Tai Chi is a branch of Chinese martial arts and has developed since the 17th century in China. The slow, supple, and graceful movement of Tai Chi is rooted in Taoism [1]. Currently, researchers took notice of the good health maintained by Asian individuals practicing Tai Chi in the Western context, prompting investigations using controlled and comparative studies to verify the benefits of Tai Chi [2]. Tai Chi practice involves continuous and slow movements, with significant shifts in body weight, providing effective stimulation to the muscles and proprioception at the knee and ankle joints. This enhances the feedback stimulus, improving balance function and increasing overall stability in the older adults [3]. In addition, participating in Tai Chi training for over a year may lead to faster leg and gastrocnemius reflex reactions and improvements in knee joint position sense [4]. Therefore, Tai Chi has been currently employed for various purposes, including fitness, recreation, self-defense, martial arts skills, meditation, character development and confidence [5], as well as to address physical and mental health related aspects [6-8]. Especially, Tai Chi has proven effective in fall prevention among older adults, and the preventive effects may enhance with increased exercise frequency [9].

Approximately 30-50% of individuals aged 65 and above experience falls each year, with about 20% requiring medical care after a fall, and around 6% leading to fractures. In addition, falls result in disability, increased fear of falling, reduced social engagement, and decreased social mobility among the elderly [10]. As age advances, balance and agility decline further, significantly increasing the risk of falls [11]. More than half of older individuals seeking medical attention due to injuries are there because of falls, which are a common health issue among the older adults. Since 1990, the National Institute on Aging (NIA) and the National Institute of Nursing Research (NINR) in the United States have sponsored the frailty and injuries: Cooperative Studies of Intervention Techniques (FICSIT) project, exploring new intervention measures to improve frailty or reduce falls in older adults. Furthermore, the American College of Sports Medicine provides recommendations on a combination of different types of exercises for older adults, primarily including aerobic exercise, resistance training, flexibility exercises, and balance activities [12].

Due to the beneficial effects of Tai Chi in fall prevention, the application of Tai Chi in older adults has been recommended in many guidelines and has drawn increased attention from researchers and leading to a growing body of related studies [13]. Therefore, it is necessary to grasp the current research trends on the effects of Tai Chi on health. This will enhance the scientific evidence applied Tai Chi research and provide valuable information for future researchers.

Bibliometric network analysis is a set of review methodology to analyze all related publications on a specific topic, which can provide a more relational, contextual, and holistic intellectual landscape concerning specific topics [14]. The combination of systematic mapping and bibliometric analysis provides a comprehensive overview of the development of a research field, including evolving research hotspots over time. This aids researchers and clinicians in understanding specific topics, identifying methodologists, gaps, and trends, and visualizing forefront issues [15], thereby offering guidance and recommendation for research [16].

Therefore, we applied the bibliometric analysis techniques to comprehensively review the research of the effect of Tai Chi on health. We also conducted an in-depth review on the research on balance control and fall prevention, focusing on the most active clusters identified from the literature. This analysis involves tracking the evolution of research hotspots over time and providing metrics for the research network, aiming to offer valuable insights for researchers in this field.

1. Data sources

We conducted a comprehensive scientific bibliometric analysis by searching the Web of Science Core Collection (WOSCC). Our search terms combined medical subject headings and keywords, and the search string was: ((TS = (tai chi*) OR TS = (tai chi quan*) OR TS = (taiji*) OR TS = (tai-ji*) OR TS = (taijiquan*) OR TS = (tai-chi*) OR TS = (tai ji quan*) OR TS = (tai chi*) OR TS = (Taichi*) OR TS = (Taiji*) OR TS = (Tai-ji*) OR TS = (Chi,Tai*) OR TS = (Tai Chii*) OR TS = (Taijiquan*)))，we restricted the database source to the Web of Science Core Collection - Expanded, publication types to “article” or “review,” and language to “English,” covering the period from 1973 to 2023. We extracted complete records of cited references until November 1, 2023, and eliminated duplicate data using CiteSpace. We performed the extraction and reported the reasons for excluding articles in the flowchart (Figure 1).

