Meta-analysis of the effects of high-intensity intermittent exercise on cardiopulmonary function rehabilitation in patients with stroke

WANG Cheng-shuo, WU Sheng-zhu, WU Liang, XU Ya-nan, ZHANG Lin-li, YONG Ming-jin

1. Tianjin University of Sport, Tianjin 301617, China

2. Beijing Xiaotangshan Hospital, Beijing 102211, China

3. Lianyungang Hospital of Traditional Chinese Medicine, Lianyungang 222001, China

Keywords:

ABSTRACT Objective: Systematically evaluate the rehabilitation effect of high-intensity intermittent exercise (HIIT) on cardiovascular function in stroke patients, in order to provide a basis for selecting the best rehabilitation plan for stroke patients.Methods: Computer retrieval of CNKI,WanFang Data, VIP, CBM, Pubmed, EMbase, Web of science, The Cochrane Library databases was conducted from the establishment of the database until March 2023.Randomized controlled trials on HIIT improving cardiovascular function in stroke patients were included,and the included literature was screened, data extracted, and bias risk evaluated.Then, metaanalysis was conducted using RevMan 5.4 software and Stata17.0 software.Results: In the end, 9 articles met the research criteria, with a total of 428 patients.The meta-analysis results showed that compared with the control group, HIIT had significant effects on peak oxygen uptake(VO2peak)[MD=3.87, 95% CI(3.43, 4.31), P<0.000 01], minute ventilation(VE)[MD=7.14, 95% CI(4.34, 9.94), P<0.000 01], peak power(WRpeak)[MD=17.13, 95% CI(13.73 20.54), P<0.000 01], 6-minute walking distance(6MWD)[MD=43.82, 95% CI(16.08, 71.56),P=0.002], The intervention effect of the 10 meter walking test (10MWT)[MD=-2.00, 95% CI(-2.91, -1.08), P<0.0001]was better than that of the control group.Conclusion: The current analysis results show that compared to conventional rehabilitation therapy or continuous aerobic exercise, HIIT has more advantages in improving the cardiopulmonary function of stroke patients.

1.Introduction

Stroke is a major non-infectious disease that endangers the health of our nationals, with more than 2 million new cases per year[1].After the stroke, the individual’s cardiorespiratory fitness(CRF) is less than half of the non-stroke individual, especially in terms of cardiopulmonary function reserves, exercise endurance, and the ability to use oxygen[2].Reduced activity tolerance and concurrent risks of respiratory diseases after stroke have become important causes of death in stroke patients[3].Research results have shown that the aerobic capacity of stroke patients continues to decline within six months of onset, and most patients do not spontaneously recover to the aerobic level of healthy individuals.Within 0 to 30 days after stroke, the maximum oxygen uptake decreases to 10 to 17mL∙kg-1∙min-1, and after 6 months, the maximum oxygen uptake does not increase to>20 mL∙kg-1∙min-1, which is 25% to 45%lower than the maximum oxygen uptake of age matched healthy individuals[4].The reduction of maximum oxygen uptake may affect the rehabilitation of stroke patients, as these patients have a greater demand for aerobic ability for walking and daily living activities.

Studies have shown that post stroke exercise training is an important component in reducing the incidence of cardiovascular events and the risk of stroke recurrence.Early initiation of exercise training to improve cardiovascular function in stroke patients can accelerate recovery in the later stages of the disease[5].The American Heart Association/American Stroke Association recommends that stroke survivors receive aerobic exercise training at least three days a week 20 to 60 min each time, to improve their activity ability, aerobic ability and cardiovascular health[6].However, at present, the routine use of continuous aerobic exercise may not be sufficient as a cardiovascular stimulus to improve neuroplasticity and cardiovascular health[7].Moreover, a single sustained aerobic exercise requires a relatively long time and a single form of exercise, making it difficult for patients to persist in the long term[8].However, using high-intensity intermittent exercise to achieve higher intensity exercise can provide effective methods for rehabilitation professionals to further improve functional recovery in stroke patients.HIIT is an efficient form of exercise that utilizes short periods of high-intensity training and intersperses with active recovery or rest during exercise[9].As early as the last century,HIIT was widely used as a method to improve athletic performance in athlete training.About 15 years ago, the medical field began studying the clinical benefits of HIIT[10].Currently, HIIT has been used as a commonly used intervention to improve cardiopulmonary function, but there is still controversy in its application in stroke patients.With the increase in the number of randomized controlled trials (RCTs) studying the impact of HIIT on cardiovascular function in stroke patients in recent years, this study uses a systematic evaluation method to explore the intervention effects of HIIT on peak oxygen uptake (VO2peak), minute ventilation (VE), peak work rate (WRpeak), 6-minute walking distance (6MWD), and 10-meter walking test (10MWT) in stroke patients, To provide a basis for the construction of a rehabilitation plan for cardiovascular function in stroke patients.

