random-effects model when heterogeneity was shown by the Q statistic.15 Publication bias was assessed using the Egger’s regression test.17 To evaluate whether an individual cohort had undue influence on the overall meta-analysis result, sensitivity analyses were performed in all 3 outcomes by omitting one of the trials at a time and determining whether statistical conclusion remained the same. All calculations were conducted with SPSS version 20, Microsoft Excel 2016, and STATA version 14.2. Subgroups analyses were performed where sufficient numbers of trials existed in subgroups in order to identify
3
potential factors influencing the effect of exercise on outcomes variables. These subgroup analyses also accounted for heterogeneity between studies: (1) age less than 65 versus age greater than or equal to 65; (2) less than 2 post-stroke years versus greater than or equal to 2 and less than 4 poststroke years versus greater than or equal to 4 post-stroke years; (3) less than or equal to 12 weeks versus greater than 12 weeks; (4) 2 days per week versus 3 days per week versus greater than or equal to 4 days per week; (5) moderate-intensity (AT: 40%-60% of heart rate reserve (HRR), RT: 50%-70% of 1RM) versus high-intensity (AT: 60%-85% of HRR, RT: 70%-90% of 1RM); (6) less than 50 sets per week versus greater than or equal to 50 sets per week; (7) less than 500 repetitions per week versus greater than or equal to 500 repetitions per week; and (8) supervised versus unsupervised training. Random effects meta-analysis regression was conducted to compare the effect estimates (ES) in different subgroups by considering the meta-analysis results from each subgroup separately. To interpret the results of subgroup analyses, P value (P < .05) between study variation was considered for the statistical difference between subgroups.
Results Study Selection and Characteristics The search resulted in 2817 potential studies (Fig 1). From the titles and abstracts, 2742 studies were excluded
Figure 1. Study search and selection process.
ARTICLE IN PRESS J. LEE AND A.J. STONE
4
based on the inclusion criteria, and then 75 full text studies were reviewed. Of these, 18 articles met the criteria. In selected studies, 3 AT groups18,19 and one Chinese martial art group20 were excluded from this analysis. One CON conducting combined AT and RT21 was included in exercise group (EX). Consequently, nineteen ET cohorts in eighteen studies were selected.
Participants Table 1 shows the characteristics of all of the studies included. Articles were published from April 199922 to March 2018.18 A total of 602 participants completed their interventions (EX: 526, CON: 157, female%: 37.8%) ranging from 923 to 12024 participants. The average age of the participants was 62.1 § 10.2 years (EX: 61.2 § 10.23, CON: 64.8 § 10.0 years). Average post-stroke period was 2.4 § 3.3 years (EX: 2.5 § 3.5, CON: 2.2 § 2.7 years).
Interventions All interventions, except for 3 studies implementing unsupervised intervention,18,24,25 were supervised by qualified trainers or researchers in research/community centers or participants’ homes. The mean training period was 15 weeks (minimum-maximum: 421-2418,24-27 weeks). The mean training frequency was 3 days per week (minimum-maximum: 226,28-518,24 days). The intensities of AT were expressed as a percent of maximum heart rate (HRmax), HRR or VO2peak or the Borg rating of perceived exertion (RPE) or revolution per minute (rpm). The intensities ranged from low (40 rpm29) to high (80% of HRmax30). The intensities of RT were expressed as a percentage of 1-RM or 10-RM (a workload which enables the participant to complete 10 repetitions) or RPE. The intensities ranged from low (50%-60% of 1RM24,25) to high (80% of 1RM22,31). All trials, with the exception of 5 studies,19-21,26,27 progressively increased their session duration or intensity over the duration of the intervention on the basis of the progression of aerobic capacity and muscle strength tested at baseline and/or mid-study.