Figure 1. Flowchart of research design

2. Data statistical analysis

For the analysis of research hotspots, sources, researchers, and their historical evolution, this study employed CiteSpace and VOSviewer, leveraging their distinct functionalities [5,17]. VOSviewer is used to generate network graphs illustrating country and institution collaboration, while CiteSpace is utilized for extracting co-citation and author collaboration analyses.

In the graphs generated by CiteSpace and VOSviewer, node size represents the quantity of papers published by an institution, author, or country/region. The connections between nodes indicate the strength of collaboration. Node size corresponds to the frequency of appearance, and in co-citation analysis graphs, it reflects the number of times a literature piece has been cited. Co-citation relationships between literature pieces are depicted by the connecting lines, reflecting the strength of co-citation [18].

The co-citation of references refers to the situation where two references are cited by a third reference simultaneously, indicating a co-citation relationship between the two. When conducting co-citation network analysis, we first perform a global cluster view analysis and then a timeline view analysis to observe the time span and activity of each cluster. By exploring the co-citation reference clusters, we can discover the time span, research hotspots, and evolution of each cluster. CiteSpace’s Log-Likelihood Ratio (LLR) algorithm was used to extract labels from the titles, keywords, and abstracts of the literature. These labels were then applied to cluster the literature, revealing hotspots and trends in the research field. Finally, we conducted in-depth analysis on the research of the major clusters, regarding the study population, intervention and measurement tools.

1. Characteristics of research regions and international collaboration

Figure 2 illustrates the top 10 countries involved in research on Tai Chi. Leading five countries in publication frequency are China (174), the United States (122), Japan (16), Canada (9), and Australia (7). Figure 2B reveals that the publications of some countries do not align with their centrality. Despite China having the highest number of publications, the United States exhibits the highest centrality, indicated by more connection lines, which means the higher degree of collaboration between countries.

Figure 2. Visualization of the major countries and regions in the research. (A) The top ten countries and regions based on publication frequency. (B) The relationships among the countries of publication by the network visualization

2. Characteristics of research institutions and institutional collaboration

Collaborative analysis of research institutions is indicated in Figure 3. Figure 3A depicts relationships among institutions, where larger nodes indicate higher publication frequency, and more connections suggest higher centrality. Figure 3B presents the top 10 institutions, with the top 5 being University of California (11), Chinese University of Hong Kong (10), Harvard University (9), Hong Kong Polytechnic University (9), and Shanghai University of Sport (9). From Figure 3C, it is evident that Harvard University (9) has the highest total link strength, indicating strong collaborative relationships with other institutions.

Figure 3. Visualization of research institutions. (A) The relationship among institutions by the network visualization. (B) The top ten institutions ranked by publication volume. (C) The top ten institutions ranked by centrality

3. Characteristics of authorship and their collaboration

The authors of the articles and the temporal network of author collaborations are shown in Figure 4. The collaborative author network visualization reveals the frequency of collaboration among authors. Figure 4B highlights the close collaborations among authors. Xu, Furong stands out with the highest publication count (6) and collaboration degree (9), indicating significant influence in the research on Tai Chi. The year with the highest publication volume is 2013, with authors such as Xu, Furong (6), Delmonico, Matthew J (5), Beebe, Nowen (5), Letendre, Jonathan (5), and Beebe, Nowen (4) (Figure 4 A and C).

Figure 4. Visualization of research author: (A) The top ten authors by publication volume. (B) The top ten authors by collaboration degree, indicating the extent of their collaborative efforts. (C) Temporal network of collaboration

4. Citation reference analysis

1) Analysis of research clusters

Through co-citation network analysis, we identified seven distinct clusters (Figure 5). Figure 5A illustrates the modularity and silhouette scores (Q=0.9582; S=0.9239) of these clusters, indicating significant values. Figure 5B presents a timeline view of the research clusters, providing an intuitive representation of the changes in research focus over time and their interrelationships [17].