2.Materials and methods

2.1 Exclusion and inclusion criteria

2.1.1 Study type: RCT

2.1.2 Research object

① Conforming to the diagnostic criteria for stroke in accordance with the 2016 Chinese Guidelines and Consensus on the Diagnosis and Treatment of Cerebrovascular Diseases.②Stable vital signs, no cognitive impairment or contraindications to exercise, and with the consent of the patient and their family members.

2.1.3 Intervention measures

The control group received continuous aerobic exercise or routine rehabilitation treatment, while the experimental group received HIIT.

2.1.4 Outcome indicators: VO2peak, VE, WRpeak, 6MWD,10MWT

Conference articles; Review literature; Dissertation Literature with incomplete outcome data; Unable to download full text literature;Literature that does not specify specific treatment methods.

2.2 Search strategy

Computer search of CNKI, WanFang Data, VIP, CBM, Pubmed,EMbase, Web of science, The Cochrane Library databases, with a search time limit from self creation to March 2023.The literature search was conducted using a combination of theme words and free words.The Chinese search terms included: high-intensity exercise, high-intensity training, high-intensity intermittent exercise,high-intensity intermittent training, stroke, stroke, hemiplegia,cerebrovascular accident, cerebral infarction, cerebral infarction,cerebral hemorrhage, randomized, randomized controlled trials.The English search terms include: high-intensity training, high-intensity exercise, high-intensity interval training, high-intensity interval exercise, HIIT, HIIE, stroke, apoplexy, hemiplegia, cerebrovascular disease, cerebral infarction, cerebral hemorrhage, cardio-pulmonary function, cardiac function, random, randomized controlled trial.

2.3 Literature screening and data extraction

Use EndNote X9 software to manage the retrieved literature.Two scholars independently conduct literature screening and data extraction.If there are any disagreements during the process, they can resolve them through negotiation or jointly discuss with the third scholar.When selecting literature, first eliminate duplicate literature,then read the article title to exclude irrelevant literature, and then read the abstract or main text to further confirm whether to include the literature.In case of missing important information, you can contact the corresponding author of the literature by email to ensure the integrity of the data, and finally input the extracted data into the self-made Excel table.

2.4 Risk assessment of bias included in the study

Two scholars evaluated the included literature using the Cochrane Collaborative Network Bias Risk Assessment Tool[11].

2.5 Statistic analysis

Statistical analysis was conducted using RevMan 5.4 software and Stata17.0.The outcome indicator of this study is a continuous variable, and the measurement data uses mean difference (MD) as the effect indicator, and provide its 95% confidence interval (CI).Extract the data of outcome indicators, and then calculate the pre treatment and post treatment differences and standard deviations according to the method provided in 16.1.3.2 of Cochrane Handbook 5.0.2.If it is already stated in the literature, it can be directly extracted and used.The heterogeneity of the included literature was evaluated using Q-test and I2test.When P>0.1 and I2<50%, it indicates no significant heterogeneity between independent studies,and a fixed effects model was used.On the contrary, a random effects model is used.Due to the fact that the number of literature included in all outcome indicators in this study is less than 10, publication bias was comprehensively evaluated using the Egger method and pruning method.

3.Results

3.1 The process and results of literature retrieval

Strictly following the inclusion and exclusion criteria, 518 articles were retrieved from the above 8 databases, including CNKI (n=35),WanFang Data (n=39), VIP (n=5), CBM (n=5), Pubmed (n=12),EMbase (n=40), Web of science (n=182), and The Cochrane Library(n=200).After deduplication, 411 articles were obtained, and those that did not meet the requirements were excluded based on inclusion criteria, ultimately including 9 articles[12-20], The process and results of literature retrieval are shown in Figure 1.