Measurements Of all 18 studies, 9 assessed CFR by measuring VO2peak by the graded exercise test18,19,23-25,27,29,32,33 and 1 used estimated VO 2peak by the submaximal stationary ergometer test.30 For the assessment of muscle strength, 6 studies measured maximum weight moved.18,20,25,26,33,34 and 5 measured maximal voluntary isometric contraction. 22,24,29,31,32 For walking capacity, 11 studies measured 6MWT,18,21,23-30,34 1 measured 12MWT,35 and 3 used gait speed.20,22,31
Effect of Exercise Training Cardiorespiratory Fitness ET significantly increased CFR in ten trials (mean ES = .41, 95% CI = .25 to .56, P < .0001) (Fig 2). The absolute increase of CFR was 12%. Univariate meta-regression did not show heterogeneity between studies (Q = 4.29, df = 9, P = .891, I2 = .0%). In subgroup analyses, younger subgroups aged less than 65 years resulted in a significantly greater improvement of CFR (P < .0001) than elderly subgroups aged greater than or equal to 65 years (P = .06). Subgroups with less than 2 post-stroke years resulted in a significantly greater CFR (P < .0001) than subgroups with greater than or equal to 2 and less than 4 post-stroke years (P = .09). Subgroups training for greater than 12 weeks resulted in a significantly greater CFR (P < .0001) than training for less than or equal to 12 weeks (P = .10). There was no significant difference in effect between subgroups training 3 days per week and greater than or equal to 4 days per week (P = .75), with moderate-intensity (40%-60% of HRR) and high-intensity (60%- 85% of HRR) (P = .34), and with supervision and nonsupervision (P = .29).
Muscle Strength ET significantly increased muscle strength in 11 trials (mean ES = .59, 95% CI = .32 to .86, P < .0001) (Fig 3). The absolute increase of muscle strength was 33%. Univariate meta-regression showed moderate heterogeneity between studies (Q = 29.83, df = 10, P = .001, I2 = 66.5%). In subgroup analyses, there was no significant difference in effect between subgroups aged less than 65 years and aged greater than or equal to 65 years (P = .84). Subgroups with less than 2 post-stroke years resulted in a significantly greater improvement of muscle strength (P < .0001) than subgroups with greater than or equal to 4 post-stroke years (P = .14). Subgroups training for less than or equal to 12 weeks resulted in a significantly greater muscle strength (P < .0001) than training for greater than 12 weeks (P < .01). Subgroups training 3 days per week resulted in a significantly greater muscle strength (P < .01) than training greater than or equal to 4 days per week (P = .10). There was no significant difference in effect between subgroups training with moderate-intensity (50%-70% of 1RM) and high-intensity (70%-90% of 1RM) (P = .71). Subgroups training with lower volume (<50 sets and <500 repetitions per week) resulted in a significantly greater muscle strength (P < .05) than training with higher volume (50 sets and 500 repetitions per week) (P = .27). Subgroups training with supervision resulted in significantly greater muscle strength (P < .001) than training with non-supervision (P < .05).
Walking Capacity ET significantly increased walking capacity in fifteen trials (mean ES = .45, 95% CI = .25 to 0.65, P < .0001) (Fig 4).