Figure 5. Analysis of co-cited references clusters using CiteSpace. (A) Network graph of co-cited references. (B) Visual timeline graph of the co-cited references network

According to the results of the LLR algorithm, we observed seven different clusters, each representing a different research topic (Figure 5A). In order of cluster size, they were numbered from #0 to #6, with the largest cluster ranked first. Specifically, these clusters were #0 “balance training,” #1 “postural stability,” #2 “anxiety,” #3 “disease,” #4 “osteoarthritis,” #5 “sport education,” and #6 “injury.”

Upon further examination, we found that the most active clusters were #0 “balance training” and #1 “postural stability,” both of which are related to fall prevention and balance control.

5. In-depth analysis of major clusters: a focus on studies investigating the effect of Tai Chi on balance control and fall prevention

Through CiteSpace’s LLR algorithm, we obtained the major two clusters, namely, #0 “balance training” and #1 “postural stability” (Figure 5A), which are associated with balance control and fall prevention. Therefore, we selected papers for in-depth analysis that investigated the effect of Tai Chi on balance control and fall prevention through randomized controlled trials (RCTs) within these two clusters for in-depth analysis. In the analysis process, our focus will be on highly cited literature, with an emphasis on research aspects such as participant characteristics, intervention duration and measurement methods.

1) Characteristics of the studies

Table 1 shows the characteristics of 16 RCTs investigated the effects of Tai Chi interventions on balance control and fall prevention; participants, intervention and control groups, duration of intervention, and measurement methods. Majority of the studies included healthy community-dwelling or facility-resident older adults, individuals who were healthy or at risk of fall, and individuals with Parkinson’s disease. The majority of these studies compared the Tai Chi intervention group with the non-exercise group control, and the duration of intervention varied widely, ranging from 4 to 96 weeks (Table 1).

Table 1 . Characteristics of the studies.

