A Systematic Review and Meta-Analysis of Tai Chi Training in Cardiorespiratory Fitness of Elderly People

Objectives The purpose of this study was to investigate the influence of Tai Chi on cardiorespiratory fitness (CRF) in elderly people using meta-analysis. Methods This study used seven electronic databases and data retrieved from randomized controlled trials (RCTs) investigating the role of Tai Chi on CRF in the elderly. All these 24 RCTs were screened and selected from 7 literature databases. The Stata 11.2 software (StataCorp, USA) was used for the meta-analysis, subgroup analysis, and bias test, while the Cochrane Collaboration's tool was used for the assessment of the risk of bias (RoB). 4 researchers independently participated in sample selection, data extraction, and RoB assessment. Results Following the inclusion criteria, 24 eligible studies were included in our analysis. The meta-analysis indicated that Tai Chi practice significantly increased the maximum rate of oxygen consumption (VO2 max) (weighted mean difference (WMD)  = 3.76, 95% CI: 1.25 to 6.26, P < 0.1), leading to an overall reduction in the heart rate (HR) (WMD  = −1.84, 95% CI: −2.04 to −1.63, P  ≤ 0.001) and an increase in the O2 pulse (WMD = 0.94, 95% CI: 0.60 to 1.28, P ≤ 0.001) in individuals who practiced Tai Chi regularly compared with those who did not. The subgroup analysis suggested that overall in those who practiced Tai Chi, males (WMD = 1.48, 95% CI: 0.85 to 2.12, P ≤ 0.001) had higher O2 pulse than females (WMD = 0.73, 95% CI: 0.33 to 1.12, P ≤ 0.001). The subgroup analysis also showed an increase in the vital capacity (VC) (WMD = 316.05, 95% CI: 239.74 to 392.35, P ≤ 0.001) in individuals practicing Tai Chi. When the samples were further stratified by Tai Chi practicing time, the subgroup analysis suggested that individuals practicing Tai Chi over a period of 24 weeks showed no significant difference in VC (WMD = 82.95, 95% CI: -98.34 to 264.23, P=0.370), while those practicing Tai Chi over a period of 48 weeks showed a significant increase (WMD = 416.62, 95% CI: 280.68 to 552.56, P ≤ 0.001). Furthermore, the subgroup analysis demonstrated that the increase in VC is significantly correlated with the Tai Chi practicing time (WMD = 344.97, 95% CI: 227.88 to 442.06, P ≤ 0.001). Conclusion Regular Tai Chi practice could improve the CRF in the elderly, as indicated by significant improvement in indicators including VO2max, O2pulse, VC, and HR. However, gender and practice time might influence the overall beneficial outcomes.


Introduction
Cardiorespiratory fitness (CRF) represents the capacity of the circulatory and respiratory systems to supply oxygen during sustained physical activity. Natural processes such as aging, senescence, and chronic diseases [1] often lead to an overall decline in the CRF [2], which is more pronounced in males than in females [3]. A positive correlation was observed between a steady decrease in CRF over time and an increase in the total mortality [4]. Maximal oxygen uptake (VO 2 max ), which decreases at an average rate of 1% per year after the age of 25, is a remarkable predictor of CRF [5]. High incidence of cardiovascular and respiratory diseases is particularly common in populations over the age of 45 [6]. Moreover, studies have found a direct correlation between poor CRF and increased risk of stroke (occurrence and recurrence) [7], atherosclerosis [8], type 2 diabetes [9], and disturbed cerebral blood flow (CBF), which can potentially impact brain structural and functional integrity and cognitive function [10].
Treatment options for cardiopulmonary rehabilitation include aerobic exercises [11], acupuncture [12], and the application of Chinese medicine [13]. However, the factors such as fear of needles and invasive therapeutic methods, and high medical expenses contributed to the avoidance of cardiopulmonary rehabilitation among patients. Aerobic exercise is widely recognized for its role in improving cardiac health and thus has always been recommended by doctors as a treatment option for cardiopulmonary rehabilitation aimed at prevention and recovery from preexisting diseases. Tai Chi involves slow-paced aerobic exercises with moderate intensity and combines delicate physical movements with rhythmic breathing [14], allowing adults of all age groups to participate. us, Tai Chi has gained popularity over the past years. In recent years, an increasing number of randomized control trials evaluating the beneficial effects of Tai Chi on balance function [15], fibromyalgia [16], and cognitive function have been carried out [17]. Some studies have reported the beneficial effects of Tai Chi on CRF in the elderly, while others have not, probably due to differences in geographic locations and practice intensity. To resolve the disparity in these studies, we conducted a systematic literature review and meta-analysis to elucidate the effects of Tai Chi on CRF in the elderly.