Study
Number (%female)
ET
CON
ET
CON ET
CON
Carr and Jones (2003)
30-82
N/A
20
N/A
N/A
N/A
VO2peak, knee isometric extension
16
3
N/A
8ⅹ2ⅹ10
Duncan et al. (2003)
68.5§ 9.0
70.2§ 44 11.40
48
.2
.2
13
3
30
7ⅹ2ⅹ10
Eng et al. (2003)
63.2§ 8.5
N/A
25
N/A
4.2§ N/A 2.9
VO2peak, knee isometric extension, 6 MWT 12 MWT
8
3
60
9ⅹN/AⅹN/A
Jørgensen et al. (2010)
60.4§ 5.7
N/A
14
N/A
2.1§ N/A 1.9
estimated VO2peak, 6MWT
12
3
N/A
4ⅹ3-5ⅹ6-12
fairly light to somewhat hard on RPE 80% of HRmax
Kim et al. (2014) Kluding et al. (2011) Lee et al. (2015)
53.95.82 54.1§ 20 11.7 63.7§ N/A 9 9.1 64.0§ 63.0§ 14 7.4 5.5
N/A
1.1§ 1.2§ 6MWT 0.2 0.3 3.2§ N/A VO2peak, 0.8 6MWT 6.0§ 5.8§ 6MWT, grip 3.3 2.5 strength
4
3
50
N/A
12
3
30
4ⅹ1ⅹ10
16
3
20
Marzolini et al. (2012)
63.6§ 13.5
N/A
41
N/A
1.4§ N/A 2.6
24
3
N/A
14ⅹ2-3ⅹ10-15 1-8wks: 50-60% light to hard Supervised on RPE 9-16wks: 60%-70% of HRR N/Aⅹ 70% of HRR 50%-60% Unsupervised N/Aⅹ10-15 of 1RM
Marzolini et al. (2014)
63.8§ 12.7
N/A
120 N/A
2.0§ N/A 3.3
24
5
20-60
10ⅹ N/Aⅹ10-15
N/A
33
24
5
20-60
11ⅹ1-2ⅹ10-15
N/A 12
N/A
Post-stroke years
N/A
Outcomes
Intervention Duration Frequency AT Volume RT Volume (A ⅹ B ⅹ C) (weeks) (week-1) (min)
VO2peak, leg extension, 6MWT VO2peak, knee isometric extension, 6MWT
AT Intensity
RT Intensity
Supervised or Unsupervised
1-5wks: 40-50% 6-10wks: 50-60% 11-16wks: 60%-70% of original test wattage 40 rpm
N/A
Supervised
N/A
Supervised
N/A
Supervised
Supervised
N/A
the highest possible training intensity N/A
Supervised
50% of VO2peak
N/A
Supervised
40%-70% of VO2peak
50-60% of 1RM
Unsupervised
Unsupervised 5
(Continued)
ARTICLE IN PRESS
Age (years)
EXERCISE TRAINING FOR STROKE SURVIVORS
Table 1. Summary of included studies
6
Table 1 (Continued) Study
Age (years)
Number (%female)
ET
ET
61.7§ 10.0
Rand et al. (2010) Rimmer et al. (2000) Rimmer et al. (2009) Tang et al. (2010) Taylor-Piliae et al. (2014) Teixeira-Salmela et al. (1999)
67.0§ 10.8 53.2§ 8.3 63.7§ 9.1 64.5§ 12.2 69.6§ 9.40 67.7§ 9.2
CON ET
CON
1.2§ 1.3
N/A
11
VO2peak,knee isometric extension, 6MWT 6 MWT, Knee strength VO2peak, leg press VO2peak
53.2§ 35 8.3 N/A 18
35 N/A
N/A
N/A
N/A
2.5§ N/A VO2peak, 2.3 6 MWT 2.8§ 3.2§ gait speed, 4.9 4.0 leg strength 7.6§ N/A gait speed, 9.4 knee isometric extension
24
2
20
N/A
12
3
30
8ⅹ1ⅹ15-20
14
3
N/A
24
2-3
30
12
3
N/A
10
3
20
7ⅹ3ⅹ10
7ⅹ3ⅹ10
68.2§ 44 10.3 N/A 13
N/A
Teixeira-Salmela 67.7§ et al. (2001) 9.2
N/A
N/A
7.7§ N/A 9.4
knee isometric extension, gait speed
10
3
20
ToledanoZarhi et al. (2011)
65.0§ 14 12.0
14
N/A
6MWT
6
2
35-55
65.0§ 10.0
13
48
N/A
Intervention Duration Frequency AT Volume RT Volume (weeks) (week-1) (min) (A ⅹ B ⅹ C)
4.4§ N/A 2.0 N/A N/A
38
N/A
Outcomes
N/A
N/A 9ⅹ2ⅹ15 N/A
N/A
AT Intensity
RT Intensity
60%-80% of VO2peak
70% of 1RM
somewhat hard on RPE 50%-70% of HRR N/A
N/A
Supervised
70% of 10RM N/A
Supervised
60%-80% of HRR N/A
60% of 1RM N/A
80% 1-5wks: of 1RM 50%-70% 6-10wks: 70% of aerobic working capacity 1-5wks: 50-70% 80% of 1RM 6-10wks: 70% of aerobic working capacity 50%-70% N/A of HRmax
Supervised or Unsupervised
ARTICLE IN PRESS
Marzolini et al. (2018)
CON
Post-stroke years
Supervised Supervised Supervised Supervised
Supervised
Supervised
J. LEE AND A.J. STONE
Abbreviations: A ⅹ B x C, number of exercise x sets x repetitions; AT, aerobic exercise training; CON, control group; ET, exercise training group; N/A, not available; RM, repetition maximum; RT, resistance exercise training; VO2peak, peak oxygen consumption; 6MWT, six minutes walking test; 12MWT, twelve minutes walking test. Values are means § SD.