	Author, year	Study population	Age of inclusion*	Intervention group	Control group	Weeks	Outcome measures	Source of the paper
1	Li F, 2005 [19]	Healthy older adults	≥70	Twenty-four forms of the classical Yang style TC	Stretching	26	Fall counts, functional balance (BBS, Dynamic Gait Index, FRT, SLS)，physical performance (50-foot speed walk and up & go)，fear of falling (SAFE)	doi: 10.1093/gerona/60.2.187
2	Tsang WW, 2004 [20]	Healthy older adults	69.1 (Mean)	The Ng style TC	No intervention (general education)	4 or 8	Balance measurement by computerized dynamic posturography; sensory organization test, dynamic standing balance (limits of stability test), center of pressure	doi: 10.1249/01.MSS.0000121941.57669.BF
3	Wolf SL, 1996 [21]	Healthy older adults	≥70	One hundred and eight form TC	Computerized balance training; no intervention (routine exercise level)	15	Strength (manual muscle test of hip, knee and ankle, grip strength), flexibility, cardiovascular endurance (12-minute walk test), body composition, instrumental ADLs, psychosocial wellbeing (CES-D), fall events	doi: 10.1111/j.1532-5415.1996.tb01432.x
4	Wolfson L, 1996 [12]	Healthy older adults	≥75	Short-term training: balance; strength; balance and strength training Maintenance program: low intensity training with brief TC supplement	No intervention (education) followed by brief TC training	12-week short-term training followed by 24-week maintenance program	Balance (losses of balance during sensory organization testing, functional base of support, single stance time), muscle strength (summed isokinetic torque of eight lower extremity movements), gait velocity	doi: 10.1111/j.1532-5415.1996.tb01433.x
5	Taylor D, 2012 [22]	Older adults at risk of falls	≥65	Modified TC program twice a week vs once a week	Lower limb exercise	20	Fall events, functional mobility (TUG), Dynamic balance (step test), lower limb muscle strength (30-second chair stand test)	doi: 10.1111/j.1532-5415.2012.03928.x
6	Faber MJ, 2006 [23]	Frail older adults in long-term care centers	63-98 (Range)	Seven therapeutic elements of TC	Functional walking; no intervention	20	Fall events, mobility (POMA), performance based physical function (walking speed, timed chair stands test, TUG, FICIST-4 balance test), disability (Groningen Activity Restriction Scale)	doi: 10.1016/j.apmr.2006.04.005
7	Wolf SL, 2000 [24]	Transitionally frail older adults in the facilities	≥70	Twenty-four simplified TC	No intervention (wellness education)	48	Fall events, function (FRT, BBS, SLS, three consecutive chair stands, 10-minute walk test), behavior (FES, Activities-specific Balance Confidence Scale, sickness impact profile, CES-D, MMSE)	doi: 10.1046/j.1532-5415.2003.51552.x.
8	Nowalk MP, 2001 [25]	Ambulatory older adults in long-term care facilities	≥65	TC with behavioral and psychotherapeutic Modulation of the fear of falling	Individualized and progressive strength-training and conditioning program; no intervention	96	Muscle strength and mobility (chair-stand time, 20-foot walk time, grip strength, quadriceps and hip flexor strength), cognitive assessments (MMSE), depression (Yesavage Geriatric Depression Scale), function (Instrumental ADL, Barthel index of ADLs), nutritional screening initiatives	https://pubmed.ncbi.nlm.nih.gov/11527475/
				All residents were provided to basic enhancement program (team management and three educational programs)
9	Nnodim JO, 2006 [26]	Older adults with mild balance impairment in the senior centers and housing facilities	≥65	Twelve unique sequences from TC the Yang short form	Combined balance and step training	10	Static balance (tandem stance, unipedal stance), gait (maximal step length, Rapid step test), TUG for balance and gait	doi: 10.1111/j.1532-5415.2006.00971.x
10	Tousignant M, 2012 [16]	Older adults at high risk for a fall with multiple disabilities, admitted to geriatric hospital	≥65	Eight-form BADUAN-JIN TC	Conventional physical therapy	15	Balance (BBS, foam and dome test), gait (TUG), fear of falling (SAFE), functional autonomy, self-actualization (measure of actualization of potential), self-efficacy (general self-efficacy scale)	doi: 10.3109/09638288.2011.591891
11	Hackney ME, 2008 [27]	Parkinson’s disease	≥40	TC of the Yang short style of Cheng Manching	No intervention	13	Unified Parkinson’s Disease Rating Scale Motor Subscale 3, balance (BBS, tandem stance test, SLS), functional mobility (TUG), gait (standard forward and backward walking; stride length, velocity, functional ambulation profile), gait endurance (6-minute walk test)	doi: 10.1016/j.gaitpost.2008.02.005
12	Li F, 2012 [28]	Parkinson’s disease	40-85	Six-TC movements integrated into an eight-form routine	Resistance training; stretching	24	Postural stability (maximum excursion, directional control by computerized dynamic posturography, FRT), gait (stride length, velocity), knee muscle strength, mobility (TUG), fall events	doi: 10.1056/NEJMoa1107911
13	Voukelatos A, 2007 [29]	Parkinson’s disease	≥60	Various TC programs mainly sun-style, Yang-style, and mixture of several styles	No intervention	16	Fall events, balance measures including sway, leaning balance (maximal balance range, coordinated stability test), lateral stability (maximal lateral sway), and choice stepping reaction time	doi: 10.1111/j.1532-5415.2007.01244.x
14	Cheon SM, 2013 [30]	Parkinson’s disease	64.2 (Mean)	Twelve movements form Sun-style TC	Combined stretching and strengthening exercise; no intervention control	8	Parkinsonism symptoms (Unified Parkinson’s Disease Rating Scale), daily activity (Schwab and England scale), depressive symptoms (Beck’s Depression Inventory), disease-specific quality of life (Bore’s PD QoL scale), functional fitness (Rikli and Jones), muscle strength (chair-stand test, arm-curl test), flexibility (back-scratch, sit-and-reach test), agility (1.5 m up-and-go test), aerobic endurance (6-minute walk test)	doi.org/10.3988/jcn.2013.9.4.237
15	Hartman CA, 2000 [15]	Lower limb osteoarthritis	49-81 (Range)	Nine-form Yang style TC	No intervention (Usual physical activities and routine care)	12	Self-efficacy (arthritis self-efficacy), quality of life (Arthritis Impact Measurement Scale), lower extremity functional mobility (one-leg stand time, 50-foot walking speed, time to rise from a chair)	doi: 10.1111/j.1532-5415.2000.tb03863.x
16	Zheng G, 2015 [31]	University students	16-25	Twenty-four forms of simplified TC	No intervention	12	Balance (Pro-kin system), flexibility (sit and reach test), Cardio-pulmonary fitness (electronic step test instrument，electronic vital capacity instrument), vital signs (blood pressure and resting heart rate), psychological assessment (self-efficacy, stress, attention, self-esteem, quality of life, etc)	doi: 10.1371/journal.pone.0132605

