Skip to main content
NCBI home page
BMC Complementary Medicine and Therapies logoLink to BMC Complementary Medicine and Therapies
. 2026 Mar 5;26:135. doi: 10.1186/s12906-026-05314-5

Modified Baduanjin improves balance function in subacute stroke survivors: a randomized controlled study

Weiwei Zhao 1,2,✉,#, Wang Yao 1,2,#, Pei Che 1, Guogang Deng 1, Ming Wu 1, Yan Cao 1, Yingying Ji 1, Guilan Huang 1,✉, Lan Chen 3,✉, Tong Wang 4,✉
PMCID: PMC13069734  PMID: 41781983

Abstract

Background

Balance dysfunction is common after stroke, yet subacute patients often are often unable to perform traditional Baduanjin exercises due to severe motor impairments. The objective of this study was to determine whether a modified Baduanjin program, adapted to sitting and standing positions, could improve balance function in subacute stroke survivors beyond conventional rehabilitation.

Methods

Fifty-four subacute stroke patients were randomly assigned to receive either conventional rehabilitation plus modified Baduanjin (40 min daily, n = 27) or conventional rehabilitation alone (80 min daily, n = 27), five days per week for six weeks. The primary outcome was performance on the Berg Balance Scale (BBS); secondary outcomes included lower extremity motor function (FMA-LE), functional independence (MBI), and dynamic balance during walking assessed by 3D motion analysis.

Results

Patients in the modified Baduanjin group demonstrated significantly greater improvements in balance (BBS: between-group difference 6.1 points, 95% CI: 2.5–9.7, p = 0.001), lower extremity motor function (FMA-LE: 2.3 points, p = 0.019), and functional independence (MBI: 8.4 points, p < 0.001) compared to controls. Dynamic balance during walking, as measured by mediolateral sway amplitude, also showed greater improvement in the intervention group (p = 0.015). The intervention was associated with high adherence (94%) and was well tolerated, with no serious adverse events reported.

Conclusion

Modified Baduanjin is expected to become a safe and viable adjunctive intervention that could help patients with subacute stroke improve their balance in addition to conventional rehabilitation. These preliminary findings warrant confirmation by multicenter trials on a larger scale.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12906-026-05314-5.

Keywords: Modified Baduanjin, Stroke, Balance, 3D motion analysis, Sacral calibration

Introduction

Advances in modern medical technology have substantially reduced the annual mortality rate among stroke patients. However, a significant proportion of survivors (70–80%) continue to experience varying degrees of physical dysfunction [1]. Balance dysfunction, one of the most common post-stroke deficits [2], not only limits mobility and physical activity but also adversely affects activities of daily living and increases the risk of falls [3]. Therefore, effective training methods to enhance post-stroke balance are critically needed. Current approaches to post-stroke balance rehabilitation include motor and occupational therapies such as static and dynamic balance training, trunk control exercises, strength training for upper and lower limbs, ambulation practice, and task-oriented activities [4]. While these methods are beneficial, they often necessitate individualized therapist supervision or specialized equipment, resulting in complex protocols, fragmented movement patterns, and limited scalability. Consequently, there remains a demand for simple, standardized, and home-based exercise programs that patients can perform independently in community or home settings.

Baduanjin, a traditional Chinese mind-body exercise composed of eight coordinated movements, has shown potential utility in stroke rehabilitation. Evidence from previous studies suggests that Baduanjin can improve balance and motor function in individuals with chronic stroke [5–7].

However, traditional Baduanjin is typically appropriate only for individuals with stable ambulation and full upper limb function. In contrast, subacute stroke survivors (defined as within 7 days to 6 months post-onset) often present with significant balance instability and motor impairments, making it difficult to complete the full standing form. Modified Baduanjin may offer several theoretical advantages over conventional rehabilitation: (1) integration of key therapeutic principles—including weight shifting, proprioceptive feedback, and core stabilization—into functionally meaningful movement sequences; (2) task-specific training that more closely mimics real-world activities than isolated balance exercises; (3) potential for sustained home-based practice following discharge, thereby bridging gaps in rehabilitation continuity; and (4) improved patient motivation and long-term adherence due to cultural relevance and familiarity among Chinese populations. Thus, adapting Baduanjin into a structured, stage-appropriate program for subacute stroke patients is both clinically justified and promising. Despite its potential, research on modified Baduanjin in subacute stroke recovery remains limited. To address this gap, we developed a modified Baduanjin protocol incorporating both seated and standing components, aiming to generate foundational evidence for future clinical investigations [8–13].

While clinical balance scales offer valuable functional insights, objective biomechanical assessment can detect subtle changes in dynamic balance during walking. Three-dimensional motion analysis using sacral marker placement offers a validated and practical method to approximate center of mass displacement [14–16], this approach has been validated showing high correlation (r > 0.90) with gold-standard segmental COM calculations. The sacral marker trajectory during walking reflects dynamic balance control—the body’s ability to maintain postural stability as the center of mass moves through space. In stroke patients, excessive sway amplitude, particularly in the mediolateral direction, is indicative of impaired balance control and increased fall risk. This method offers several advantages for assessing subacute stroke patients: minimal setup time, good patient tolerance, continuous data collection during natural gait, and sensitivity to subtle improvements that may not be detected by discrete-task clinical scales—thereby providing quantitative measures of balance control during functional activities [17–21]. Although this approach has been successfully applied in neurological populations, it has not previously been used to evaluate Baduanjin interventions in stroke patients.

Therefore, this study aimed to investigate whether modified Baduanjin improves balance function in subacute stroke patients more effectively than conventional rehabilitation alone, using both clinical outcome measures and objective biomechanical assessment. We hypothesized that modified Baduanjin would lead to greater improvements in static balance, functional mobility, and dynamic balance during walking compared to conventional therapy.

Methods

Experimental design

Following the guidelines laid out in the Declaration of Helsinki, this assessor-blinded, randomized controlled trial was approved by the Wuxi Mental Health Center’s Ethics Committee (Approval No. WXMHCIRB2023LLKY046). The study was registered with the Chinese Clinical Trial Registry (ChiCTR2300069394). Written informed consent was obtained from all participants prior to enrollment. The study follows CONSORT guidelines for nonpharmaceutical interventions, with the checklist available in supplemental table S1.