ARTICLE IN PRESS EXERCISE TRAINING FOR STROKE SURVIVORS
7
Figure 2. Forest plot of effect sizes and 95% confidence intervals for 10 cohorts representing cardiorespiratory fitness, based on the fixed effects meta-analysis results.
The absolute increase of walking capacity was 14%. Univariate meta-regression showed moderate heterogeneity between studies (Q = 23.82, df = 14, P = .048, I2 = 41.2%). In subgroup analyses, there was no significant difference in effect between subgroups aged less than 65 years and aged greater than or equal to 65 years (P = .52). Subgroups with greater than or equal to 4 post-stroke years resulted in a significantly greater improvement in walking capacity (P < .01) compared to subgroups with greater than or equal to 2 and less than 4 post-stroke years (P = .09). Subgroups training for greater than 12 weeks resulted in a significantly greater walking capacity (P < .001) than training for less than or equal to 12 weeks (P < .05). Subgroups training 3 days per week resulted in a significantly greater walking capacity (P < .0001) than training 2 days (P = .28) and greater than or equal to 4 days per week (P = .17). Significant difference in effect was not found between subgroups training with moderate-intensity and high-intensity (P = .60). Subgroups training with supervision resulted in a significantly greater walking capacity (P < .001) than training with nonsupervision (P = .06).
Quality Assessment and Potential Bias In the quality assessment, 89% reported appropriate generation of a random allocation sequence (16 of 18),
33% presented concealment of the allocation sequence (6 of 18), 56% described blinding of the assessment and collection outcomes (10 of 18), 94% explained proportion of participants lost to follow-up (17 of 18), 100% exhibited complete outcome data (18 of 18), and 39% reported that the intention-to-treat principle was used for statistical analyses (7 of 18). Egger’s test showed no significant publication bias for all CFR, muscle strength, and walking capacity (P = .94, P = .98, and P = .19 respectively) (Fig 5).
Sensitivity Analysis Sensitivity analysis reported that by excluding any of all cohorts from the meta-analysis, the estimated effects would still be within the 95% CI of the mean ES in all three outcomes, suggesting the results of the meta-analysis would not significantly change after the removal of any one cohort.
Adverse Events The presence or absence of adverse events was recorded in 7 of the 18 studies. Five studies reported there were no adverse events. One study reported 2 participants experienced complications during the exercise testing.33 Another study reported 3 participants had diagnoses of second strokes, but there were no deaths or heart attacks.29
ARTICLE IN PRESS J. LEE AND A.J. STONE
8
Figure 3. Forest plot of effect sizes and 95% confidence intervals for 11 cohorts representing muscle strength, based on the random effects meta-analysis results.