Fig 1 Literature Retrieval Process and Results

3.2 Basic characteristics of included literature

9 articles included in the literature[12-20], There are a total of 428 patients.Mainly from 6 countries including China, the United States,Norway, etc., published from 2015 to 2021.The basic characteristics included in the study are summarized in Table 1.

Tab 1 Basic characteristics of included studies

3.3 Bias risk assessment results

Randomization was mentioned in 9 included papers[12-20], of which 6 papers[12-14,16,18,19]reported the generation method of random sequence, rated as low bias risk, and 3 papers[15,17,20] did not report, rated as uncertain bias risk.Six articles[12,15,16,18-20] did not report blinding of researchers and subjects, indicating an uncertain risk of bias.Three articles[13,14,17] did not blinding researchers and subjects, indicating a high risk of bias.Seven articles[12-15,17,18,20]described blinding outcome measures and rated them as having a low risk of bias.The remaining two articles[16,19]were not reported and rated as having an uncertain risk of bias.Four articles[13-15,17]described allocation concealment and rated it as low bias risk, while the remaining five articles[12,16,18-20] did not describe it and rated it as uncertain bias risk.The results of 9 references[12-20] are complete and no selective reporting of research results has been found; Other sources of bias have not been described.The detailed results of the bias risk assessment included in the study are shown in Table 2.

3.4 Meta analysis results

3.4.1 VO2peak

A total of 7 studies have reported baseline and post intervention data for VO2peak[12,13,15-19], involving 308 patients.The analysis results showed low heterogeneity among the studies (I2=0%, P=0.57).A fixed effects model was used for meta-analysis.Compared with the control group, the experimental group showed a significant increase in VO2peak, with a statistically significant difference [MD=3.87,95% CI (3.43, 4.31), P<0.00001], as shown in Figure 2.

Tab 2 Bias risk assessment results for inclusion in the study

Fig 2 Comparison of VO2peak between the experimental group and the control group

3.4.2 VE

Two studies reported baseline and post intervention data for VE[13,17],involving 100 patients.The analysis results showed low heterogeneity among the studies (I2=0%, P=0.44).A fixed effects model was used for meta-analysis, and compared with the control group, the experimental group showed a significant increase in VE,with a statistically significant difference [MD=7.14, 95% CI (4.34,9.94), P<0.000 01], as shown in Figure 3.

Fig 3 Comparison of VE between the experimental group and the control group

3.4.3 WRpeak

Two studies reported baseline and post intervention data for WRpeak[19,20], involving 176 patients.The analysis results showed low heterogeneity among the studies (I2=0%, P=0.58).A fixed effects model was used for meta-analysis.Compared with the control group, the experimental group showed a significant increase in WRpeak, with a statistically significant difference [MD=17.13, 95%CI (13.73 20.54), P<0.000 01], as shown in Figure 4.

3.4.4 6MWD

A total of 5 studies reported baseline and post intervention data for 6MWD[12,14-16,20], involving 185 patients.The analysis results showed low heterogeneity among the studies (I2=25%, P=0.25).A fixed effects model was used for meta-analysis, and compared with the control group, the experimental group showed a significant increase in 6MWD, with a statistically significant difference[MD=43.82, 95% CI (16.08, 71.56), P=0.002], as shown in Figure 5.

Fig 4 Comparison of WRpeak between the experimental group and the control group

Fig 5 Comparison of 6MWD between the experimental group and the control group

3.4.5 10MWT

A total of 4 studies reported baseline and post intervention data for 10MWT[12,14,15,20], involving 151 patients.The analysis results showed that there was a high degree of heterogeneity (I2=85%,P=0.000 2) among the four articles.Sensitivity analysis was conducted on the four articles, and after exclusion, it was found that Gjellesvik’s study caused significant fluctuations in the effect quantity results[14].Excluding this study, meta-analysis was conducted on the remaining three articles, and heterogeneity was significantly reduced (I2=0%, P=0.65).A fixed effects model was used for meta-analysis, and compared with the control group, the experimental group had a significant decrease of 10MWT, The difference was statistically significant [MD=-2.00, 95% CI (-2.91,-1.08), P<0.000 1], as shown in Figure 6.