Participants

An independent researcher, uninvolved in grouping, intervention delivery, or outcome assessment, recruited stroke patients admitted to the Department of Rehabilitation at Wuxi Central Rehabilitation Hospital between March 2023 and January 2024. Inclusion criteria were as follows: (1) diagnosis of cerebrovascular diseases as per the 2019 criteria revised by the National Cerebrovascular Disease Conference of the Chinese Medical Association, with confirmatory head CT or MRI; (2) First-ever stroke occurring 2 weeks to 6 months prior to enrollment; (3) Age between 40 and 80 years; (4) Ability to walk independently for at least 10 m under therapist supervision; (5) Brunnstrom stage ≥ IV in the upper extremities and ≥ III in the lower extremities, without significant paresthesia; (6) Stable vital signs, no progression of neurological symptoms, and capacity to tolerate 40 min of modified Baduanjin practice; (7) Ability to follow simple verbal commands; (8) provision of informed consent by the patient or legal guardian. Exclusion criteria included: (1) Prior experience with Baduanjin; (2) Pre-existing limb motor dysfunction before stroke onset; (3) Lower extremity skin or muscle injuries that could interfere with sEMG measurements; (4) Complications such as shoulder subluxation, tendinitis, or shoulder-hand syndrome; (5) Major organ failure (heart, liver, kidney), autoimmune disorders, malignancies, or other unstable conditions; (6) Severe diseases affecting liver or kidney function, and psychiatric conditions; (7) Severe cognitive or communication impairments; (8) Participation in other clinical trials; (9) Anticipated poor adherence to the intervention protocol; (10) Other conditions deemed unsuitable by the assessors.

Randomization and blinding

Eligible participants were randomly assigned to either the control group (conventional rehabilitation) or the experimental group,, with 27 participants per group. A computer-generated random allocation sequence with variable block sizes (4–6) was created using SPSS 22.0 by an independent statistician with no clinical involvement in the study. Group assignments were placed in sequentially numbered, opaque, sealed envelopes, which were stored in a locked cabinet. An independent research coordinator opened envelopes only after confirmation of eligibility and completion of baseline assessments. The coordinator informed the treating therapist of group allocation after baseline assessment while outcome assessors and data analysts remained blinded throughout the trial. Block randomization was employed to ensure balanced group sizes across recruitment periods [22, 23]. Specifically, when six participants were enrolled, randomization was performed within a block of size six at a 1:1 allocation ratio [24]. Outcome assessors and researchers with over five years of clinical experience remained blinded to group allocation. The statisticians responsible for data analysis also remained blinded; the group codes were disclosed only after final statistical analysis was completed.

Intervention therapy

The control group received conventional rehabilitation, which included resistance training for upper and lower limbs, respiratory exercises, balance training, trunk stabilization exercises, and transfer and posture-specific training. Each intervention was tailored to the individual’s functional abilities. Participants were actively engaged in each session, with therapists providing continuous verbal encouragement and real-time corrective feedback [25]. This program was delivered for 40 min twice daily, five days per week, over six consecutive weeks, by certified physiotherapists.

The experimental group received an additional 40-minute session of modified Baduanjin exercise following 40 min of conventional rehabilitation, totaling 80 min of therapy per day. Each Baduanjin session began with a 5-minute warm-up, followed by eight specific movement forms, each performed six times, except for the eighth form, which was repeated seven times, and concluded with a 5-minute cool-down. Rest intervals of 30–60 s were provided between each movement form to prevent fatigue. Additional breaks were permitted based on individual tolerance, and sessions were paused immediately if participants reported discomfort or exhibited abnormal vital signs. Overall, The entire session lasted 80 min and was conducted once daily, five days per week, for six weeks. Both groups received equal total therapy time (80 min daily).

Specifically, the experimental group received 40 min of conventional rehabilitation plus 40 min of modified Baduanjin. The control group received two 40-minute sessions of conventional rehabilitation separated throughout the day. All sessions were conducted five days per week for six weeks. Rest periods of 30–60 s were provided between movement forms, with extra breaks allowed based on individual tolerance. Sessions were suspended if participants showed signs of excessive fatigue or discomfort. Pre- and post-session vital signs were monitored, and fatigue was assessed using a 0–10 verbal rating scale. One week prior to the intervention, patients assigned to the experimental group underwent training in the modified Baduanjin movements under the supervision of a qualified instructor. Only those who demonstrated mastery of all movement sequences and coordinated breathing techniques were enrolled. Figure 1 illustrates the eight modified Baduanjin movements. During actual sessions, participants practiced following standardized video instructions, with real-time guidance from therapists; family members were also permitted to assist when appropriate.

Fig. 1.

Fig. 1

Open in a new tab

Eight movement forms of the modified Baduanjin. Forms 1-4 and Form 7 involved adjusting the horse stance to a partial squat to symmetrically engage the upper extremities; Forms 5 and 6 consisted of adapted sitting exercises; and in Form 8, participants placed both hands on the back of a chair for support during execution. A family member or caregiver allowed to assist during sessions to ensure safety

The modified Baduanjin protocol included the following components: (1) Two hands holding up the heaven (to regulate the triple burner); (2) Drawing the bow to shoot the eagle; (3) Separating heaven and earth; (4) Wise owl gazing backwards; (5) Swaying the head and shaking the tail; (6) Two hands holding the feet; (7) Clenching the fists and glaring; (8) Bouncing on the toes. The exercise intensity and speed were adjusted according to each participant’s physical condition. Instead of full squatting positions, micro-squats with slight knee flexion (approximately 10–20 degrees) were performed, preserving therapeutic benefits while minimizing balance demands. For participants with significant lower limb weakness, supportive equipment such as chairs was used to ensure stability. Therapists offered hands-on assistance when necessary, and the unaffected limb was used to guide movements of the affected limb during bilateral exercises. Forms 5 and 6, which are physically demanding and may not be suitable for all subacute patients, were adapted for sitting positions. In form 5, sat upright and leaned 45° toward the unaffected side, tightened the muscles on the affected side of the waist, slightly lifted the hip, turned their head and torso toward the affected side, looked upward, retracted the chin, rotated back to center, and then repeated the sequence on the opposite side—each side performed three times alternately. In form 6, participants remained seated, placing their hands over the lower abdomen with fingertips pointing downward. The unaffected hand guided the affected hand to rub the abdomen vertically from bottom to top, followed by circular motions bilaterally. The hands were then placed on the waist, fingers interlaced, palms facing outward. The unaffected hand guided the affected hand to move from the waist to the hips, then forward into a forward bend, sliding the palms along the outer legs to the ankles, crossing the hands over the ankles, hooking the feet, lifting the heels, rising slowly to an upright position, and guiding the hands upward along the inner legs. This sequence was repeated six times [26]. For form 8, participants stood holding the back of a chair for support, lifted their heels, paused briefly, and then gently lowered them to lightly tap the ground—repeated seven times in total. Throughout the exercise, if the affected upper limb could not perform movements independently, it was guided by the unaffected limb or assisted manually by the therapist, with simultaneous instruction on maintaining even, coordinated breathing.

Clinical outcomes

All assessments were conducted at three time points: pre-intervention (T0), and at 3 weeks (T1) and 6 weeks (T2) post-intervention, by non-participating physiotherapists who did not participate in the intervention delivery. These assessors had over five years of clinical experience and were fully trained in administering the outcome measures.