Materials and Methods
Our study design followed the guidelines for reporting systematic reviews in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.

Literature Search Strategy.
e protocol of this study was registered with the International Prospective Register of Systematic Reviews (PROSPERO) (registration number: CRD42021272968). e relevant literature studying the relationship between Tai Chi and cardiorespiratory fitness in elderly people was searched in 7 databases, including PubMed, Web of Science, EMBASE, Cochrane Library, Chinese Scientific Citation Database (CSCD), China National Knowledge Infrastructure Database (CNKI), and WanFang Database. e date of literature searching is from inception to June 9, 2021. Relevant systematic reviews and the reference list of included articles were searched to identify any further relevant studies. e search keywords used in Chinese were as follows: "Tai Chi," "cardiorespiratory function," and "aged." Based on similar studies [18], the search keywords in English used were used as follows (

Study Selection Criteria.
e articles were primarily screened based on their titles and abstracts. en, the full texts of these articles were further reviewed by 4 researchers. In the case of disagreement for study inclusion, the researchers would discuss until a consensus was reached. Studies were considered eligible if: (1) e mean age of patients was >50 years.
(2) Tai Chi training was the sole intervention method irrespective of the style. (3) e outcomes included CRF parameters such as VO 2 , vital capacity (VC), and heart rate (HR). (4) Paired groups, including the control group (sedentary lifestyle) and the comparison group (practicing other forms of exercise such as walking or maintaining usual physical activity), were included in the study. (5) Language of publication was either English or Chinese. (6) e study was an RCT.
Studies were excluded if: (1) e study was a review, case study, or report describing a method or protocol. (2) e study cases were already included in another study we have selected. (3) Missing control groups or comparison groups. (4) Incomplete data. (5) e intervention group had a combinatorial exercise regime involving other forms of exercise training (e.g., strength training).

Data Extraction and Risk-of-Bias (RoB) Assessment.
Two independent researchers participated in the data extraction. In case of disagreement, the researchers would discuss until a consensus was reached. e key data extracted from each study were as follows: (1) author details; (2) year of publication; (3) country; (4) sample size (M/F); (5) mean age (Tai Chi group/control group); (6) style of Tai Chi practiced; (7) frequency of exercise; (8) daily duration of exercise; (9) total time of Tai Chi training; and (10) the outcomes of CRF should include at least of the following core outcomes such as VO 2 max (mL kg −1 min −1 , VC (mL), HR (beats per min), and O 2 pulse (mL beat −1 ). e two researchers independently assessed the methodologies of the studies using the Cochrane Collaboration's tool for the assessment of RoB. e RoB assessment involved random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome data, marking of incomplete outcome data, selective reporting, and screening of other existing biases.

Statistical Analysis.
e Stata 11.2 software (StataCorp, USA) was used for conducting the meta-analysis. Regarding continuous variables, several analyses, such as combined effects, heterogeneity analysis, subgroup analysis, and publication bias analysis, were carried out. e calculated results were expressed as weighted mean difference (WMD).
e I 2 and χ 2 homogeneity tests were conducted before the combined effects were evaluated. When I 2 < 50% or P > 0.1, the variables were considered to possess low heterogeneity. When I 2 < 50% or P < 0.1, the variables were considered to possess high heterogeneity. A fixed meta-analysis was performed when I 2 < 50%, and a random meta-analysis was performed when I 2 ≥50%. e publication bias analysis was conducted using Egger's and Begg's plots in the Stata 11.2, and the results are represented using a funnel chart.

Study Selection.
A total of 471 articles were identified from the search results of the 7 electronic databases. A total of 126 articles were excluded due to duplicate representation ( Figure 1). After reading the titles and abstracts, we rescreened the remaining 52 articles. We excluded 10 articles due to lack of inclusion of control group; 16 articles due to the presence of unrelated data not pertinent to this study; 1 article due to inclusion of non-elderly; and 1 article due to missing of information on CRF measurements. Finally, a total of 24 articles that met the eligibility criteria for the systematic review were included.