Fig 6 Comparison of 10MWT between the experimental group and the control group

3.5 Publication bias analysis

Due to the fact that the number of literature included in the outcome indicators of this study is less than 10, the Egger method was used to analyze the publication bias of the outcome indicators using Stata17.0 software.The analysis results showed that VO2peak P=0.861, 6MWD P=0.716, and 10MWT P=0.206.The P-values of the above outcome indicators are all greater than 0.05, indicating that there is no publication bias.

4.Discussion

VO2peak has been recognized as one of the important indicators that can scientifically and accurately reflect cardiopulmonary function[21].Existing research shows that for every improvement of 1 mL·kg-1·min-1 in VO2peak, the all-cause mortality rate can be reduced by 15%[22].After a stroke, extremely low levels of VO2peak may restrict patients from continuously engaging in lower levels of daily activities and may prohibit patients from engaging in higher levels of daily activities[23].But through reasonable exercise training,the VO2peak of stroke patients can be increased by 9% to 34.8%[24].The results of this study show that compared with the control group,HIIT can significantly increase VO2peak in stroke patients, which undoubtedly has important clinical significance for the recovery of physical function and the reduction of cardiovascular disease risk in patients after stroke.This conclusion is consistent with Anjos’s research conclusion[25].VE refers to the amount of gas entering and exiting the lungs per minute, which is one of the indicators for evaluating lung ventilation function and can also be used to evaluate the amount of exercise a person can perform.The results of this study show that HIIT improves VE in stroke patients better than the control group.Some studies have shown that the pulmonary dysfunction after stroke is mainly manifested as decreased ventilation function and respiratory muscle weakness[26], which is because the patients’respiratory center is damaged, leading to the central control disorder of the diaphragm, abnormal respiratory mode, respiratory rate and respiratory rhythm[27].HIIT can improve the respiratory function and activity tolerance of stroke patients, and reduce the risk of respiratory diseases.In addition, after HIIT, WRpeak significantly increased compared to the control group, which may be due to the strong stimulation brought by HIIT, which enables stroke patients to obtain higher workload during exercise, thereby improving cardiovascular adaptability to exercise and increasing exercise endurance.However,due to the limited number of literature on VE and WRpeak outcome indicators, more and higher quality RCTs are needed in the future to explore the intervention effects of HIIT.6MWD is mainly used to evaluate the maximum distance that subjects can walk within 6 minutes after a stroke, and it is standardized according to the guidelines of the American Thoracic Society[28].6MWD is one of the commonly used methods for measuring CRF[29].In addition, we also indirectly reflect the changes in CRF in stroke patients through 10MWT.10MWT is a method of measuring walking speed that has been proven to evaluate the reliability and effectiveness of stroke patients, specifically measuring the time it takes subjects to complete a straight line distance of 10 meters at their fastest walking speed.The meta-analysis results showed that the experimental group was more effective in improving 6MWD and 10MWT compared to the control group, and the difference was statistically significant,consistent with the meta-analysis results conducted by Wiener[30].