Discussion The primary results of this meta-analysis study were that ET significantly increased CRF, muscle strength, and walking capacity by 12%, 32%, and 14%, respectively, in ambulatory persons with stroke of an average age of 61.2 years. Moderate heterogeneity was found between studies on muscle strength and walking capacity except for CFR. In subgroup analyses for CFR, longer training duration was significantly associated with larger ES. For muscle strength, moderate weekly frequency and lower training volume were significantly associated with larger ES. For walking capacity, moderate weekly frequency and longer training duration were significantly associated with larger ES. Interestingly, subgroups training for middle/long-term (>12 weeks) resulted in significantly greater effects of ET both on CRF and walking capacity, whereas muscle strength in short-term (12 weeks) training of subgroups increased significantly more than middle/long-term training. This may be a result of physiological adaptations in muscular responses to exercise, which can diminish in middle/longterm training.36 These results suggest that middle/longterm is an essential training element for improving both CRF and walking capacity with ET, but it is necessary to change variables of RT during mid/long-term to consistently obtain ET-induced strength gains.
In terms of CRF, there was no significant difference in effect between subgroups training 3 days per week and greater than or equal to 4 days per week, and training with moderate and high intensity. Subgroup analyses were not conducted based on training volume, as most studies conducted AT with 20-30 minutes of session duration. Thus, 3 days per week and moderate-intensity of AT for relatively short session duration combined with RT can be appropriate to improve CRF in stroke patients. These results seem to contradict the results of a meta-analysis37 suggesting that high weekly volume of treadmill ET were significantly associated with greater improvement in VO2peak. In this study, supplementing RT to AT might contribute to increasing CRF by enhanced muscle strength. Marzolini et al.18 found that combined AT and RT resulted in similar VO2peak, but a 2-fold improvement in VO2 at the ventilatory threshold as well as greater muscle strength and mass gains compared with AT alone. Accordingly, moderate weekly training frequency and volume of AT combined with RT could be sufficient for significant improvement of CRF in this study. Contrary to expectations, training with lower volume (<50 sets and <500 repetitions per week) significantly increased muscle strength, whereas subgroups training with higher volume (50 sets and 500 repetitions per week) had no significant effects on strength. Similarly,
ARTICLE IN PRESS EXERCISE TRAINING FOR STROKE SURVIVORS
9
Figure 4. Forest plot of effect sizes and 95% confidence intervals for 15 cohorts representing walking capacity, based on the random effects meta-analysis results.
subgroups training 3 days per week resulted in significantly greater muscle strength than training greater than or equal to 4 days per week. No significant differences between subgroups training with moderate-intensity (50%-70% of 1RM) and high-intensity (70%-90% of 1RM) were identified. As already mentioned, short-term training (12 weeks) was associated with greater strength gains. Consequently, it is possible to suggest that RT 3 days per week with moderate-intensity and lower training volume for short-term combined with AT can induce
significant strength gains in stroke patients. On the other hand, a meta-analysis study investigating the effects of RT alone on strength after stroke reported that moderatehigh intensity and long duration is required for strength gains.11 A possible interpretation for the discrepancy is adding AT on RT as well as the different characteristics of participants (such as age, BMI, post-stroke period, and levels of baseline strength), might influence the change of strength with ET. Future research is warranted to ascertain the influence of these variables on muscle strength.
Figure 5. Funnel plots of publication bias in all 3 outcomes. Abbreviations: SE, standard error; SMD, standardized mean difference.