Primary outcome measure

The BBS is a widely adopted clinical tool for assessing both static and dynamic balance capabilities in patients, as well as their risk of falling. It boasts a high inter-rater reliability of 0.98 [27, 28]. Each of the 14 items on the scale is scored from 0 to 4, yielding a maximum total score of 56, with higher scores indicating better balance performance.

Secondary outcome measures

Secondary outcome measures included the mean sway amplitudes (MSA) during walking assessed via sacral calibration in the three-dimensional (3D) motion analysis system, the Fugl-Meyer assessment of lower extremity (FMA-LE), the modified Barthel Index (MBI) and sEMG to evaluate the strength of major lower limb muscle groups.

The FMA-LE is a well-established and recommended scale for the assessment of post-stroke motor impairment, which has attained high item-level reliability (percentage of agreement ≥ 75%) [29], and is recognized as the gold standard for both clinical use and research worldwide [30]. It exhibits excellent psychometric properties in stroke populations, with high inter-rater reliability (ICC > 0.95) and test-retest reliability (ICC > 0.90). The scale has been validated for use in subacute stroke patients and is highly sensitive to changes in motor recovery during rehabilitation.

The MBI serves as a key instrument for measuring a patient’s capacity to perform activities of daily living. It consists of 10 items, with a total score ranging from 0 to 100; higher scores reflect greater functional independence [31]. The MBI was utilized to evaluate the subject’s capacity to perform activities of daily living (ADL). This index, which ranges from 0 to 100 points, assesses various functional tasks, with higher scores denoting better functional independence. The MBI has shown robust reliability (ICC = 0.89–0.95) and internal consistency (Cronbach’s alpha > 0.90) in subacute stroke populations. It is a validated and sensitive measure for detecting changes in functional independence during post-stroke rehabilitation.

MSA was determined using sacral calibration in a 3D motion analysis system (Qualisys, Sweden). This system comprised eight infrared cameras mounted on the walls of the laboratory at an average height of 2.2 m. Motion capture was performed using reflective marker balls (2 cm in diameter). System calibration was achieved using a right-angle frame and a 0.3-m-long T-shaped pole, with data sampled at a frequency of 100 Hz. The sacral marker (SM) was placed at the midpoint between the two posterior superior iliac spines (as illustrated in Fig. 2), following the SM method developed by Zhu et al. [20, 21]. The subjects were instructed to walk naturally in a straight line on a 7-meter walkway at their usual speed (see Fig. 2), completing 3 rounds. Positional data of the sacral marker in the anteroposterior (X), mediolateral (Y), and vertical (Z) directions were recorded in triplicate. MSA, defined as the mean absolute deviation of the SM position from its overall mean position over time, was computed for each trial. Lower MSA values indicate better postural stability during walking [32].

Fig. 2.

Fig. 2

Open in a new tab

Sacral calibration method in the 3D motion analysis; SM, Sacral marker; X, anteroposterior directions; Y, mediolateral directions; Z,vertical directions. A Sacral bone calibration method for 3D motion analysis scene diagram; B Marker ball placement points for the sacral calibration method; C Sacral marker points in the 3D motion analysis system

Formula: Inline graphic  

The electromyographic data were collected using using the FREEEMG 300, a wireless sEMG system manufactured by BTS Bioengineering (Milan, Italy). The root mean square (RMS) values were recorded bilaterally from four key lower limb muscles: rectus femoris, biceps femoris, tibialis anterior, and gastrocnemius, during maximal isometric contractions in flexion and extension. Each measurement was repeated three times, and the average RMS value was used for analysis [33]. As an objective indicator of muscle strength, RMS provides more sensitive and reliable assessment of muscle strength changes compared to clinical manual testing, offering valuable quantitative data on neuromuscular function [34].

Statistical analysis

A formal a priori power analysis was conducted using GPower 3.1 software (Heinrich Heine University, Düsseldorf, Germany). The primary outcome was the Berg Balance Scale (BBS) score at 6 weeks post-intervention. Based on previous studies evaluating Baduanjin interventions in stroke populations [35], a conservative effect size of 0.2 (Cohen’s d) was assumed for calculation. Using a two-tailed independent t-test with α = 0.05 and power (1-β) = 0.80, the required sample size was calculated as 42 participants (21 per group). To account for an anticipated 20% dropout rate—based on our institutional experience with subacute stroke rehabilitation trials —the target sample size was increased to 54 participants (27 per group). The actual dropout was 7.4% (4/54), with 50 participants completing the study, which maintained sufficient statistical power for the primary outcome analysis. Given the anticipated clinical attrition rate of 20%, the minimum required sample size was determined to be 53; thus, enrolling 54 participants ensured adequate power.

All statistical analyses were conducted using SPSS 22.0. Baseline differences between the experimental and control groups in demographic characteristics and outcome indicators were compared using independent t-tests, Mann-Whitney U tests, or χ2 tests, depending on data distribution and scale of measurement. Comparisons of the effect between experimental and control treatments were conducted using a two-factor repeated-measures ANOVA (treatment protocol × measurement time point). If the interaction effects were significant, separate effects analysis was performed, followed by pairwise comparisons using the Bonferroni correction method. The Mauchly test was used to verify the sphericity assumption. Non-sphericity was corrected using the Greenhouse-Geisser method when required. A two-tailed test was employed, P < 0.05 was considered statistically significant. Both intention-to-treat (ITT) and per-protocol analyses were conducted. The ITT analysis included all randomized participants (n = 54) using last observation carried forward for missing data. Per-protocol analysis included only completed participants(n = 50).

Results

Between March 2023 and January 2024, 54 participants meeting the inclusion criteria were randomly assigned to either the experimental group (Baduanjin, n = 27) or the control group (n = 27). During the intervention period, four participants (two from each group) withdrew due to non-intervention-related reasons. leaving 50 participants who completed all assessments. The trial flow is summarized in Fig. 3. Adherence to the intervention was high in both groups. Participants in the experimental group attended an average of 28.3 ± 1.8 sessions (94.3%) out of the planned 30 sessions, while those in the control group attended 28.6 ± 1.5 sessions (95.3%). Non-attendance was primarily due to temporary medical consultations or personal reasons, with no evidence of systematic withdrawal related to the intervention. Fatigue levels were monitored before and after each session using a verbal rating scale (0–10). Mean post-session fatigue scores in the experimental group were 3.2 ± 1.4, indicating mild-to-moderate fatigue levels that resolved with rest. No participant reported persistent or cumulative fatigue affecting daily activities, nor was any modification to the intervention required. No serious adverse events occurred in either group throughout the study. In the experimental group, minor adverse events included transient muscle soreness (n = 3), mild knee discomfort (n = 2), and one episode of dizziness that resolved immediately upon sitting. These did not require medical attention or lead to withdrawal. The control group reported similar minor events, including muscle soreness (n = 2). Overall, the intervention was well tolerated and safe.

Fig. 3.