The plan for high-intensity intermittent exercise specifically includes intensity, type, duration, and frequency, and the formulation of the plan plays a decisive role in the impact of CRF on stroke patients.The intensity of exercise is a key component in improving cardiovascular health, helping to maintain and improve peripheral muscle oxidative capacity and exercise performance[31].In the study included in this article, VO2peak, heart rate reserve, maximum power, Borg scale, and peak heart rate were used as indicators of exercise intensity.However, some studies have shown that specifying exercise intensity based on power output related to VO2peak percentage may be a more effective method for HIIT[32], as it better represents individual exercise ability, providing valuable reference for us in clinical practice.The 9 studies included in this article[12-20] mainly include three types of exercise: treadmill training, power cycling training, and skater exercise.Among them, treadmill training may be the best type of exercise for stroke patients, which can greatly improve their motor ability.However, for patients with severely limited walking function, power cycling training may be more advisable, but existing studies have shown that HIIT in both modes appears to be effective.It is worth noting that one of the studies included in this article focuses on skater exercise, which can provide greater assistance to patients in improving proprioception and balance function compared to the previous two types of exercise.However, skater exercise is only applicable to patients with mild stroke.HIIT has significant advantages in terms of time and frequency compared to traditional continuous aerobic exercise, and short-term training is a very attractive exercise method for stroke patients, especially elderly patients.In Crozier’s study[7], HIIT time was mentioned, which refers to the range of high-intensity exercise time from 30 seconds to 4 min, the interval recovery stage time range from 30 seconds to 3 min, and a single intervention lasting for 25 to 30 min.This provides a reference for the formulation of HIIT plans.Finally, the recovery time after HIIT is also crucial, as there is evidence to suggest that the training interval for elderly people(with an average age of 63.0±3.4 years) is at least 3 days, which can reduce the risk of fatigue and achieve optimal recovery results[33].Therefore, in the early stages of training, it is possible to perform training twice a week, especially for elderly patients with stroke, and gradually increase the frequency of exercise while tolerating it.In summary, HIIT may bring a series of benefits to the cardiovascular function of stroke patients, but personalized treatment plans are necessary, and in clinical practice, treatment plans can be flexibly selected based on the specific conditions of the subjects.

The safety issue of high-intensity intermittent exercise has always been one of the main obstacles to its widespread clinical application.Some scholars believe that high-intensity exercise increases the risk of acute myocardial infarction and sudden cardiac death[34].However, the studies included in this article strictly implemented various security measures during the HIIT intervention to ensure the safety of the subjects.Out of 9 studies[12-20], 7 recorded adverse events[12-16,19,20], and none of these 7 studies reported any serious adverse events related to HIIT during or after the intervention.Additionally, participants had high compliance with the intervention and subsequent evaluations.In Carl’s study[35], the safety of three different HIIT regimens for stroke patients was evaluated through electrocardiogram, blood pressure, heart rate, and orthopedics.The results showed that cardiovascular intolerance, severe arrhythmia,ST segment changes, and orthopedic disease issues did not lead to early termination of HIIT, indicating that HIIT for patients with chronic stroke is reasonable and safe.However, more randomized controlled trials are needed to investigate the safety of high-intensity intermittent exercise in a wider range of stroke patients, such as those with subacute or chronic stroke accompanied by other diseases.

The limitations of this study mainly include the following points:①The number of included studies is relatively small, with a total sample size of only 428 cases, and most studies are single center trials.Therefore, the reliability of the conclusions of this study may be affected to some extent.②The 9 studies[12-20] included in this paper all mentioned randomization, but 3 studies[15,17,20]did not report the generation method of random sequence in detail, 2 studies[16,19] did not blind the outcome measurers, and 5 studies[12,16,18-20] did not conduct allocation hiding, which will produce selection, implementation and measurement bias to a certain extent.③There are significant differences in the HIIT schemes applied in each study, which may lead to bias and heterogeneity.④The stroke patients included in the study are mainly concentrated in the subacute and chronic stages.There is limited evidence on the efficacy and safety of HIIT in the acute stage of stroke, and further exploration is needed in the future.⑤Only a few studies included in this article involve follow-up, therefore, the long-term effectiveness of HIIT on cardiovascular function in stroke patients needs further confirmation.

In summary, high-intensity intermittent exercise can improve the cardiopulmonary function indicators of stroke patients.Therefore,it can be concluded that improving the cardiopulmonary function of stroke patients through high-intensity intermittent exercise is feasible and safe with good compliance and mild adverse events such as muscle soreness.However, the number of literature included in this study is limited, and the severity of the patient’s disease and the timing of high-intensity intermittent exercise intervention in the literature have not been unified.Therefore, in the future, higher quality and larger sample size RCTs are still needed to optimize the exercise prescription of HIIT, in order to further improve the effectiveness of rehabilitation treatment.

Authors’ contribution

WANG Cheng-shuo: the lead author, who puts forward the theme and ideas of the article and is responsible for the writing of the article.WU Liang: corresponding author, responsible for the guidance, review and revision of articles.WU Sheng-zhu, XU Yanan: responsible for literature screening, collecting and organizing relevant data.ZHANG Lin-li, YONG Ming-jin: Responsible for organizing and typesetting articles, and participating in article writing.

All authors declare that there is no conflict of interest relationship.