ARTICLE IN PRESS J. LEE AND A.J. STONE
10
Walking capacity significantly increased in subgroups training 3 days per week, whereas there was no significant effect in subgroups training 2 or greater than or equal to 4 days per week. There was no significant difference in effect between subgroups training with moderate and high intensity. Walking capacity was continuously improved in proportion to training duration. Consequently, moderate frequency and longer training duration should be prioritized over high-intensity for greater improvement in walking capacity. In contrast with this study, a meta-analysis by Veerbeek et al.38 found intense, highly repetitive, and taskoriented ET induces the largest positive effect on walking capacity, suggesting that different methods of ET can be applied depending on the purpose of rehabilitation. In terms of supervision, subgroups training under supervision significantly increased muscle strength and walking capacity compared to subgroups training without supervision; however, there was no significant difference in effect of CFR between the two subgroups. Although the participants in unsupervised training cohorts were asked to record their interventions, there is a possibility they did not completely perform their assigned training without supervision, which would diminish the effects of exercise. In addition, some adverse events were reported during or after interventions in included studies. Supervision under exercise specialists is integral to safely maintaining greater ET-induced positive effects. Compared to subgroups aged greater than or equal to 65 years, subgroups aged less than 65 years had greater effects of ET both on CRF and walking capacity. These results suggest that aging might attenuate physiological response or adaptation to exercise,39 consequently indicating that ET is necessary for the elderly population, distinct from the younger stroke population. Other results suggest that shorter poststroke duration (<2 years) was significantly associated with greater improvement in CRF and muscle strength. This suggests that aging without rehabilitation might diminish the effects of ET in stroke patients; thus, ET early after stroke is important for greater ETinduced those benefits.39,40 On the contrary, the largest ES of walking capacity was shown in subgroups with a longest post-stroke duration (4 years). As most of these subgroups had lowered walking capacity at baseline, these subgroups might increase walking capacity greater than subgroups with shorter post-stroke duration and higher baseline walking capacity after ET. Moreover, there was no significant difference in the effect of walking capacity between the younger and elderly subgroups. Therefore, it can be suggested that walking capacity can be significantly improved by ET, and is less affected by a long post-stroke duration and aging. There are some limitations in this study. First, studies only published in English were retrieved, which could increase a risk of bias; however, significant publication bias was not found in all 3 outcomes. Second, all studies were conducted on stroke patients who ambulate independently, thus
limiting the generalizability of this review to non-ambulatory stroke patients. Third, most included studies did not involve CON. However, differences in ESs were insignificant between randomized and nonrandomized trials.41 Lastly, this meta-analysis did not involve the facilitators and barriers to exercise. There are various factors influencing the participation in exercise, such as stroke severity, comorbid conditions, fatigue, and learning ability.7 Further studies are warranted to investigate those crucial elements to increase participation with exercise. Despite these limitations, this meta-analysis study has significant strengths. This was the first study to investigate the effects of ET on CRF, muscle strength, and walking capacity in stroke patients. To enhance the validity of the research, this review excluded studies conducting solely AT or RT and/or targeting nonambulatory stroke patients. Furthermore, this review determined that all gains in CRF, muscle strength, and walking capacity in stroke patients can be significantly induced by ET, and that longer duration for AT and moderate-intensity and lower volume for RT are the priority for greater effects. Thus, it is believed that this study expands our knowledge to provide a more optimized exercise strategy for stroke patients.
Summary and Conclusion This systemic review and meta-analysis found that ET significantly improved all CRF, muscle strength, and walking capacity in ambulatory persons with stroke. Through subgroup analyses, this study suggests that moderate-intensity (AT: 40%-60% of HRR, RT: 50%-70% of 1RM) and 3 days per week for 20 weeks should be considered as a priority in ET program for greater effect on all 3 outcomes. For detailed training variables, a volume of 30 minutes (AT), 2 sets of 10-12 repetitions, and 8 exercises (RT) can be additionally recommended, and ET should start early after a stroke has occurred. To continuously improve muscle strength, it is necessary to change training elements during mid/long-term training (>12 weeks). Different ET is necessary for the elderly stroke patients aged 65 years and should be under supervision. In future investigations, researchers may wish to examine various training components in detail to establish further optimized exercise guidelines.
Authors’ Contributions Junghoon Lee: Study design, data collection, data analysis, data interpretation, manuscript writing; Audrey J Stone: Manuscript writing, manuscript review.
Ethics Approval and Consent to Participate Not applicable
Consent for Publication Not applicable
ARTICLE IN PRESS EXERCISE TRAINING FOR STROKE SURVIVORS
Availability of Data and Material The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Acknowledgments We would like to thank the authors from the included studies who provided additional information.