Characteristics of Selected Studies.
e 24 articles selected for the meta-analysis reported data from 1995 to 2020 and represented individuals geographically localized in countries such as China, the Netherlands, Mexico, and the United States. e sample size of these studies ranged from 20 to 380. A total of 2155 participants were included, with ages ranging from 50 to 89. e most common frequency of Tai Chi training reported was 4 times per week (20%). e longest duration of Tai Chi practice reported was 11 years, and the shortest was 4 weeks (Table 1).

RoB Results.
e results of the RoB of these RCTs are summarized in Figures 2 and 3, respectively. In summary, 6 studies (25%) showed low-risk bias due to random sequence generation; 5 studies (20%) showed low-risk bias attributed to allocation concealment; 24 studies (100%) exhibited lowrisk bias due to blinding of the participants or personnel; 2 studies (8%) showed low-risk bias, which was attributed to blinding of the assessment outcomes; 20 studies (80%) showed a low-risk bias due to incompleteness of the outcome data; 1 study (4%) showed low-risk bias due to selective reporting; and 12 studies (48%) showed low-risk bias due to the presence of other factors or biases.

Tai Chi for VO 2 max .
Two studies including 122 patients contributed to the meta-analysis of the VO 2 max (Figure 4). Tai Chi training significantly increased the VO 2 max compared with the control (WMD � 3.76, 95% CI: 1.25 to 6.26, P � 0.003).

Tai
Chi for HR. 1,492 participants from 15 studies were used for the meta-analysis of HR. As shown in Figure 5, the HR was significantly reduced in participants who practiced Tai Chi compared with those who did not (WMD � -1.84, 95% CI: -2.04 to -1.63, P � 0.001). For these studies, I 2 � 30.9%, χ 2 � 31.83 (d.f. � 22), and P � 0.008, indicating a low heterogeneity across these studies. e funnel plots for several outcomes were not fully symmetrical ( Figure 6). e P value for Egger's test was 0.026. e Z value for Begg's test was 0.79 ( Figure 6).

Tai Chi for O 2 pulse . 267 participants from 4 studies
were used for the meta-analysis of O 2 pulse . As shown in Figure 7, the O 2 pulse was significantly increased in participants who practiced Tai Chi compared with those who did not (WMD � 0.94, 95% CI: 0.60 to 1.28, P � 0.001). For these studies, I 2 � 16.5%, χ 2 � 8.38 (d.f. � 7), and P � 0.300, indicating a low heterogeneity across these studies. e subgroup analysis was performed to compare the effects of Tai Chi across different genders in the test population. e results suggested that Tai Chi practice in males resulted in a significant increase in the O 2 pulse (WMD � 1.48, 95% CI: 0.85 to 2.12, P � 0.001) ( Figure 8). No heterogeneity was observed in these studies that included male participants, as indicated by I 2 � 0.0% ( Figure 8). Similarly, Tai Chi practice in females significantly increased the O 2 pulse as well (WMD � 0.73, 95% CI: 0.33 to 1.12, P � 0.001) ( Figure 9). No heterogeneity was observed in these studies that included female participants (I 2 � 0.0%) (Figure 9) 3.4.4. Tai Chi for VC. 748 participants from 8 studies were used for the meta-analysis of VC. As shown in Figure 10, the VC was significantly increased in participants who practiced Tai Chi compared with those who did not (WMD � 3 16.05, 95% CI: 239.74 to 392.35, P � 0.001). For these studies, the I 2 � 40.7%, heterogeneity χ 2 � 5.17 (d.f. � 9), and P � 0.086, indicating a low heterogeneity across the studies. e funnel plots for several outcomes were not fully symmetrical Records excluded by titles and abstracts n=293 Full-text articles assessed for eligibility n=52 Meeting the inclusion criteria n=24 No control group n=10 Not required data n=16 Not including CRF measurement n=1 No older people n=1 Trials included in the systematic review n=24 Identification Included Eligibility Screening Figure 1: Flowchart representing the study selection criteria.      Evidence-Based Complementary and Alternative Medicine ( Figure 11). P value for Egger's test was P � 0.464, and Z value for Begg's test was 0.09. e subgroup analysis was performed to compare the effects of Tai Chi practice of different exercise durations ( Figure 12). e results suggested that no statistically significant difference existed between the control and comparison groups with participants undergoing Tai Chi training less than 24 weeks (WMD � 82.95, 95% CI: -98.34 to 264.23, P � 0.370).
e VC was increased in the Tai Chi practice group with a duration of 48 weeks compared with that of the control group (WMD � 416.62, 95% CI: 280.68 to 552.56, P ≤ 0.001).
In these studies, I 2 � 27.9%, indicating a low heterogeneity across the studies (Figure 13). e Tai Chi training for mixed duration significantly increased the VC compared with the control (WMD � 344.97, 95% CI: 227.88 to 442.06, P ≤ 0.001). ere was no heterogeneity across the studies included in the meta-analysis (I 2 � 3.9%) (Figure 14).