*In cases where the inclusion criteria regarding participants’ age were not specified, the average age or age range of the participants is provided based on the available data in the respective studies..

TC, Tai Chi; BBS, Berg Balance Scale; FRT, Functional Reach Test; SLS, Single leg stance, SAFE, survey of activities and fear of falling in the elderly; ADL, activities of daily living; CES-D, Centers for Epidemiologic Studies Depression Scale; TUG, timed up and go; POMA, performance oriented mobility assessment; FICIST, the frailty and injuries: cooperative studies of intervention techniques; FES, Falls Efficacy Scale; MMSE, mini-mental state examination..

2) Measurement methods in the studies

A variety of measurement tools were employed to assess balance, primarily including the Timed Up and Go Test (TUG) [16,22,23,26-28], Berg Balance Scale (BBS) [16,19,24,27], Single Leg Stance (SLS) [19,24,27], and Functional Reach Test (FRT) [19,24,28]. In terms of fall metrics, the number of fall events [21,23,24,28,29], and the fear of falling, measured by the survey of activities and fear of falling in the elderly [16,19], were commonly reported (Table 1).

3) Most cited papers and papers with transformative significance

We retrieved the top 5 most commonly co-cited references in these two clusters (Table 2). The most frequently cited articles were authored by Wolf SL [21], followed by Li F [19,28]. Li F published articles on the correlation between Tai Chi and balance/fall in 2012 and 2005, respectively, ranks among the top 5 in citation frequency for both papers.

Table 2 . The top 5 most cited references about studies associated with balance and fall.

No. of citations in the network*	No. of citations in the literature**	Author, year	Journals	Title	Major findings
4	1,657	Wolf SL, 1996 [21]	J Am Geriatr Soc	Reducing frailty and falls in older persons: An investigation of Tai Chi and computerized balance training	Grip strength declined in all groups; fear of falling reduced in Tai Chi group
3	1,073	Li F, 2012 [28]	New Engl J Med	Tai Chi and postural stability in patients with Parkinson’s disease	Better in maximum excursion, directional control, stride length, and functional reach in Tai Chi group than other groups; reduced fall events in Tai Chi group than stretching group
2	940	Li F, 2005 [19]	J Gerontol A-Biol Sci Med Sci	Tai Chi and fall reductions in older adults: A randomized controlled trial	Reduced fall in the Tai Chi group compared with the stretching control group. Improvement of functional balance, physical performance, and reduced fear of falling
2	763	Wolfson L, 1996 [12]	J Am Geriatr Soc	Balance and strength training in older adults: intervention gains and Tai Chi maintenance	Improved all balance measures in balance training group. Improved low extremity movement in strengthening group
2	684	Faber MJ, 2006 [23]	Arch Phys Med Rehabil	Effects of exercise programs on falls and mobility in frail and pre-frail older adults: A multicenter randomized controlled trial	Lower fall incidence rate in Tai Chi group compared with functional walking group, statistically not significant. Small, but significant improvement in their POMA and physical performance scores in functional walking and Tai Chi group, compared to control group.