Fig. 3

Open in a new tab

A CONSORT diagram of patient flow through the study

Baseline data

Table 1 outlines the demographic characteristics of the 50 participants, with a mean age of 61.76 (SD 8.75) years and a mean stroke duration of 49.59 (SD 26.79) days. No significant differences were observed between the two groups in terms of age, gender distribution, duration since stroke, stroke type, or affected hemiplegic side (all P > 0.05), indicating baseline comparability.

Table 1.

Demographic and clinical characteristics of patients at baseline

Characteristic Experimental group
(n = 25)		 Control
group
(n = 25)		 P value
Age (y) 62.12 ± 8.23 61.40 ± 9.26 0.582
Gender (n) Male 15 14 0.774
Female 10 11
Stroke onset duration (d) 46.67 ± 24.62 52.50 ± 28.96 0.530
Stroke type (n) Ischemic 21 18 0.306
Hemorrhagic 4 7
Hemiparetic side (n) Left 15 14 0.774
Right 10 11
Open in a new tab

Values are expressed as mean ± SD

Primary outcome

Table 2 presents the clinical scores results analyzed using a two-way repeated-measures ANOVA, with the intervention mode as the intergroup factor and the three time points (T0, T1, T2) as the intragroup factor. Results of BBS score, the primary outcome measure, revealed significant interaction effects between group and time point (F[2, 96] = 22.203, P = 0.000, η2p = 0.316). Simple effects tests revealed no significant between-group differences at T0 (P = 0.073) and T1 (P = 0.460), but a significant difference emerged at T2 (*P* = 0.002), favoring the Baduanjin group. Within-group comparisons showed significant improvements across all time point. ITT analysis confirmed findings from per-protocol analysis. For the primary outcome (BBS at T2), ITT showed between-group difference of 5.8 points (95% CI: 2.3–9.3, P = 0.002), compared to 6.1 points in per-protocol analysis (95% CI: 2.5–9.7, P = 0.001). Secondary outcomes also demonstrated similar patterns across both analytical approaches.

Table 2.

Repeated-measures variance analysis results of FMA-LE, BBS and MBI scores

T0 95% CI T1 95% CI T2 95% CI Group effect Time effect Group*Time effect
M ± SD M ± SD M ± SD F P η2 F P η2 F P η2
BBS
 Experimental group 39.04 ± 3.46 37.44, 40.33 44.88 ± 3.06 43.60, 46.03 50.72 ± 1.97† 49.85, 51.43 0.001 0.972 0.000 440.724 0.000* 0.902 22.203 0.000* 0.316
 Control group 40.76 ± 3.18 39.64, 42.07 45.52 ± 3.02 44.28, 46.76 48.28 ± 3.13 47.03, 49.69
FMA-LE
 Experimental group 22.80 ± 3.97 21.31, 24.32 25.44 ± 3.12 24.07, 26.52 28.24 ± 2.17 27.44, 29.28 1.015 0.319 0.021 187.854 0.000* 0.796 4.910 0.019* 0.093
 Control group 24.28 ± 3.53 22.69, 25.61 26.64 ± 3.44 25.12, 28.00 28.28 ± 3.49 26.74, 29.58
MBI
 Experimental group 60.60 ± 12.70 56.14, 65.93 76.80 ± 10.83 72.73, 81.20 90.56 ± 6.21† 88.23, 93.29 0.146 0.704 0.003 324.117 0.000* 0.871 16.067 0.000* 0.251
 Control group 66.44 ± 11.80 61.65, 71.02 79.00 ± 9.86 74.19, 82.37 85.60 ± 9.10 81.67, 89.13
Open in a new tab

Abbreviations: BBS Berg Balance Scale, FMA-LE Fugl-Meyer Assessment Lower Extremity, MBI Modified Barthel Index

* indicate statistically significant of interaction or main effect (P < 0.05)

† indicate significant differences between the pre-intervention (T0), and 3weeks (T1) or 6weeks (T2) post-intervention compared with the control group

Secondary outcomes

The ANOVA also indicated a significant group × time interaction (F[2, 96] = 4.910, P = 0.019, η2p = 0.093) for FMA-LE scores, along with a significant main effect of time (F[2, 96] = 187.854, P < 0.001, η2p = 0.796). However, there was no significant main effect of group (F[1, 48] = 1.015, P = 0.319, η2p = 0.021). Post hoc analysis revealed significant effects across different assessment time points in both groups, with no significant differences between each time point.

For MBI scores, a significant interaction between group and time was found (F[2, 96] = 16.067, P < 0.001, η2p = 0.251), with a strong main effect of time (F[2, 96] = 324.117, P < 0.001, η2p = 0.871). Simple effects tests indicated no significant group differences at T0 (P = 0.099) and T1 (P = 0.456), but a significant effect at T2 (P = 0.029).

Table 3 details the RMS results of the rectus femoris, biceps femoris, anterior tibialis, and gastrocnemius muscles, measured by surface electromyography (sEMG). A significant interaction effect between group and time was observed for the RMS values of the rectus femoris muscle (F[2, 96] = 3.883, P = 0.024, η2p = 0.075). Post-hoc Bonferroni comparisons revealed that T1 (P < 0.001) and T2 (P < 0.001) in the Baduanjin group were significantly higher than T0, with no significant difference between T1 and T2 (P = 1.000). In contrast, the control group showed no significant changes over time, nor any significant differences between time points. No significant interaction effects were found for the biceps femoris, anterior tibialis, or gastrocnemius muscles.

Table 3.

Repeated-measures variance analysis results of sEMG

T0 95% CI T1 95% CI T2 95% CI Group effect Time effect Group*Time effect
M ± SD M ± SD M ± SD F P η2 F P η2 F P η2
rectus femoris
 Experimental group 108.38 ± 51.52 89.56, 127.09 135.79 ± 76.89 104.91, 165.79 140.64 ± 75.51 110.16, 169.97 0.028 0.869 0.001 14.809 0.000* 0.236 3.883 0.024* 0.075
 Control group 118.11 ± 66.63 89.38, 146.86 131.13 ± 62.78 104.34, 158.49 126.54 ± 61.14 100.20, 152.95
Biceps femoris
 Experimental group 82.43 ± 45.62 60.78, 95.02 101.13 ± 64.35 73.54, 125.03 106.95 ± 58.46 81.37, 128.41 0.730 0.397 0.015 6.642 0.002* 0.122 3.075 0.051 0.060
 Control group 81.87 ± 48.32 61.36, 103.03 86.08 ± 48.79 66.60, 108.28 86.26 ± 49.29 66.58, 108.69
tibialis anterior
 Experimental group 157.95 ± 85.50 126.53, 195.92 188.23 ± 81.24 157.37, 221.93 195.31 ± 74.73 166.79, 226.12 0.824 0.369 0.017 26.488 0.000* 0.356 0.297 0.689 0.006
 Control group 140.62 ± 90.06 99.97, 177.16 163.26 ± 94.58 120.27, 201.10 173.20 ± 89.10 132.89, 209.19
Gastrocnemius
 Experimental group 57.28 ± 45.37 37.16, 71.45 80.02 ± 61.77 52.49, 102.48 95.47 ± 67.10 65.04, 119.68 0.057 0.812 0.001 10.220 0.000* 0.176 1.415 0.248 0.029
 Control group 64.88 ± 45.46 47.93, 86.07 75.63 ± 53.50 55.78, 100.59 82.31 ± 53.63 62.82, 107.46
Open in a new tab