Discussion
Tai Chi is originated from traditional Chinese martial arts and medicine [43] and was practiced to maintain physical    Evidence-Based Complementary and Alternative Medicine and mental health. e unity of opposites representing yinyang is also an integral part of the symbol representing Tai Chi. Tai Chi is also known as the "moving meditation" [44]. is study aimed to evaluate the effectiveness of Tai Chi in improving CRF in the elderly using a meta-analysis approach, which included 2155 participants from 24 RCTs. Based on our analyses, we concluded that overall the Tai Chi training could significantly improve the CRF in the elderly. However, the beneficial effects of Tai Chi are influenced by many factors, including gender and practice time.
VO 2 max and O 2 pulse were indicators of comprehensive circulatory and respiratory ability; in particular, VO 2 max was the gold parameters of CRF. e parameters of cardiorespiratory fitness are various, such as maximal minute ventilation (MMV) and cardio output (CO). However, there were very few literatures including MVV and CO. us, we were unable to perform meta-analysis. Blood pressure was a vital sign, and the change in blood pressure was not influenced by a single factor of Tai Chi training. erefore, blood pressure was not selected as a CRF indicator in this study.

VO 2 max .
e VO 2 max represents the oxygen consumed during a maximum intensity exercise, which can be analyzed using a cardiopulmonary exercise test (CEPT). e VO 2 max is an indicator of CRF [45]. e results of the meta-analysis suggested that Tai Chi training could significantly improve the VO 2 max in individuals who practiced Tai Chi compared  Evidence-Based Complementary and Alternative Medicine 9 with those who did not (WMD � 3.76, 95% CI: 1.25 to 6.26, P � 0.003). e improvement of VO 2 max might be attributed to the distinct movement patterns performed during the practice of Tai Chi. Tai Chi training requires the center of gravity to move down, the waist to rotate slowly, and the upper and lower limbs to coordinate simultaneously. e overall rhythm of Tai Chi training is regular, involving movements of the abdominal muscles, pectoralis major, and sternocleidomastoid muscle (breathing muscles). e Tai Chi movements also involve trained and rhythmical breathing. e fusion of body exercise and effective breathing enhances the contractility and endurance of the diaphragm.
Although VO 2 max is currently the "gold standard" for CRF, there are very few published articles that used CEPT to evaluate the efficacy of Tai Chi. is may be due to the fact that the presence of preexisting chronic diseases in the elderly hindered the performance of CEPT, as this test method involves high-risk events such as palpitations, loss of consciousness, and, more seriously, a sudden death.

HR.
HR is a commonly measured vital sign, which is regulated by the autonomic nervous system [46]. Our analyses showed that the Tai Chi training significantly reduced the HR in those who practiced Tai Chi (WMD � −1.84, 95% CI: −2.04 to −1.63, P ≤ 0.001). A previous study has proved that Tai Chi could enhance parasympathetic activity and decrease sympathetic activity [47]. e neurophysiological mechanism of Tai Chi may involve the activation of the parasympathetic nervous system, which is known to decrease HR [48] and play an active role in relieving anxiety and fear [49]. Elevated HR can increase the risk of sudden death [50]. Based on these studies, we believe that long-term and regular Tai   e subgroup analysis suggested that males (WMD � 1.48, 95% CI: 0.85 to 2.12, P ≤ 0.001) had higher O 2 pulse than females (WMD � 0.73, 95% CI: 0.33 to 1.12, P ≤ 0.001). Females, on average, have smaller body sizes and organs, while their body fat is usually higher than males [52]. Additionally, females seem to be more vulnerable to cardiovascular diseases caused by obesity [53]. High body fat is a negative factor [54] limiting the performance of females during Tai Chi training. Also, for participants who practiced Tai Chi, males have better CRF than females, a difference estimated to be 20% [55], which is consistent with our findings. e average O 2 pulse in females is less than what is observed in males.
ere was limited literature in this context, preventing us from drawing more precise conclusions.
Our results also suggested that gender could affect the effectiveness of Tai Chi in individuals. After Tai Chi-based interventions were initiated, males showed higher O 2 pulse than females. e factors such as gender, body fat, and fat metabolism should be incorporated to develop a more personalized Tai Chi training regime in rehabilitation clinics.