*Based on CiteSpace, **Based on Google scholar..

POMA, performance oriented mobility assessment..

We comprehensively outlined research on Tai Chi for fall prevention and balance control using two different software tools (CiteSpace and VOSviewer, 1973-2023). Studies in this field remain a focal point in international academia, exhibiting sustained research activity.

An analysis of the most frequently cited papers, national cooperation characteristics, and collaborative author networks in the Tai Chi field revealed key trends and research dynamics. The national analysis indicated that although China, where Tai Chi originated, leads in the number of publications, the United States exhibits the highest centrality. High centrality implies key papers or research findings that hold the greatest influence within a particular field. Identifying these centrally important papers or research can assist scholars in pinpointing the main focal points and hot topics within a research area, providing guidance and insights for future studies. In addition, Harvard University demonstrated outstanding influence and cohesion in the institutional collaboration network. This underscores the United States’ pivotal role in international collaboration.

In this study, co-citation network analysis reveals the significant modularity and silhouette scores of the clusters, indicating the accuracy and tightness of the clustering. In addition, among the seven distinguished clusters that represent different research topics, the most active clusters are #0 “balance training” and #1 “postural stability.” Both of these clusters are related to fall prevention and balance control, indicating significant attention to these topics in the literature. Therefore, conducting further research on these themes is crucial for understanding and addressing issues in these areas.

Through the in-depth analysis, we revealed that studies investigating the effects of Tai Chi on balance and fall yielded various results according to participants, intervention types, or outcome variables. The study population included not only community-dwelling older adults but also those residing in facilities, primarily encompassing individuals who were either at potential or current risk of falling and who had specific diseases. This suggests that Tai Chi can be broadly applicable across a diverse elderly population with varying health conditions.

Notably, focused on studies with specific diseases, there was a predominant inclusion of studies targeting Parkinson’s disease, known for its significant risk of falls [32]. A previous meta-analysis revealed that Tai Chi has shown a significant effect on balance scores for individuals with Parkinson’s diseases compared to a non-exercise control group, not significant when compared with active treatment. In addition, Tai Chi significantly reduced fall incidents both in Parkinson’s disease in this meta-analysis [33].

Given the extensive history of Tai Chi, a variety of styles have emerged. Within the scope of articles analyzed for this study, several researchers has specifically identified the traditional or modified Yang and Sun styles of Tai Chi. These are part of the four official standardized forms for competition: 40-form Yang style, 56-form Chen style, 54-form Wu style and 73-form Sun style) [34]. Meanwhile, other researchers have focused on simplified and modified forms derived from various traditional styles or have been purposefully created. The remainder have implanted Tai Chi interventions without specifying a particular style. In addition, despite a wide range of the duration of intervention, most studies were conducted over a period of 12 to 24 weeks. However, particularly extensive interventions, such as those lasting 48 or 96 weeks [24,25], were carried out in long-term care facilities, which suggests a dedication to improving the health of residents and demonstrates the practical feasibility of such prolonged interventions in these settings.