Abbreviations: sEMG surface electromyography, RMS root mean square

* indicate statistically significant of interaction or main effect (P < 0.05)

Balance during walking was assessed using Mean Square Amplitudes (MSAs) in the X (anteroposterior), Y (mediolateral), and Z (vertical) directions via sacral-mounted 3D motion analysis (Table 4). Significant group × time interaction effects were observed only in the Y direction (F[2, 96] = 4.690, P = 0.015, η2p = 0.089). Group comparisons showed no significant differences at T0 (P = 0.428) and T1 (P = 0.696), but a significant improvement in mediolateral stability in the Baduanjin group at T2 (P = 0.047). Within both groups, pairwise comparisons revealed significant reductions in MSA-Y across time points, suggesting improved lateral balance control. No significant interactions were found in the X or Z directions.

Table 4.

Repeated-measures variance analysis results of three-dimensional gait sacral coordinate test

T0 95% CI T1 95% CI T2 95% CI Group effect Time effect Group*Time effect
M ± SD M ± SD M ± SD F P η2 F P η2 F P η2

Walking

(m)

 X
 Experimental group 1.793 ± 0.085 1.737, 1.814 1.804 ± 0.092 1.754, 1.840 1.792 ± 0.067 1.768, 1.822 1.106 0.298 0.023 0.661 0.519 0.014 0.262 0.770 0.005

Control

 group

 1.768 ± 0.080 1.729, 1.801 1.788 ± 0.062 1.757, 1.804 1.787 ± 0.059 1.760, 1.809
Y
 Experimental group 0.048 ± 0.007 0.044, 0.050 0.038 ± 0.010 0.034, 0.042 0.031 ± 0.009† 0.027, 0.034 0.425 0.517 0.009 71.806 0.000* 0.599 4.690 0.015* 0.089

Control

 group

 0.046 ± 0.010 0.042, 0.050 0.039 ± 0.011 0.034, 0.044 0.036 ± 0.010 0.032, 0.040
Z
 Experimental group 0.016 ± 0.005 0.013, 0.017 0.015 ± 0.005 0.013, 0.017 0.015 ± 0.005 0.013, 0.017 2.559 0.116 0.051 0.538 0.586 0.011 0.137 0.872 0.003

Control

 group

 0.013 ± 0.004 0.011, 0.014 0.013 ± 0.004 0.012, 0.015 0.013 ± 0.004 0.011, 0.015
Open in a new tab

Abbreviations: MSA Mean sway amplitude, X anteroposterior direction, Y mediolateral direction, Z vertical direction

* indicate statistically significant of interaction or main effect (P < 0.05)

† indicate significant differences between the pre-intervention (T0), and 3weeks (T1) or 6weeks (T2) post-intervention compared with the control group

Adherence, fatigue, and safety outcomes

Overall adherence was high in both groups. Participants in the Baduanjin group completed an average of 28.3 ± 1.8 sessions (94.3%), while those in the control group completed 28.6 ± 1.5 sessions (95.3%) out of 30 scheduled sessions. Post-session fatigue scores (0–10 scale) averaged 3.2 ± 1.4 in the experimental group, indicating mild-to-moderate levels that resolved with rest. No participant reported persistent fatigue interfering with daily activities.

No serious adverse events occurred in either group. Minor events in the experimental group included transient muscle soreness (n = 3), mild knee discomfort (n = 2), and one episode of dizziness that resolved immediately upon sitting. None required medical intervention or led to withdrawal. The control group reported comparable minor events, including muscle soreness (n = 2). These findings indicate that both interventions were well tolerated and safe.

Discussion

In this study, we evaluated 50 patients with subacute stroke and observed considerable enhancements in the BBS, FMA-LE, MBI, and MSA in the Y direction for both groups following treatment. A more marked recovery was noted in the modified Baduanjin group as contrasted with the control group, with significantly greater enhancements in the BBS, FMA-LE, MBI scores, MSA (Y direction) values, as well as rectus femoris sEMG indices compared with the control group.

BBS scores in both groups increased over the course of the intervention, indicating that both conventional rehabilitation and modified Baduanjin exercises effectively improved balance function in stroke patients. However, the experimental group showed significantly greater improvement (p < 0.05), suggesting superior efficacy of modified Baduanjin. Several factors may account for these results: First, the symmetrical movement patterns in modified Baduanjin—particularly the micro-squat posture—promote activation of lower limb muscles and enhance proprioceptive feedback. Second, the practice emphasizes controlled shifting of the center of gravity, engaging key postural control strategies such as the “hip strategy” and “ankle strategy” [36]. Third, upper limb movements like “Supporting Heaven with Palms to Regulate the Triple Burner”, incorporate diaphragmatic breathing, which increases intra-abdominal pressure and activates core stabilizers such as the transversus abdominis and rectus abdominis, thereby improving trunk stability [37]. Previous studies incorporating Berg balance as a primary outcome have reported that Baduanjin, combined with rehabilitation exercises, improves BBS scores in patients with subacute stroke [7, 38–40], consistent with our findings, which underscore the effectiveness of modified Baduanjin training in improving balance function in stroke patients.

Human balance encompasses static balance, self-initiated dynamic balance, and reactive dynamic balance. Stroke survivors often experience impaired central nervous system regulation, resulting in compromised postural control and motor coordination, leading to balance deficits. During rehabilitation, the focus is on maintaining stability when the center of gravity shifts without external forces, termed self-dynamic balance. Although various clinical tools are available for assessing balance, many fail to fully capture dynamic aspects of balance, particularly those involved in functional tasks such as walking.

To address this limitation, we employed the 3D motion analysis sacrum calibration method. In recent years, researchers globally have utilized sacral marker points (the upper edge of the sacral midline) in the 3D motion analysis system to evaluate balance stability and regulatory capacity [17, 41], streamlining the complex flow of the globally recognized standard body segmentation method [16], and detecting and collecting body COM. The parameters obtained are in close agreement with those from the body segment method [20, 42], and allow an accurate assessment of balance function and stability under various functional movements. Therefore, in this study, the 3D motion analysis sacral scaling method was employed to assess changes in self-homeostasis during walking among stroke subjects. Studies have shown that the Y-directional MSA of the sacral marker in both groups gradually decreased over time, indicating that both modified Baduanjin and routine rehabilitation training could effectively enhance coronal postural stability in stroke subjects. The MSA score of the experimental group was significantly reduced in the Y-direction, indicating an advantage over the conventional rehabilitation group to improve balance functions.