VC.
e VC is one of the most commonly used indicators for evaluating the physiologic and pathophysiologic state of the lungs, due to the simple and fast measurement process. VC is an indicator for the inspiratory reserve capacity, expiratory reserve capacity, and tidal volume of an individual [56]. VC is also used for the diagnosis of lung diseases such as chronic obstructive pulmonary disease (COPD) [57] and asthma [58]. Our results suggested that Tai Chi significantly increased the VC in individuals (WMD � 316.05, 95% CI: 239.74 to 392.35, P ≤ 0.001). In 8 studies for individuals with a short practice duration of 4 weeks, the WMD was 670.00 with a 95% CI ranging from 215.60 to 1124.40. In individuals who underwent a longer training duration (96 weeks), the WMD was 265.92, with a 95% CI ranging from 72.18 to 604.00. is sharp increase in VC may be associated with a cardiorespiratory stress reaction to short periods of high-frequency Tai Chi-based training (7 times per week, 60 min per day, and a total duration of 4 weeks). When the training lasted 96 weeks, the beneficial effect on the VC may not be evident. Overtraining often reduces skeletal muscle strength [59] and induces oxidative stress [60], which may trigger a ceiling effect. According to our subgroup analysis, VC was higher in individuals who underwent Tai Chi training for 48 consecutive weeks (WMD � 416.62, 95% CI: 280.68 to 552.56, P ≤ 0.001) than those who practiced Tai Chi for 24 consecutive weeks (WMD � 82.95, 95% CI: -98.34 to 264.23, P � 0.370). e increase in VC for individuals practicing Tai Chi for 24 weeks was reversed when the training was done over a 48week period. We hypothesized that the improvement of VC could only be achieved by compounding the effects of exercise over a longer training duration. However, VC turned out to be an indicator with reduced sensitivity, which changed significantly with long durations of Tai Chi training. Tai Chi involves aerobic exercises of low-to-moderate intensity. Determining the training duration may provide better clinically significant insight. According to Begg's test (P � 0.464) and the funnel plot, there was no bias in the published studies.

Conclusions
To the best of our knowledge, our study is the first to report the correlation between CRF and Tai Chi training in the elderly using meta-analysis. Our findings suggest that Tai Chi training effectively improved the CRF in older adults. We demonstrated that practicing Tai Chi could benefit body function by enhancing factors such as VO 2 max , O 2 pulse , VC, and HR. Additionally, we found that gender and practice time can also influence the outcome of Tai Chi practice. Compared with females, males may benefit to a greater extent showing better CRF. We also demonstrated that longer practice time could improve the CRF. us, this study contributes to the existing knowledge and provides a new direction for further study.

Limitations.
ere were several limitations of this study. (1) Although we included studies from multiple databases, we only considered studies published in English and Chinese, which might undergo the risk of miss studies; (2) limiting the number of studies by our inclusion criteria may lead to bias; and (3). e quality of included literature was low methodological. e descriptions of the 18 studies regarding the random sequence generation were not detailed.
ere were no descriptions of allocation concealment in 19 studies. 22 studies have the risk in blinding of the assessment outcomes.
ese limitations could possibly attribute to multiple factors. First, how Tai Chi could relate to CRF has not attracted enough attention. Second, Tai Chi as the exercise therapy was unable to be blinded. Lastly, Tai Chi training requires disciple and it is rather difficult to adhere to a regular training regimen. (4) e subgroup analysis of the control group was not performed, which may further contribute to biased results. erefore, additional RCTs with larger sample sizes would be essential in future studies. Data Availability e data for supporting this review were taken from previously reported and datasets, which have been cited. Data are available upon request to the corresponding author.

Conflicts of Interest
e author(s) declare that there are no conflicts of interest regarding the publication of this paper.

Authors' Contributions
Yulong Wei contributed to developing the overall design of study. Tianyang Tan performed the literature review from relevant databases, assessed the quality of the study, and wrote the manuscript under the guidance of Yulong Wei. Tianyang Tan and Yanyan Meng contributed equally to this article. Yanyan Meng was the Co-first author. Chengchao Wang and Chaoyang Zhang contributed to screening the articles independently. Meng Liu and Xirui Zhao contributed to data extraction and reviewed methodological quality independently. Yulong Wei arbitrated the disagreements. Jiaxuan Lyu and Tianyi Lyu supported their valuable advice and optimized the language accuracy.