Measurement tools in Tai Chi studies for balance control were various including self-report questionnaires, functional assessments, or specific balance measurement tools. TUG test, one of the most frequently employed tools, involves a sequence of actions including rising from a chair, walking three meters, turning, walking back and sitting down again [35]. This test is designed to evaluate balance, walking speed and functional ability for older adults and recommended for the screening of gait and balance disorders [36]. BBS is another example to assess balance in the elderly by evaluating their ability to perform 14 daily activities that challenge their equilibrium, which range from transitioning from sitting to standing, unsupported standing, standing with eyes closed, picking up an object from the floor, standing on one foot, etc [37]. In addition, several studies focused on decrease of fall events [21-24,29], since the ultimate goal of balance control for older adults is fall prevention.

In Tousignant’ study, both Tai Chi and physiotherapy significantly improved scores on BBS and TUG test, with the Tai Chi group showing an additional decreased incidence of falls [16]. Meanwhile, both Tai Chi and low limb exercise groups exhibited reduced fall rates, although no significant differences in TUG scores was observed [22]. In patients with Parkinson’s disease, Tai Chi led to decrease both in TUG test scores and fall rates, compared to stretching group [28]. In Hackney’s trial, the Tai Chi group showed significant improvement in BBS and TUG test compared to the control group [27].

Although Tai Chi research is abundant, data regarding its impact on preventing severe events, such as falls leading to fractures or death, remain sparse. Furthermore, evidence regarding the effectiveness of Tai Chi in improving balance and reducing falls in populations other that those with Parkinson’s disease is limited. Critical gap in knowledge also exists due to lack of studies investigating the optimal frequency, intensity or duration of intervention of Tai Chi for fall prevention. For instance, a study involving a 13-week Tai Chi program conducted once or twice weekly compared to lower limb exercise did not demonstrate a significant difference in falls rates [22]. Therefore, to effectively incorporate Tai Chi into various health promotion and fall prevention programs, further research is essential, focusing on holistic outcomes in a variety of populations.

Among the top 5 most cited papers on balance and fall, the works of Wolf SL and Li F garnered significant attention. In addition, two articles [19,28] on the correlation between Tai Chi and balance and falls by Li F rank among the top 5 in citation frequency, indicating that the author holds a prominent position in Tai Chi research related to balance and falls.

There are several limitations in this study. First, scientific measurement research has a significant limitation in its reliance on citation-related metrics, potentially introducing biases, particularly citation bias. The sole purpose of citing references is to highlight manuscript quality, which might result in underutilization of existing evidence [38]. Another limitation is that the data collected are solely from WOSCC, potentially leading to a relatively incomplete retrieval of publications [39]. While WOSCC is considered the most appropriate database for scientific measurement research, future advancements in software may enable the simultaneous analysis of results from different databases, reliably automating the removal of duplicate data from sources such as PubMed, Scopus, Springer, EBSCOhost, ProQuest, etc.

Nonetheless, this study represents the first application of CiteSpace and VOSviewer in a scientific measurement analysis of Tai Chi’s role in fall prevention and balance control, providing a novel perspective on the research landscape. CiteSpace and VOSviewer offer time-series analysis capabilities, allowing for the visualization of evolution of the research field over time, which aids in understanding the developmental stages and trends within the research domain and in determination of future research direction. Furthermore, identifying key authors and journals in the field of martial arts and health promotion provides valuable information for stakeholders and policymakers. Additionally, the inclusion of scientific measurement graphs, along with the analysis of intervention intensity, participant characteristics, and major cluster papers, has the potential to inform future trials.

In conclusion, this scientometric study is the first to focus on the major clusters of Tai Chi in the prevention of falls and balance control, providing in-depth analysis of participant characteristics, intervention and measurement methods. Future research should aim to comprehensively identify the optimal Tai Chi regimen, including the appropriate duration and intensity, to maximize its benefits. Moreover, studies should extend their scope to investigate Tai Chi’s effectiveness across diverse health conditions and demographic groups, with the ultimate goal of reducing the fall-related mortality.

This work was supported by a Research Grant of Pukyong National University (2023).

The authors declare no conflicts of interest.