From a movement analysis perspective, the second paradigm of modified Baduanjin, —characterized by stable trunk control and dynamic left-right shifts in the center of gravity [43], strengthens the load-bearing response of the lower limbs, activates the bilateral lumboabdominal and hip muscle groups, and enhances lateral stability mechanisms in stroke patients with single-leg support. The fifth and sixth paradigms emphasize pronounced lateral flexion of the cervical and lumbar spine, thereby improving core muscle strength and pelvic flexibility. Engagement of the hip musculature during these movements contributes to improved core control, which supports better coronal plane stability of the pelvis during walking. The eighth paradigm involves a gentle heel strike following landing, which stimulates plantar proprioception; repeated execution of this action enhances lower limb weight-bearing capacity and standing balance [44]. These biomechanical features collectively account for the observed improvements in dynamic balance function, particularly the significant increase in lateral excursion amplitude during walking among participants in the experimental group.

However, no significant differences were found between groups in MSA scores along the X and Z axes at any time point. This absence of change may be attributed to limited anteroposterior displacement of the body’s center of gravity in both groups. Additionally, the modified Baduanjin simplifies the traditional squatting posture by adopting an elevated “horse stance,” reducing the demand on vertical center-of-gravity adjustments compared to conventional forms. Concurrently, the control group’s balance training primarily targeted horizontal plane movements with minimal emphasis on vertical weight shifting. These factors likely contributed to the lack of significant effects on motion symmetry in the anteroposterior and vertical directions.

As for motor function and MBI scores, both groups demonstrated significant improvements over time, but the experimental group showed superior gains. In the practice of modified Baduanjin, emphasis is placed on controlled weight shifting and sustained semi-squat positions while performing coordinated upper limb movements. This training method induces simultaneous contraction of the quadriceps and hamstrings, requiring the knee joint muscles to perform concentric and eccentric coordinated actions, thus enhancing lower limb functional movement and muscle group coordination control abilities. Additionally, the movements closely resemble common daily activities such as sitting down, standing up, bending, reaching, toileting, and dressing, thereby engaging dynamic balance mechanisms across functional tasks. Therefore, our findings indicate that the modified Baduanjin may be more effective in promoting recovery of lower limb motor function and improving independence in activities of daily living after stroke.

Surface electromyography was employed to ascertain the RMS values for lower extremity muscles. Results revealed significantly higher RMS values in the rectus femoris in the experimental group compared to the control group, suggesting that modified Baduanjin may more effectively facilitate neuromuscular recovery of the paralyzed lower limb and enhance quadriceps strength. Strengthening the knee extensors is also a key factor in improving balance function. Although the group-by-time interaction did not reach statistical significance for the biceps femoris, a trend toward improvement was observed, approaching the threshold for significance. Additionally, no significant difference in RMS values was observed between the two groups for the tibialis anterior and gastrocnemius muscles, which might be explained by the nature of the training: modified Baduanjin involves symmetrical upper limb movements performed from a slight squat position, mimicking closed kinetic chain activities. In this context, the distal lower limbs remain fixed, placing greater emphasis on coordination and control of knee joint dynamics (concentric/eccentric contractions), leading to preferential activation of proximal rather than distal musculature. Consequently, the training may elicit broader and faster recruitment of muscles surrounding the knee compared to those around the ankle.

Unlike other studies, our study is a randomized controlled trial specifically designed for modified Baduanjin tailored for individuals in the subacute phase of stroke recovery. It addresses key limitations of prior work and provides robust evidence supporting the clinical efficacy of this customized exercise regimen in improving balance function among Chinese subacute stroke survivors. Notably, to our knowledge, this is the first study to employ 3D motion analysis of sacral coordinates to objectively assess changes in dynamic balance during walking in this population, offering novel biomechanical insights. It is important to note that this study does have some limitations. First, the widespread familiarity with Baduanjin among Chinese populations, participant blinding was not feasible, which may have heightened expectations of benefit in the intervention group. To minimize bias, we implemented blinded outcome assessment, adhered to standardized intervention protocols, and incorporated objective biomechanical measurements—such as 3D motion analysis and surface electromyography (sEMG)—to strengthen the validity of our findings. Secondly, the sample size was relatively small, limiting statistical power and generalizability. Third, the 6-week intervention period captures only short-term effects and does not allow evaluation of sustained benefits after program cessation, optimal maintenance strategies, long-term community adherence, or impact on clinically important outcomes such as fall incidence and hospitalization rates—all essential for informing evidence-based practice. Fourth, the single-center design and relatively narrow inclusion criteria —specifically requiring Brunnstrom stages III–IV for both upper and lower limbs—further restrict the generalizability of findings to broader stroke populations. Fifth, the intensive nature of the intervention (80 min per day) may not be feasible in all clinical or community settings with limited resources. Finally, we did not measure long-term functional outcomes beyond the intervention period. The absence of extended follow-up assessments at 6 and 12 months post-intervention represents a significant limitation, as it precludes evaluation of the durability of treatment effects, patterns of continued practice, and potential for ongoing functional gains in home and community environments. These questions are critical for understanding the real-world applicability of modified Baduanjin in chronic stroke management. Future research should prioritize larger samples and include long-term follow-up assessments at 3, 6, and 12 months post-intervention to examine sustained efficacy, adherence patterns, and factors influencing long-term practice continuation.

This preliminary study demonstrates that modified Baduanjin is a feasible, safe, and promising adjunctive intervention for enhancing balance and functional independence in subacute stroke survivors, outperforming conventional rehabilitation alone. However, these findings require confirmation in larger, multicenter trials with longer follow-up before definitive clinical recommendations can be established. For clinicians working with subacute stroke patients who have achieved basic mobility, modified Baduanjin may serve as a valuable complementary approach—particularly for those responsive to structured physical activity or interested in traditional mind-body practices. Moreover, its adaptability makes it suitable for implementation in community and home-based rehabilitation settings, potentially promoting long-term engagement and self-management. Furthermore, the sacral calibration method in 3D motion analysis offers an objective assessment of dynamic balance during walking, warranting further dissemination in clinical practice. The inability to blind participants and therapists to group allocation may be a potential source of performance bias, a phenomenon common in behavioral intervention studies. Participants’ level of knowledge about the acceptance of a new intervention may affect their level of effort or expectations of benefits, while the therapist’s knowledge may affect the implementation of treatment or incentive effects. To mitigate these concerns, a number of safeguards were put in place: outcome assessors remained blinded throughout data collection and analysis; both groups received the same level of attention from therapists and standardized protocols with fidelity monitoring; and we utilized objective biomechanical measures (3D motion analysis, sEMG) alongside clinical scales. Nevertheless, the possibility of such performance biases cannot be totally excluded and must therefore be taken into account when interpreting the results.