Conflict of Interest The authors declare that they have no conflict of interest.
References 1. Kurl S, Laukkanen JA, Rauramaa R, et al. Cardiorespiratory fitness and the risk for stroke in men. Arch Intern Med 2003;163:1682-1688. 2. Macko RF, Katzel L, Yataco A, et al. Low-velocity graded treadmill stress testing in hemiparetic stroke patients. Stroke 1997;28:988-992. 3. Mohan KM, Wolfe CD, Rudd AG, et al. Risk and cumulative risk of stroke recurrence: a systematic review and meta-analysis. Stroke 2011;42:1489-1494. 4. Bohannon RW. Muscle strength and muscle training after stroke. J Rehabil Med 2007;39:14-20. 5. Yang Y-R, Wang R-Y, Lin K-H, et al. Task-oriented progressive resistance strength training improves muscle strength and functional performance in individuals with stroke. Clin Rehabil 2006;20:860-870. 6. Hill TR, Gjellesvik TI, Moen PMR, et al. Maximal strength training enhances strength and functional performance in chronic stroke survivors. Am J Phys Med Rehabil 2012;91:393-400. 7. Billinger SA, Arena R, Bernhardt J, et al. Physical activity and exercise recommendations for stroke survivors: a statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2014;45:2532-2553. 8. Committee PAGA. Physical activity guidelines advisory committee report. Washington (DC): US Department of Health and Human Services; 2008. 9. Gebruers N, Vanroy C, Truijen S, et al. Monitoring of physical activity after stroke: a systematic review of accelerometry-based measures. Arch Phys Med Rehabil 2010;91:288-297. 10. Boyne P, Welge J, Kissela B, et al. Factors influencing the efficacy of aerobic exercise for improving fitness and walking capacity after stroke: a meta-analysis with metaregression. Arch Phys Med Rehabil 2017;98:581-595. 11. Wist S, Clivaz J, Sattelmayer M. Muscle strengthening for hemiparesis after stroke: a meta-analysis. Ann Phys Rehabil Med 2016;59:114-124. 12. Lloyd M, Skelton DA, Mead GE, et al. Physical fitness interventions for nonambulatory stroke survivors: a mixed methods systematic review and meta analysis. Brain Behav 2018;8:e01000. 13. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Med 2009;6:e1000100.
11 14. Cochran WG. Some methods for strengthening the common x 2 tests. Biometrics 1954;10:417-451. 15. Higgin J, Thompson S, Deeks J, et al. Measuring inconsistency in meta-analysis. Br Med J 2003;327:557-560. 16. Cohen J. Statistical power analysis for the behavioral sciences. Lawrence Earlbaum Associates; 1988. p. 20-26. 17. Egger M, Smith GD, Schneider M, et al. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315:629-634. 18. Marzolini S, Brooks D, Oh P, et al. Aerobic with resistance training or aerobic training alone poststroke: a secondary analysis from a randomized clinical trial. Neurorehabil Neural Repair 2018;32:209-222. 19. Rimmer JH, Rauworth AE, Wang EC, et al. A preliminary study to examine the effects of aerobic and therapeutic (nonaerobic) exercise on cardiorespiratory fitness and coronary risk reduction in stroke survivors. Arch Phys Med Rehabil 2009;90:407-412. 20. Taylor-Piliae RE, Hoke TM, Hepworth JT, et al. Effect of Tai Chi on physical function, fall rates and quality of life among older stroke survivors. Arch Phys Med Rehabil 2014;95:816-824. 21. Kim J, Park JH, Yim J. Effects of respiratory muscle and endurance training using an individualized training device on pulmonary function and exercise capacity in stroke patients. Med Sci Monit 2014;20:2543. 