Conclusion

This preliminary single-center study suggests that modified Baduanjin, when combined with conventional rehabilitation, may provide additional benefits for improving balance and lower extremity motor function in subacute stroke patients with hemiplegia. While both groups showed improvements, the modified Baduanjin group demonstrated greater gains in several outcome measures. Modified Baduanjin may therefore be considered a safe and potentially effective complementary intervention in stroke rehabilitation. However, confirmation in larger, multicenter trials with longer follow-up periods is necessary before definitive clinical recommendations can be made.

Supplementary Information

Supplementary Material 1. (197.5KB, doc)

Acknowledgements

We express our gratitude to all the patients involved in this study, heartfelt appreciation for the guidance and assistance provided by the Baduanjin coach, and also thank the research team members who contributed to data collection and other aspects.

Abbreviations

3D

Three-dimensional

BBS

Berg Balance Scale

FMA-LE

Fugl-Meyer Assessment of Lower Extremity

MBI

Modified Barthel Index

sEMG

Surface Electromyogram

MSA

Mean Sway Amplitudes

COM

Center of Mass

Authors’ contributions

WZ, WY, MW, and LC contributed to the conception and design of the study; PC, YC, and GD were responsible for intervention and assessment; GH conducted data acquisition and analysis; WZ, GH, and TW participated in drafting and revising the manuscript. All authors reviewed, revised, and approved the final version of the article.

Funding

This work is supported by the Jiangsu Provincial Program for TCM Science and Technology Development (ZX2021A1), Youth Project of Wuxi Health Committee (Q202444 and T202525), Wuxi Municipal Health Commission Major Research Project Z202512.

Data availability

The datasets generated during and/or analysed during the this study can be obtained by the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committee of Wuxi Mental Health Center (Approval No. WXMHCIRB2023LLKY046) and registered with the Chinese Clinical Trial Registry (ChiCTR2300069394). All participants provided written informed consent prior to enrollment.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Weiwei Zhao and Wang Yao contributed equally to this work and shared first authorship.

Contributor Information

Weiwei Zhao, Email: 13814284945@163.com.

Guilan Huang, Email: huangguilan211@163.com.

Lan Chen, Email: chenlan080501@163.com.

Tong Wang, Email: wangtong60621@163.com.

References

  • 1.Yao T, Wang N, Chen Z. Evidence-based medical research on motor function training for stroke patients. Chin J Rehabilitation Med. 2010;25(6):565–70. [Google Scholar]
  • 2.Tyson SF, Hanley M, Chillala J, Selley A, Tallis RC. Balance disability after stroke. Phys Ther. 2006;86(1):30–8. [DOI] [PubMed] [Google Scholar]
  • 3.Harris JE, Eng JJ, Marigold DS, Tokuno CD, Louis CL. Relationship of balance and mobility to fall incidence in people with chronic stroke. Phys Ther. 2005;85(2):150–8. [PubMed] [Google Scholar]
  • 4.van Duijnhoven HJ, Heeren A, Peters MA, Veerbeek JM, Kwakkel G, Geurts AC, et al. Effects of Exercise Therapy on Balance Capacity in Chronic Stroke: Systematic Review and Meta-Analysis. Stroke. 2016;47(10):2603–10. [DOI] [PubMed] [Google Scholar]
  • 5.Erickson KI, Voss MW, Prakash RS, Basak C, Szabo A, Chaddock L, et al. Exercise training increases size of hippocampus and improves memory. Proc Natl Acad Sci USA. 2011;108(7):3017–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Zheng G, Chen B, Fang Q, Lin Q, Tao J, Chen L. Baduanjin exercise intervention for community adults at risk of ischamic stroke: A randomized controlled trial. Sci Rep. 2019;9(1):1240. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Bai Y, Mao H, Guo J, Zhang M. Study on improving balance function of stroke subjects by combining Ba Duan Jin with functional training. J Tradit Chin Med. 2011;26(10):1231–2. [Google Scholar]
  • 8.Wu F-L, Yang Q, Jiang J, Yu J, Jin Y-C. Targeted exercise interventions on stress, anxiety, depression, and sleep disorders in PhD students. World J Psychiatry. 2025;15(12):109558. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Zunin ID, Ring AJ, Sakamoto AM, Haverkort-Yeh RD. Effects of Tummo Meditation and Niguma Yoga on Brain Activity. Hawaii J Health Soc Welf. 2025;84(9):195–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Sun J, Yao K, Zhao R, Li H, Cicchella A. Tai Chi as a preventive intervention for improving mental and physical health in non-depressed college students with high perceived stress. Front Public Health. 2025;13:1613384. [DOI] [PMC free article] [PubMed]
  • 11.Yu X-M, Jin X-M, Lu Y, Gao Y, Xu H-C, Xue X, et al. Effects of Body Weight Support-Tai Chi Footwork Training on Balance Control and Walking Function in Stroke Survivors with Hemiplegia: A Pilot Randomized Controlled Trial. Evidence-Based Complement Altern Med. 2020;2020(1):9218078. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Miller SM, Hui-Lio C, Taylor-Piliae RE. Health Benefits of Tai Chi Exercise: A Guide for Nurses. Nurs Clin North Am. 2020;55(4):581–600. [DOI] [PubMed] [Google Scholar]
  • 13.Walter AA, Van Puymbroeck M, Bosch P, Schmid AA. Complementary and integrative health interventions in post-stroke rehabilitation: a systematic PRISMA review. Disabil Rehabil. 2022;44(11):2223–32. [DOI] [PubMed] [Google Scholar]
  • 14.Sun Z, Wang T. Application progress of three-dimensional motion analysis system in rehabilitation medicine assessment and testing. Chin J Rehabilitation Med. 2018;33(2):234–8. [Google Scholar]
  • 15.McKinon W, Hartford C, Di Zio L, van Schalkwyk J, Veliotes D, Hofmeyr A, et al. The agreement between reaction-board measurements and kinematic estimation of adult male human whole body centre of mass location during running. Physiol Meas. 2004;25(6):1339–54. [DOI] [PubMed] [Google Scholar]
  • 16.Gullstrand L, Halvorsen K, Tinmark F, Eriksson M, Nilsson J. Measurements of vertical displacement in running, a methodological comparison. Gait Posture. 2009;30(1):71–5. [DOI] [PubMed] [Google Scholar]
  • 17.Kerrigan DC, Viramontes BE, Corcoran PJ, LaRaia PJ. Measured versus predicted vertical displacement of the sacrum during gait as a tool to measure biomechanical gait performance. Am J Phys Med Rehabil. 1995;74(1):3–8. [DOI] [PubMed] [Google Scholar]
  • 18.Zhu X, Zhu Y, Wang S, Meng D, Wang T. Reliability and validity of a new balance test with three-dimensional motion analysis system. Chin J Rehabilitation Med. 2012;27(4):315–9. [Google Scholar]
  • 19.Napier C, Jiang X, MacLean CL, Menon C, Hunt MA. The use of a single sacral marker method to approximate the centre of mass trajectory during treadmill running. J Biomech. 2020;108:109886. [DOI] [PubMed] [Google Scholar]
  • 20.Zhu Y, Zhu X, Meng D, Wang S, Wang T. A comparison and analysis on the consistencies of three kinds of measurement methods with three-dimentional-motion analysis system in location of center of mass of healthy person for balance assessment. Chin J Rehabilitation Med. 2012;27(3):216–22. [Google Scholar]
  • 21.Yang F, Pai YC. Can sacral marker approximate center of mass during gait and slip-fall recovery among community-dwelling older adults? J Biomech. 2014;47(16):3807–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Broglio K. Randomization in Clinical Trials: Permuted Blocks and Stratification. JAMA. 2018;319(21):2223–4. [DOI] [PubMed] [Google Scholar]
  • 23.Suresh K. An overview of randomization techniques: An unbiased assessment of outcome in clinical research. J Hum reproductive Sci. 2011;4(1):8–11. [DOI] [PMC free article] [PubMed] [Google Scholar] [Retracted]
  • 24.Yuen M, Ouyang HX, Miller T, Pang MYC. Baduanjin Qigong Improves Balance, Leg Strength, and Mobility in Individuals With Chronic Stroke: A Randomized Controlled Study. Neurorehabilit Neural Repair. 2021;35(5):444–56. [DOI] [PubMed] [Google Scholar]
  • 25.Koch G, Bonnì S, Casula EP, Iosa M, Paolucci S, Pellicciari MC, et al. Effect of Cerebellar Stimulation on Gait and Balance Recovery in Patients With Hemiparetic Stroke: A Randomized Clinical Trial. JAMA Neurol. 2019;76(2):170–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Chen J, Chen Q, Chen C, Li S, Liu L, Wu C, et al. Effects of modified Ba Duan Jin physical activity on cardiopulmonary function, motor function and daily living activities in stroke patients. hinese J Rehabilitation Theory Pract. 2024;30(1):74–80. [Google Scholar]
  • 27.Meseguer-Henarejos AB, Rubio-Aparicio M, López-Pina JA, Carles-Hernández R, Gómez-Conesa A. Characteristics that affect score reliability in the Berg Balance Scale: a meta-analytic reliability generalization study. Eur J Phys Rehabil Med. 2019;55(5):570–84. [DOI] [PubMed] [Google Scholar]
  • 28.Maeda N, Urabe Y, Murakami M, Itotani K, Kato J. Discriminant analysis for predictor of falls in stroke patients by using the Berg Balance Scale. Singapore Med J. 2015;56(5):280–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Hernández ED, Forero SM, Galeano CP, Barbosa NE, Sunnerhagen KS, Alt Murphy M. Intra- and inter-rater reliability of Fugl-Meyer Assessment of Lower Extremity early after stroke. Braz J Phys Ther. 2021;25(6):709–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Kwong PWH, Ng SSM. Cutoff Score of the Lower-Extremity Motor Subscale of Fugl-Meyer Assessment in Chronic Stroke Survivors: A Cross-Sectional Study. Arch Phys Med Rehabil. 2019;100(9):1782–7. [DOI] [PubMed] [Google Scholar]
  • 31.Shah S, Vanclay F, Cooper B. Improving the sensitivity of the Barthel Index for stroke rehabilitation. J Clin Epidemiol. 1989;42(8):703–9. [DOI] [PubMed] [Google Scholar]
  • 32.Wang S, Zhu Y, Zhu X, Meng D, Wang T. Validity of sacrum-marked method on evaluating the third-level balance performance of health people and hemiplegic patients. Chin J Rehabilitation Med. 2012;27(3):223–6. [Google Scholar]
  • 33.Rainoldi A, Bullock-Saxton JE, Cavarretta F, Hogan N. Repeatability of maximal voluntary force and of surface EMG variables during voluntary isometric contraction of quadriceps muscles in healthy subjects. J Electromyogr kinesiology: official J Int Soc Electrophysiological Kinesiol. 2001;11(6):425–38. [DOI] [PubMed] [Google Scholar]
  • 34.Jing C, Li K, Li Z, Sun Y, Wu J, Li Y, et al. Tai Chi postural training for dyskinesia rehabilitation: a study protocol for a randomised controlled trial in convalescent ischaemic stroke patients. BMJ open. 2021;11(5):e046003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Song R, Park M, Jang T, Oh J, Sohn MK. Effects of a Tai Chi-based stroke rehabilitation program on symptom clusters, physical and cognitive functions, and quality of life: a randomized feasibility study. Int J Environ Res Public Health. 2021;18(10):5453. [DOI] [PMC free article] [PubMed]
  • 36.Morasso P. Integrating ankle and hip strategies for the stabilization of upright standing: An intermittent control model. Front Comput Neurosci. 2022;16:956932. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Kim SH, Shin HJ, Cho HY. Impact of types of breathing on static balance ability in healthy adults. Int J Environ Res Public Health. 2022;19(3):1205. [DOI] [PMC free article] [PubMed]
  • 38.Zhang M, Guo J, Bai Y, Feng X. Study on the effect of Ba Duan Jin combined with rehabilitation exercise on balance function of stroke subjects. Lishizhen Med Materia Med Res. 2013;24(10):2438–9. [Google Scholar]
  • 39.Zhang Y, Li L. Ba Duan Jin combined with balance training in the treatment of 31 cases of balance dysfunction in stroke subjects. Shandong J Traditional Chin Med. 2016;35(8):716–8. [Google Scholar]
  • 40.Liu W, Zhao Y, Yang D, Fu Q, Zhou J. Effect of Ba Duan Jin exercise prescription on self-dynamic balance in patients with stroke recovery. Lishizhen Med Materia Med Res. 2022;33(8):1936–9. [Google Scholar]
  • 41.Domagalska-Szopa M, Szopa A. Gait Pattern Differences Among Children With Bilateral Cerebral Palsy. Front Neurol. 2019;10:183. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Wang S, Zhu X, Zhu Y, Meng D, Wang T. Reliability and validity analysis of sacral calibration method for detecting balance function in stroke hemiplegic subjects. Chin J Phys Med Rehabilitation. 2013(4):273–7.
  • 43.Li M, Lan X, Yan X, Lin Q, Zheng G. The influence of Health Qigong Ba Duan Jin on the balance function of college students. J Fujian Univ Traditional Chin Med. 2015;25(1):14–8. [Google Scholar]
  • 44.Dursun Ö, Mavuş AB. Effect of plantar vibration on ankle proprioception in chronic stroke patients: a randomized controlled trial. Physiother Theory Pract. 2025;41(11):2317-27. [DOI] [PubMed]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1. (197.5KB, doc)

Data Availability Statement

The datasets generated during and/or analysed during the this study can be obtained by the corresponding author upon reasonable request.

Articles from BMC Complementary Medicine and Therapies are provided here courtesy of BMC

RESOURCES