22. Teixeira-Salmela LF, Olney SJ, Nadeau S, et al. Muscle strengthening and physical conditioning to reduce impairment and disability in chronic stroke survivors. Arch Phys Med Rehabil 1999;80:1211-1218. 23. Kluding PM, Tseng BY, Billinger SA. Exercise and executive function in individuals with chronic stroke: a pilot study. J Neurol Phys Ther 2011;35:11. 24. Marzolini S, Tang A, McIlroy W, et al. Outcomes in people after stroke attending an adapted cardiac rehabilitation exercise program: does time from stroke make a difference? J Stroke Cerebrovasc Dis 2014;23:1648-1656. 25. Marzolini S, Oh P, McIlroy W, et al. The effects of an aerobic and resistance exercise training program on cognition following stroke. Neurorehabil Neural Repair 2013;27:392-402. 26. Rand D, Eng JJ, Liu-Ambrose T, et al. Feasibility of a 6month exercise and recreation program to improve executive functioning and memory in individuals with chronic stroke. Neurorehabil Neural Repair 2010;24:722-729. 27. Tang A, Marzolini S, Oh P, et al. Feasibility and effects of adapted cardiac rehabilitation after stroke: a prospective trial. BMC Neurol 2010;10:40. 28. Toledano-Zarhi A, Tanne D, Carmeli E, et al. Feasibility, safety and efficacy of an early aerobic rehabilitation program for patients after minor ischemic stroke: a pilot randomized controlled trial. NeuroRehabilitation 2011;28:85-90. 29. Duncan P, Studenski S, Richards L, et al. Randomized clinical trial of therapeutic exercise in subacute stroke. Stroke 2003;34:2173-2180. 30. Jørgensen JR, Bech-Pedersen DT, Zeeman P, et al. Effect of intensive outpatient physical training on gait performance and cardiovascular health in people with hemiparesis after stroke. Phys Ther 2010;90:527-537. 31. Teixeira-Salmela LF, Nadeau S, Mcbride I, et al. Effects of muscle strengthening and physical conditioning training on temporal, kinematic and kinetic variables during gait in chronic stroke survivors. J Rehabil Med 2001;33:53-60. 32. Carr M, Jones J. Physiological effects of exercise on stroke survivors. Top Stroke Rehabil 2003;9:57-64. 33. Rimmer JH, Riley B, Creviston T, et al. Exercise training in a predominantly African-American group of stroke survivors. Med Sci Sports Exerc 2000;32:1990-1996.
ARTICLE IN PRESS 12 34. Lee YH, Park SH, Yoon ES, et al. Effects of combined aerobic and resistance exercise on central arterial stiffness and gait velocity in patients with chronic poststroke hemiparesis. Am journal Phys Med Rehabil 2015;94:687-695. 35. Eng JJ, Chu KS, Kim CM, et al. A community-based group exercise program for persons with chronic stroke. Med Sci Sports Exerc 2003;35:1271. 36. Herrick AB, Stone WJ. The effects of periodization versus progressive resistance exercise on upper and lower body strength in women. J Strength Cond Res 1996;10:72-76. 37. Abbasian S, MM MR. Is the intensity or duration of treadmill training important for stroke patients? A meta-analysis. J Stroke Cerebrovas Dis 2018;27:32-43.
J. LEE AND A.J. STONE 38. Veerbeek JM, van Wegen E, van Peppen R, et al. What is the evidence for physical therapy poststroke? A systematic review and meta-analysis. PLoS One 2014;9:e87987. 39. Mehta S, Pereira S, Viana R, et al. Resistance training for gait speed and total distance walked during the chronic stage of stroke: a meta-analysis. Top Stroke Rehabil 2012;19:471-478. 40. Stoller O, de Bruin ED, Knols RH, et al. Effects of cardiovascular exercise early after stroke: systematic review and meta-analysis. BMC Neurol 2012;12:45. 41. Benson K, Hartz AJ. A comparison of observational studies and randomized, controlled trials. N Engl J Med 2000;342:1878-1886.